A deeper insight into direct trifluoromethoxylation with trifluoromethyl triflate

Article abstract of DOI:10.1016/j.jfluchem.2009.11.006

Commercially available fluorides (silver fluoride and n-tetrabutylammonium triphenyldifluorosilicate), combined with TFMT, allow a simple generation, in situ, of silver and n-tetrabutylammonium trifluoromethoxides which were able to react with electrophilic substrates. Silver trifluoromethoxide, which is usually more efficient than n-tetrabutylammonium trifluoromethoxide, converts, under mild conditions, primary aliphatic bromides and iodides, as well as primary and secondary benzylic or allylic bromides to the corresponding trifluoromethoxylated compounds. Several trifluoromethyl ethers, which could be valuable building-blocks, were prepared in such a way.

Full text of DOI:10.1016/j.jfluchem.2009.11.006

Journal of Fluorine Chemistry 131 (2010) 200–207
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
A deeper insight into direct trifluoromethoxylation with trifluoromethyl triflate
Olivier Marrec a,b,c,d,e,f, Thierry Billard a,b,c,d,e,f, Jean-Pierre Vors g, Sergii Pazenok g,
Bernard R. Langlois
,
*
a,b,c,d,e,f,**
a ICBMS, Institut de Chimie et Biochimie Mole´culaires et Supramole´culaires, Equipe SERCOF, 43 boulevard du 11 novembre 1918, Villeurbanne, F-69622, France
b CNRS, UMR5246, Villeurbanne, F-69622, France
c
´
Universite de Lyon, Lyon, F-69622, France
d
´
Universite Lyon 1, Lyon, F-69622, France
e INSA-Lyon, Villeurbanne, F-69622, France
f CPE Lyon, Villeurbanne, F-69616, France
g Bayer CropScience AG, Product Technology, Process Research, Building 6550, Alfred-Nobel-Str. 50, 40789 Monheim am Rhein, Germany
A R T I C L E I N F O
A B S T R A C T
Article history:
Commercially available fluorides (silver fluoride and n-tetrabutylammonium triphenyldifluorosilicate),
combined with TFMT, allow a simple generation, in situ, of silver and n-tetrabutylammonium
trifluoromethoxides which were able to react with electrophilic substrates. Silver trifluoromethoxide,
which is usually more efficient than n-tetrabutylammonium trifluoromethoxide, converts, under mild
conditions, primary aliphatic bromides and iodides, as well as primary and secondary benzylic or allylic
bromides to the corresponding trifluoromethoxylated compounds. Several trifluoromethyl ethers, which
could be valuable building-blocks, were prepared in such a way.
Received 1 September 2009
Received in revised form 9 November 2009
Accepted 11 November 2009
Available online 17 November 2009
Keywords:
Trifluoromethoxylation
ß 2009 Elsevier B.V. All rights reserved.
Trifluoromethoxide
Trifluoromethyl triflate
Trifluoromethyl trifluoromethanesulfonate
Trifluoromethoxy-substitued compounds
1. Introduction
substituent looks like chlorine [9] in the sense that it is electron-
withdrawing by induction (
but more than chlorine ( I = +0.47 [11a]), and electron-donating by
resonance (
R = À0.13 to À0.18 [11]), but less than chlorine
x = 3.7 [10], sI = +0.51 to +0.60 [11]),
The intrinsic properties of fluorine atom (small size, high
electronegativity, formation of strong C-F bonds) induce dramatic
changes in the electronic, steric and hydrophobic parameters of the
molecules which bear fluorinated moieties. Thus, their pharmaco-
dynamic and pharmacokinetic properties are deeply modified
[1,2]. This is the reason why fluorine chemistry is now so popular in
the life sciences.
Among the fluorinated moieties currently used, the trifluor-
omethoxy group (OCF3) becomes more and more prominent [3,4].
For example, in addition to trifluoromethoxy-substituted liquid
crystals [5] and dyes [6], several trifluoromethoxylated major
pesticides and pharmaceuticals are now present on the market
[5,7,8] (Fig. 1).
s
s
(
s
R = À0.25 [11a]). For this reason, it has been named ‘‘super-
halogen’’ [12] or ‘‘pseudo-halogen’’ [13]. On the other hand, OCF3 is
one of the most hydrophobic substituent, just after SCF3, as
indicated by its Hansch-Leo parameter (PR (SCF3) = +1.44, PR
(OCF3) = +1.04, PR (CF3) = +0.88, PR (OCH3) = À0.02) [14]. Thus,
such a hydrophobic substituent dramatically increases the
bioavailability of the products bearing it.
Another characteristic of the OCF3 moiety is that the electron
density of the non-bonding p-orbitals of oxygen is very low. The
first evidence of that came from the study of the UV spectrum of
(trifluoromethoxy)benzene, which is very similar to that of
(trifluoromethyl)benzene [15,16]. Later, Anderson speculated that
‘‘the p atomic orbital of oxygen is drawn towards the CF3 group by
acceptance of its p-electrons into the antibonding orbitals of the
perfluoroalkyl group’s C–F bonds’’ [17]. The first consequence is
that bonding/non-bonding resonance and ionic limiting structures
cannot be written for OCF3, despite tiny differences in bond lengths
sometimes observed between experiment and calculation (the
accuracy of which were not precised). As far as aryl trifluoromethyl
ethers are concerned, the second consequence is that oxygen non-
bonding orbitals are not conjugated with the aromatic nucleus, as
This growing interest for trifluoromethyl ethers is related to the
very peculiar characteristics of the CF3O group. On one hand, this
* Corresponding author. Tel.: +49 2173384808; fax: +49 2173384880.
´
** Corresponding author at: ICBMS, Institut de Chimie et Biochimie Moleculaires
´
et Supramoleculaires, Equipe SERCOF, 43 boulevard du 11 novembre 1918,
Villeurbanne, F-69622, France.
E-mail addresses: sergiy.pazenok@bayercropscience.com (S. Pazenok),
Bernard.Langlois@univ-lyon1.fr (B.R. Langlois).
0022-1139/$ – see front matter ß 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2009.11.006

O. Marrec et al. / Journal of Fluorine Chemistry 131 (2010) 200–207
201
Fig. 1. Some bioactive trifluoromethoxylated products.
Scheme 1. Chlorination/fluorination sequence of phenol derivatives.
also deduced from molecular photoelectron spectroscopy [18] or
dipole moments studies [19,20,11b]. Thus, the O–CF3 group of Ar–
OCF3 can rotate freely out of the nucleus plane. Consequently, in
order to minimize electronic repulsions, (trifluoromethoxy)ben-
zene adopts a conformation in which the O–CF3 bond is orthogonal
to the nucleus plane (Fig. 2). Such a conformation was first
anticipated from modelling studies [11b] and NMR spectroscopy
[21], then directly observed by X-ray spectroscopy [22,23].
Of course, this conformation is suspected to be important in the
interaction between trifluoromethoxylated substrates and their
biological receptors. Moreover, it could explain the strong para-
orientating effect observed in electrophilic substitution of (tri-
fluoromethoxy)benzene. For the previously mentioned reasons,
such an orientating effect cannot be due to conjugation between
chlorine/fluorine exchange on trichlorinated precursors, action
of sulfur tetrafluoride on fluoroformates (Sheppard’s method),
Hiyama’s oxidative fluorodesulfurization, electrophilic trifluoro-
methylation of hydroxyl functions, and nucleophilic trifluoro-
methoxylation.
Chlorination/fluorination was developed as soon as 1955 by
Yagupolskii et al. [24], a research group which remained on the fore
front of CF3O chemistry until now. This method was improved
when Yarovenko and Vasileva, from the same Ukrainian laborato-
ry, discovered a new access to aryl trichloromethyl ethers from aryl
thionochloroformates [25] (Scheme 1). However, this chlorination/
fluorination technique, which is currently employed on the
industrial scale, can be only applied to aromatics. Moreover,
Yarovenko’s modification is difficult to scale-up because of the
high percutaneous toxicity of aryl thionochloroformates [26].
The three following methods can be, in principle, applied to
aromatic as well as aliphatic substrates (Scheme 2). Though
deoxyfluorination of fluoroformates [27,11a] is greatly deserved by
the high toxicity of SF4 (and that of fluorophosgene when
fluoroformates are prepared from it), oxidative fluorodesulfuriza-
tion [28] opened a new access to trifluoromethyl ethers, which can
be used without specific equipment and successfully applied to
oxygen and the
P-system, though sometimes written. More
probably, as the very electronegative fluorine atoms are sur-
rounded by an intense electric field, a strong coulombian repulsion
occurs between the fluorine atoms (one of which being always
close from the nucleus) and the mobile
P-electrons which are
repelled as far as possible (+I effect), that means towards the
P
para-position of the cycle which, consequently, exhibits the higher
electron density (Fig. 2). Through-space field interactions were
effectively anticipated by Sheppard from measurements of the
Dewar constant (F) of the OCF3 group [12].
Until now, essentially aryl trifluoromethyl ethers are described.
Their different preparations have been recently reviewed [7] and
can be divided into five methods that are, chronologically,
Scheme 2. Conversion of alcohols into trifluoromethyl ethers.
Fig. 2. Conformation and through-space interactions in (trifluoromethoxy)benzene.

202
O. Marrec et al. / Journal of Fluorine Chemistry 131 (2010) 200–207
is not an electrophilic trifluoromethylating agent and is attacked
by nucleophiles, especially hard ones, at its harder electrophilic
site, that is sulfur (VI). For example, fluorides gave rise to
trifluoromethanesulfonyl fluoride and CF3OÀ, which usually
collapses into C(O)F2 and FÀ, so that catalytic amounts of FÀ can
cause the complete decomposition of TFMT [39c]. Taking advan-
tage of this observation, Kolomeitsev et al. performed the same
reaction with bulky fluorides in order that the generated
trifluoromethoxide salt was stable enough to substitute some
leaving groups on reactive substrates (Scheme 4).
Scheme 3. Decomposition of the trifluoromethoxide anion.
Several trifluoromethoxides were generated from several
fluorides but their relative stabilities were not rationalized and
their reactivities are very difficult to compare since, if different
substrates were opposed to some of them, only ethyl 2-
(trifluoromethanesulfonyloxy)-propionate has been reacted with
all the prepared trifluoromethoxide. Thus, the scope of the method
does not seem very clear and other potentialities have to be
explored.
Scheme 4. Generation of trifluoromethoxide from TFMT.
phenols, primary and even secondary alcohols. Nevertheless, this
latter method is not tolerant with many functional substituents.
Electrophilic trifluoromethylation of alcohols and phenols has
been first reported by Umemoto et al. [29], using O-trifluoromethyl
dibenzyl furanium salts, but these trifluoromethylating agents
cannot be isolated and must be prepared in situ and at low
temperature from trifluoromethylated biphenyls. An outstanding
breakthrough in electrophilic trifluoromethylation has been
recently brought by Togni et al. [30] with easily accessible
trifluoromethyl-containing iodine (III) reagents. When opposed to
phenols, aromatic trifluoromethylation competes with O-trifluor-
omethylation [31] but, in the presence of sub-stoichiometric
amounts of zinc triflimide, such reagents allow trifluoromethyla-
tion of aliphatic alcohols [32], though a large excess of the latter is
required.
Obviously, the best synthesis of trifluoromethyl ethers would
be the direct introduction of the whole OCF3 moiety. Apart the
addition of trifluoromethyl hypofluorite upon olefins [33], which is
highly hazardous and toxic, numerous attempts to generate
trifluoromethoxide salts failed since, generally, this anion collapses
into fluoride and fluorophosgene, even at low temperature
(Scheme 3).
As already mentioned, we started working on the same
reaction, concomitantly but totally independently, more than 1
year before Kolomeitsev’s paper appeared. Our major goal was to
compare the reactivities of different trifluoromethoxides, prepared
from TFMT and different nucleophiles, towards different classes of
electrophiles, especially alkyl halides which are more accessible
and less expensive than alkyl triflates.
2. Results and discussion
The first attempt to activate TFMT with phosphines, instead of
fluoride, was unsuccessful: if triphenylphosphine was able to react
with TFMT, as expected, this reaction led exclusively to Ph3PF2.
Also, activation of TFMT with chlorides was not satisfying. For
example, it was expected that benzyltriphenylphosphonium
chloride, prepared in situ from benzyl chloride and triphenylpho-
sphine, could generate triflyl chloride and CF3OÀ (BnPPh3)+ and,
finally, provide benzyl trifluoromethyl ether. The result was
different since triflyl fluoride was the major product, along traces
of triflyl chloride but not any trace of BnOCF3 was detected. That
means that the benzyltriphenylphosphonium cation, though
bulky, is not able to stabilize CF3OÀ and, maybe, that bulkiness
is not the only parameter involved in the stabilization of CF3OÀ.
Thus, we focused our efforts on the use of n-tetrabutylammo-
Several teams tried to draw this equilibrium towards CF3OÀ and
demonstrated that, if fluorides are assoÀciated with rather bulky
cations (CsF [34,35], (Me2N)3S+ Me3SiF2 [36,37a]), the resulting
trifluoromethoxide could be stable enough, in solution, to react
with reactive electrophiles such as benzyl bromide [35] as well as
primary triflates and bromides [37]. Nevertheless, the develop-
ment of this reaction is limited by the high toxicity of gaseous
fluorophosgene which must be used under pressure.
nium triphenyldifluorosilicate (TBAT, DeShong’s reagent),
a
commercially available, cheap, anhydrous and bulky fluoride
which was not used by Kolomeitsev, as we remarked later. Silver
fluoride was also evaluated since, as our goal was to substitute
alkyl halides, it was anticipated that precipitation of silver halides
could contribute to success.
Indeed, the corresponding methoxides (Bu4N+ CF3OÀ, Ag+
CF3OÀ) were formed but they were not reactive enough towards
chlorides, except benzyl chloride and allyl chloroformate which
reacted with CF3OAg in a medium yield (Scheme 5). Other
chlorides, such as aroyl chlorides, benzenesulfonyl chloride and
chlorotrimethylsilane, led quite exclusively to the corresponding
fluorides. Probably, silver trifluoromethoxide was too soft to react
Recently, Kolomeitsev et al. [38] have described, during our
own work on the same topics, a more convenient generation of
trifluoromethoxide salts from trifluoromethyl triflate (TFMT),
which is a volatile liquid, far less toxic than fluorophosgene. It is
usually prepared from triflic acid or anhydride [17,39] and is
commercially available on the bench scale. Indeed, it was known
for more than two decades that, in contrast to other triflates, TFMT
Scheme 5. Reaction of CF3OAg with chlorides.

O. Marrec et al. / Journal of Fluorine Chemistry 131 (2010) 200–207
203
Table 1
Trifluoromethoxylation of alkyl bromides with nBu4NOCF3 or AgOCF3
.
Entry
1
Substrate
Method A nBu4N+ ÀOCF3
Substrat eq.
Method B Ag+ ÀOCF3
Substrat eq.
Yielda
Yielda
1 eq., 1.5 eq., 2 eq.
45%, 60%, 75%
0.9 eq., 1.5 eq.
100% (76%), 100%
Br
2
1.5 eq.
53%
0.9 eq.
57% (47%)
Br
3
4
Not examined
1.5 eq., 2 eq.
0.9 eq.
1%, 2%
80%
Br
0.9 eq.
50%
Br
8
5
6
0.9 eq.
0.9 eq.
36%
41%
0.9 eq.
0.9 eq.
70% (45%)
O
Br
OBn
47%
O
O
Br
7
0.9 eq.
58%
0.9 eq.
60% (58%)
O
Br
Br
N
O
Ph
8
9
1.5 eq.
1.5 eq.
1%c
Not examined
0.9 eq.
100%
92% (42%), Mono-F 8%b
O
Br
10
0.9 eq.
11%
0.9 eq.
74% (50%)
Br
11
12
0.9 eq.
40%
0.9 eq.
11%
44%
Br
Not examined
0.9 eq.
O
Br
OEt
Br
13
14
1.5 eq.
1.5 eq.
0%
0.9 eq.
0.9 eq.
2%
a: 3%,
b
: 24%
b: 97% (75%)
O
Br
OAc
AcO
AcO
OAc

204
O. Marrec et al. / Journal of Fluorine Chemistry 131 (2010) 200–207
Table 1 (Continued )
Entry
Substrate
Method A nBu4N+ ÀOCF3
Method B Ag+ ÀOCF3
Substrat eq.
0%
Substrat eq.
Yielda
Yielda
15
Not examined
0.9 eq.
O
Br
OEt
a
Crude yield determined from 19F NMR with PhCF3 as internal standard. Isolated yield in parentheses.
Mono-F = monofluorinated compound without any CF3O moiety.
b
c
Elimination reaction may predominate since styrene has been detected.
Table 2
Trifluoromethoxylation of alkyl iodides with nBu4NOCF3 or AgOCF3
O O
O O
Q+ F-
R-I
S
CF3
O
S
+
F3C O Q
F3C
O
R
CF3
F3C
F
-30 °C
2h
-30 °C then r.t.
Q+ = Bu4N+, Ag+
.
Entry
1
Substrate
Method A nBu4N+ ÀOCF3
Substrat eq.
Method B Ag+ ÀOCF3
Yielda
4%
Substrat eq.
0.9 eq.
Yielda
0.9 eq.
64% (56%)
I
2
0.9 eq.
33%
0.9 eq.
85% (69%)
60% (40%)
O
I
I
3
4
0.9 eq.
35%
0.9 eq.
O
O
O
Not examined
0.9 eq.
84% (80%)
O
O
I
N
Ph
5
6
0.9 eq.
40%
0.9 eq.
0%
22%
I
Not examined
0.9 eq.
I
a
Crude yield determined from 19F NMR with PhCF3 as internal standard. Isolated yield in parentheses.
with chlorinated substrates. It can be also noticed that with
benzeneselenenyl chloride as substrate, decomposition was only
observed.
The results were far more satisfying with alkyl bromides, as
shown in Table 1.
Generally speaking, alkyl bromides reacted more efficiently
than alkyl chlorides and, as expected, silver trifluoromethoxide
offered better yields than n-tetrabutylammonium trifluorometh-
oxide.
The resulting products could be valuable starting materials for
more sophisticated trifluoromethoxylated compounds.
Both silver trifluoromethoxide and n-tetrabutylammonium
trifluoromethoxide were able to substitute secondary benzylic
bromides (entry 10) and secondary allylic bromides (entry 11),
though with lower yields, but very poor yields were obtained from
secondary aliphatic bromides (entries 12 and 13) and no reaction
was observed with tertiary bromides (entry 15).
It can be also mentioned that, in sharp contrast with benzoyl
chloride, benzoyl bromide led to trifluoromethyl benzoate in good
yields, through an addition-elimination process, whatever the
trifluoromethoxide employed (entry 9). The reaction with 2,3,4,6-
Indeed, excellent yields were obtained from primary bromides
(entries 1–7), except phenethyl bromide for which
b-elimination
was probably the major reaction (entry 8). Obviously, primary
benzyl and allyl bromides were the most reactive (entries 1–3) but,
tetra-O-acetyl-a-D-glucopyranosyl bromide (entry 14) is worthy
interestingly, primary bromine atoms in
a
position to a carbonyl
to mention since this substrate has been already reacted with TAS+
CF3OÀ [37a] and, under these conditions, delivered the expected
function could be substituted in a satisfactory way (entries 5–7).

O. Marrec et al. / Journal of Fluorine Chemistry 131 (2010) 200–207
205
Scheme 6. Trifluoromethoxylation of N-benzyl trifluoromethanesulfonimide.
b
-trifluoromethyl glucoside (14%) along a far larger quantity of
glucosyl fluoride (34%). In our hands, this substrate gave a higher
yield of -trifluoromethyl glucoside with n-tetrabutylammonium
trifluoromethoxide (27%, = 1:8) and an excellent yield with
silver trifluoromethoxide which, on the other hand, offered a
complete stereoselectivity in favor of the stereomer.
b
-
electrophilic sulfur (VI) of TFMT, but this hypothesis does not seem
very realistic to explain such a sharp drop in the results. It could be
also supposed that, instead of substituting acetate or mesylate,
CF3OÀ Bu4N+ could be quenched by the carbonyl function or the
sulfonyl function of these ‘‘leaving groups’’, without evolution of
the system until work-up.
b
a:b
b
The reactivity of alkyl iodides was also examined (Table 2).
As usual, alkyl iodides delivered better results than alkyl
bromides (i.e. Table 2, entry 2 vs. Table 1, entry 6; Table 2, entry 4
vs. Table 1, entry 7) and, again, CF3OAg was more efficient than
CF3ONBu4, except with geranyl iodide (entry 5). This apparent
contradiction could be related to the moderate stability of this
substrate which would be decreased by the Lewis acidity of the
silver cation. The same kind of reason could explain that
substitutive trifluoromethoxylation was slightly better with
benzyl and allyl bromides than with the corresponding iodides.
As far as neomenthyl iodide is concerned, a hindered secondary
iodide, no reaction occurred.
After bromides and iodides, the reactivity of some other leaving
groups towards Q+ CF3OÀ (Q+ = Bu4N+, Ag+) was examined.
Aromatic nucleophilic substitution of 2-fluoropyridine and penta-
fluoropyridine failed but N-benzyl trifluoromethanesulfonimide
was quite quantitatively transformed by n-tetrabutylammonium
trifluoromethoxide into benzyl trifluoromethyl ether. In must be
noticed that, in this case, silver trifluoromethoxide provided a far
lower yield than n-tetrabutylammonium trifluoromethoxide
(Scheme 6). This result contrasts with those of the substitutive
trifluoromethoxylation of alkyl halides with the same reagents.
This result was rationalized by the fact that, because of the
oxophilic character of Ag+, silver fluoride (which is probably in
equilibrium with CF3OAg) could partly reacted with the triflyl
moieties of the N(SO2CF3)2 leaving group and generate triflyl
fluoride.
Comparatively, it must be reminded that Kolomeitsev et al. [38]
reported a successful substitution of alkyl triflates (which bear a
more electrophilic sulfur than mesylates) by tetramethylammo-
nium and triethylammonium trifluoromethoxides, which look
very similar to n-tetrabutylammonium trifluoromethoxide. How-
ever, in Kolomeitsev’s experiments, it seems (though not clearly
mentioned) that trifluoromethoxides were prepared in situ and
separated from triflyl fluoride prior addition of the electrophilic
substrate. Because of the announced trifluoromethoxylation
yields, this also implies that the preformed trifluoromethoxides
were not contaminated at all by fluoride residues. Are these
experimental differences significative enough to explain such
changes? Nevertheless, such intermediate treatments and pur-
ifications make Kolomeitsev’s process and equipment more
complicated than ours.
In conclusion, the present work confirms that trifluoromethyl
triflate can be used as a generator of trifluoromethoxide anions
when activated by fluoride anions, of course, but also by chlorides
or phosphines. Nevertheless, such trifluoromethoxides, which
collapse rapidly into fluoride and fluorophosgene when ‘‘free’’,
must be stabilized by bulky counter-cations to have a chance to
trifluoromethoxylate electrophilic substrates, through nucleophil-
ic substitution or addition-elimination, and deliver trifluoromethyl
ethers. Moreover, the reaction must be carried out in a closed
vessel to counterbalance the possible decomposition of CF3OÀ into
FÀ and gaseous fluorophosgene. Commercially available fluorides
(silver fluoride and n-tetrabutylammonium triphenyldifluorosili-
cate), combined with TFMT, allow a simple generation, in situ, of
silver and n-tetrabutylammonium trifluoromethoxides which
were able to react with electrophilic substrates. Silver trifluor-
omethoxide, which is usually more efficient than n-tetrabuty-
lammonium trifluoromethoxide, converts, under mild conditions,
primary aliphatic bromides and iodides as well as primary and
secondary benzylic or allylic bromides. Several trifluoromethyl
So, in order to avoid such a side-reaction, the substitution of
cinnamyl acetate and citronellyl mesylate was tested with n-
tetrabutylammonium trifluoromethoxide only. Surprisingly, tri-
fluoromethoxylation of these two substrates completely failed. Of
course, it could be argued that even Bu4NF (in equilibrium with
CF3OÀ Bu4N+) can attack the carbonyl or the sulfonyl moieties of
these starting materials, prior attacking the harder and far more
Fig. 3. Prepared trifluoromethyl ethers.

206
O. Marrec et al. / Journal of Fluorine Chemistry 131 (2010) 200–207
1
ethers, which could be valuable building-blocks, were prepared in
such a way (Fig. 3).
120.1 (q, JC-F = 256.3). 19F NMR: À57.48 (s). Anal: Calcd for
C8H5F3O2: C (50.54), H (2.65); Found: C (50.22), H (2.36).
3.1.4. (2E)-3-phenylprop-2-en-1-yl trifluoromethyl ether (cinnamyl
trifluoromethyl ether)
3. Experimental part
THF was distilled over sodium/benzophenone prior to use.
Dichloromethane was dried over molecular sieves. Other reagents
were used as received.
Colourless oil (47%). 1H NMR: 7.45–7.29 (massif, 5H), 6.72 (bd,
1H, 2JH-H = 15.8), 6.29 (dt, 1H, 2JH-H = 15.8, 3JH-H = 6.4), 4.65 (dd, 2H,
3JH-H = 6.4, 4JH-H = 1.3). 13C NMR: 135.8, 135.3, 128.8, 128.6, 126.9,
121.9 (q, 1JC-F = 255.2), 121.5, 68.1 (q, 3JC-F = 3.5). 19F NMR: À60.54
(s). Anal: Calcd for C10H9F3O: C (59.41), H (4.49); Found: C (59.53),
H (4.40).
1H, 13C (CPD and DEPT 135) and 19F NMR spectra were generally
recorded in CDCl3 at respectively 300, 75 and 282 MHz, unless
specified. Attribution of 1H and 13C NMR peaks have sometimes
required 2D NMR experiments (COSY, HSQC, HMBC, NOESY,
HOESY). Chemical shifts are given in ppm relative to TMS (1H and
13C) or CFCl3 (19F), used as internal references. Coupling constants
are given in Hertz. For 19F NMR titration, PhCF3 has been used as
internal standard. The following abbreviations are used: ‘‘s’’
singlet, ‘‘bs’’ broad singlet, ‘‘d’’ doublet, ‘‘bd’’ broad doublet, ‘‘t’’
triplet, ‘‘q’’ quadruplet, ‘‘m’’ multiplet, ‘‘Cq’’ quaternary carbone.
Flash chromatography was performed on silica gel Geduran
3.1.5. Benzyl (trifluoromethoxy)acetate
Colourless oil (45%). 1H NMR: 7.39–7.35 (massif, 5H), 5.25 (s,
2H), 4.52 (s, 2H). 13C NMR: 166.0, 134.8, 128.9, 128.8, 128.7, 121.6
1
3
(q, JC-F = 256.9), 67.7, 63.2 (q, JC-F = 3.5). 19F NMR: À61.86 (s).
Anal: Calcd for C10H9F3O3: C (51.29), H (3.87); Found: C (51.21), H
(4.18).
60 M (40–60
in capillary tubes on a Bu¨chi apparatus.
m
m). Melting points (uncorrected) were determined
3.1.6. 2,6-Dimethyl-8(trifluoromethoxy)oct-2-ene (citronellyl
trifluoromethyl ether)
4(S)-Benzyl-3-(bromoacetyl)-1,3-oxazolidin-2-one was pre-
pared according to [40]. Its iodo analog was obtained by exchange
with NaI in acetone. 2-Iodoacetophenone and 2-(iodoacetyl)furan
were prepared according to [41]. 2,6-Dimethyl-8-iodo-oct-2-ene
(citronellyl iodide), (2E)-1-iodo-3,7-dimethylocta-2,6-diene (ger-
anyl iodide) and (1S,2S,4R)-2-iodo-1-isopropyl-4-methylcyclohex-
ane (neomenthyl iodide) were prepared according to [42]. N-
benzyl triflimide was prepared according to [43].
Colourless oil (58%). 1H NMR: 5.09 (m, 1H), 4.00 (m, 2H), 1.98
(m, 2H), 1.63–1.12 (massif, 11H); 0.91 (d, 3H, 3JH-H = 6.4). 13C NMR:
131.7, 124.5, 121.9 (q, 1JC-F = 253.6), 66.0 (q, 3JC-F = 3.1), 37.0, 35.7,
29.1, 25.8, 25.5, 19.3, 17.7. 19F NMR: À61.13 (s). Anal: Calcd for
C11H19F3O: C (58.91), H (8.54); Found: C (59.12), H (8.45).
3.1.7. Trifluoromethyl 2,3,4,6-tetra-O-acetyl-
b-D-glucopyranoside
White solid (75%). M.P.: 120–122 8C. 1H NMR: 5.25–5.07
2
3
(massif, 4H), 4.31 (dd, 1H, JH-H = 12.5, JH-H = 4.8), 4.14 (dd, 1H,
3
3
3
3
3.1. Nucleophilic trifluoromethoxylation with TFMT and
AgF: general procedure
2JH-H = 12.5, JH-H = 2.3), 3.82 (ddd, 1H, JH-H = 9.8, JH-H = 4.8, JH-
H = 2.3), 2.10 (s, 3H), 2.07 (s, 3H), 2.04 (s, 3H), 2.02 (s, 3H). 13C NMR:
1
3
170.9, 170.1, 169.3, 169.0, 121.0 (q, JC-F = 259.6), 96.9 (q, JC-
F = 3.3), 72.8, 72.2, 70.2, 67.6, 61.4, 20.6, 20.5, 20.4. 19F NMR:
À59.57 (s). Anal: Calcd for C15H19F3O10: C (43.28), H (4.60); Found:
C (43.37), H (4.76).
In a 10 mL round bottomed flask, equipped with a rubber
septum and a magnetic stirrer, silver fluoride (1 mmol) was
introduced. Under nitrogen atmosphere, anhydrous CH3CN (2 mL)
were added and the heterogeneous mixture was cooled to À30 8C.
TFMT (300
m
L) was then added, the vessel was tightly closed
3.1.8. (4S)-4-Benzyl-3-[(trifluoromethoxy)acetyl]-1,3-oxazolidin-2-
one
Viscous yellow oil (80%). 1H NMR: 7.38-7.19 (massif, 5H), 5.13
(d, 1H, 2JH-H = 17.3), 5.07 (d, 1H, 2JH-H = 17.3), 4.71 (m, 1H), 4.29 (m,
2H), 3.32 (dd, 1H, 2JH-H = 13.5, 3JH-H = 3.2), 2.86 (dd, 1H, 2JH-H = 13.5,
3JH-H = 9.3). 13C NMR: 165.3, 153.5, 134.6, 129.4, 129.1, 127.7,
(autogenous pressure of COF2 is needed to allow the reaction to
proceed) and the reaction mixture was stirred for 2 h at À30 8C.
After addition of the electrophile (neat when liquid or dissolved in
the minimum of CH3CN when solid) by the mean of a gas-tight
syringe, stirring was continued at À30 8C for 30 min then at r.t. for
24 h (in the dark). Finally, the vessel was depressurised and the
reaction mixture was filtered over celite. The filtrate was
concentrated in vacuo, the residue was dissolved in dichloro-
methane, washed with brine, dried over MgSO4, filtered and
concentrated in vacuo. Purification by chromatography over silica
gel finally afforded the pure corresponding trifluoromethyl ether.
1
3
121.6 (q, JC-F = 256.7), 67.8, 65.7 (q, JC-F = 3.3), 54.9, 37.5. 19F
NMR: À61.38 (s). Anal: Calcd for C13H12F3NO4: C (51.49), H (3.99),
N (4.92); Found: C (51.73), H (4.22), N (5.21).
3.1.9. 1-Phenyl-2-(trifluoromethoxy)ethanone [
(trifluoromethoxy)acetophenone]
a-
Yellow oil (69%). 1H NMR: 7.91 (m, 2H), 7.65 (m, 1H), 7.52 (m,
2H), 5.18 (s, 2H). 13C NMR: 190.2, 134.4, 133.8, 129.1, 127.9, 121.8
3.1.1. Benzyl trifluoromethyl ether
Colourless oil (76%). 1H NMR: 7.40–7.35 (massif, 5H), 4.99 (s,
(q, JC-F = 256.3), 68.4 (q, JC-F = 2.9). 19F NMR: -61.44 (s). Anal:
1
3
1
2H). 13C NMR: 134.0, 129.1, 128.7, 127.5, 121.8 (q, JC-F = 255.4),
Calcd for C9H7F3O2: C (52.95), H (3.46); Found: C (53.12), H (3.15).
69.2 (q, 3JC-F = 3.5). 19F NMR: -60.78 (s). Anal: Calcd for C8H7F3O: C
(54.55), H (4.01); Found: C (54.34), H (3.90).
3.1.10. 1-(2-Furyl)-2-(trifluoromethoxy)ethanone
3
4
Colourless oil (40%). 1H NMR: 7.63 (dd, 1H, JH-H = 1.8, JH-
3
4
3
3.1.2. 1-Phenylethyl trifluoromethyl ether
H = 0.6), 7.34 (dd, 1H, JH-H = 3.6, JH-H = 0.6), 6.60 (dd, 1H, JH-
Colourless oil (50%). 1H NMR: 7.41–7.30 (massif, 5H), 5.30 (q,
H = 3.6, JH-H = 1.8), 5.01 (s, 2H). 13C NMR: 179.6, 150.2, 147.4,
3
3
3
1
3
1H, JH-H = 6.6), 1.63 (d, 3H, JH-H = 6.6). 13C NMR: 140.6, 128.7,
121.7 (q, JC-F = 254.2), 119.0, 112.0, 67.8 (q, JC-F = 3.0). 19F NMR:
À61.62 (s). Anal: Calcd for C7H5F3O3: C (43.41), H (2.60); Found: C
(43.32), H (2.89).
128.5, 125.9, 121.8 (q, JC-F = 255.0), 77.3 (q, JC-F = 2.6), 23.4. 19F
NMR: À58.44 (s). Anal: Calcd for C9H9F3O: C (56.85), H (4.77);
Found: C (57.01), H (4.89).
1
3
Acknowledgements
3.1.3. Trifluoromethyl benzoate
Colourless oil (50%). 1H NMR: 8.07 (m, 2H), 7.70 (m, 1H), 7.52
(m, 2H). 13C NMR: 159.2, 135.3, 130.7, 129.1, 126.8 (q, 4JC-F = 1.6),
One of the authors (O.M.) gratefully acknowledge the Bayer
CropScience Co. for granting his PhD work.

O. Marrec et al. / Journal of Fluorine Chemistry 131 (2010) 200–207
207
References
(d) L.M. Yagupolskii, E.B. Dyachenko, V.I. Troitskaya, Ukrain. Khim. Zh. 27 (1961)
77-79; Chem. Abstr. 55 (1961) 21029a..
[25] N.N. Yarovenko, A.S. Vasileva, J. Gen. Chem. USSR 29 (1959) 3747–3748.
[26] B. Langlois, M. Desbois, Ann. Chim. Fr. 9 (1984) 729–741. This toxicity has been
mentioned in reference [7] but without citing any source.
[27] (a) W.A. Sheppard, J. Org. Chem. 29 (1964) 1–11;
(b) P.E. Aldrich, W.A. Sheppard, J. Org. Chem. 29 (1964) 11–15;
(c) W. Dmowski, R.A. Kolinski, Polish J. Chem. 52 (1978) 547–559.
[28] (a) M. Kuroboshi, K. Suzuki, T. Hiyama, Tetrahedron Lett. 33 (1992) 4173–4176;
(b) K. Kanie, Y. Tanaka, M. Shimizu, M. Kuroboshi, T. Hiyama, Chem. Commun.
(1997) 309–310;
(c) K. Kanie, Y. Tanaka, K. Suzuki, M. Kuroboshi, T. Hiyama, Bull. Chem. Soc. Jpn. 73
(2000) 471–484;
(d) M. Kuroboshi, K. Kanie, T. Hiyama, Adv. Synth. Catal. 343 (2001) 235–250;
(e) M. Shimizu, T. Hiyama, Angew. Chem. Int. Ed. 44 (2005) 214–231.
[29] T. Umemoto, Chem. Rev. 96 (1996) 1757–1777;
[1] T. Hiyama, Organofluorine Compounds, Chemistry and Applications, Springer,
Berlin, 2000.
[2] J.P. Be´gue´, D. Bonnet-Delpon, Chimie bioorganique et medicinal du fluor, CNRS
Editions & EDP Sciences, Paris, 2005.
[3] P. Jeschke, ChemBioChem 5 (2004) 570–589.
[4] P. Jeschke, E. Baston, F.R. Leroux, Mini-Rev. Med. Chem. 7 (2007) 1027–1034.
[5] F. Leroux, P. Jeschke, M. Schlosser, Chem. Rev. 105 (2005) 827–856.
[6] (a) L.M. Yagupolskii, V.I. Troitskaya, J. Gen. Chem. USSR 27 (1957) 587–594;
(b) L.M. Yagupolskii, M.S. Marenets, J. Gen. Chem. USSR 27 (1957) 1477–
1480.
[7] F.R. Leroux, B. Manteau, J.-P. Vors, S. Pazenok, Beilstein J. Org. Chem. 4 (13) (2008)
1–15, doi:10.3762/bjoc.4.13.
[8] C.D.S. Tomlin, Pesticide Manual, 13th ed. British Crop Protection Council, Farham,
2003.
T. Umemoto, K. Adachi, S. Ishihara, J. Org. Chem. 72 (2007) 6905–6917.
[30] (a) P. Eisenberger, S. Gischig, A. Togni, Chem. Eur. J 12 (2006) 2579–2586;
(b) I. Kieltsch, P. Eisenberger, A. Togni, Angew. Chem. Int. Ed. 46 (2007) 754–757;
(c) P. Eisenberger, I. Kieltsch, N. Armanino, A. Togni, Chem. Commun. (2008)
1575–1577.
[31] K. Stanek, R. Koller, A. Togni, J. Org. Chem. 73 (2008) 7678–7685.
[32] R. Koller, K. Stanek, D. Stolz, R. Aardoom, K. Niedermann, A. Togni, Angew. Chem.
Int. Ed. 48 (2009) 4332–4336.
[9] G.A. Olah, T. Yamato, T. Hashimoto, G. Shih, N. Trivedi, B.P. Singh, M. Piteau, J.A.
Olah, J. Am. Chem. Soc. 109 (1987) 3708–3713.
[10] M.A. McClinton, D.A. McClinton, Tetrahedron 48 (1992) 6555–6666.
[11] (a) W.A. Sheppard, J. Am. Chem. Soc. 83 (1961) 4860–4861;
(b) I.W. Serfaty, T. Hodgins, E.T. McBee, J. Org. Chem. 37 (1972) 2651–2655;
(c) R.W. Taft, E. Price, J.R. fox, I.C. Lewis, K.K. Andersen, G.T. Davis, J. Am. Chem.
Soc. 85 (1963), 709 and 3146;
(d) L.M. Yagupolskii, V.F. Bystrov, A.U. Stepanyants, Yu.A. Fialkov, J. Gen. Chem.
USSR 34 (1964) 3731–3738.
[33] S. Rozen, Chem. Rev. 96 (1996) 1717–1736.
[34] M.E. Redwood, C. Willis, Can. J. Chem. 43 (1965) 1893–1898.
[35] M. Nishida, A. Vij, R.L. Kirchmeier, J.M. Shreeve, Inorg. Chem. 34 (1995) 6085–
6092.
[36] W.B. Farnham, B.E. Smart, W.J. Middleton, J.C. Calabrese, D.A. Dixon, J. Am. Chem.
Soc. 107 (1985) 4565–4567.
[37] (a) G.L. Trainor, J. Carbohydr. Chem. 4 (1985) 545–563;
(b) W.B. Farnham, W.J. Middleton (DuPont de Nemours Co.), Eur. Pat. Appl. EP
164124 A2 (1985).
[38] A.A. Kolomeitsev, M. Vorobyev, H. Gillandt, Tetrahedron Lett. 49 (2008) 449–453.
[39] (a) M. Oudrhiri Hassani, A. Germain, D. Brunel, A. Commeyras, Tetrahedron Lett.
22 (1981) 65–68;
[12] W.A. Sheppard, J. Am. Chem. Soc. 85 (1963) 1314–1318.
[13] A. Haas, Adv. Inorg. Chem. Radiochem. 28 (1984) 167–202.
[14] (a) C. Hansch, A. Leo, Substituent Constants for Correlation Analysis in Chemistry
and Biology, John Wiley & Sons, New, New York, 1979;
(b) A. Leo, P.Y.C. Jow, C. Silipo, C. Hansch, J. Med. Chem. 18 (1975) 865–868.
[15] W.A. Sheppard, J. Org. Chem. 29 (1964) 1–11.
[16] A.E. Lutskii, L.M. Yagupolskii, S.A. Volchenok, J. Gen. Chem. USSR 34 (1964) 2749–
2757.
[17] A.G. Anderson, PhD thesis (University of Utah), 1977; University Microfilms
International 77-20.316, Ann Arbor, Michigan, USA.
[18] A.D. Baker, D.P. May, D.W. Turner, J. Chem. Soc. (B) (1968) 22–34.
[19] A.E. Lutskii, L.M. Yagupolskii, E.M. Obukhova, J. Gen. Chem. USSR 34 (1964) 2663–
2667.
[20] W.A. Sheppard, J. Am. Chem. Soc. 87 (1965) 2410–2420.
[21] F.E. Herkes, J. Fluorine Chem. 9 (1977) 113–126.
[22] S.V. Sereda, M.Yu. Antipin, T.V. Timofeeva, Yu.T. Struchkov, S.V. Cheliajenko,
Kristallografiia 32 (1987) 1165–1174.
[23] F. Rose-Munch, R. Khourzoum, J.P. Djukic, E. Rose, B. Langlois, J. Vaisserman, J.
Organometal. Chem. 470 (1994) 131 (and references therein).
[24] (a) L.M. Yagupolskii, Dokl. Akad. Nauk SSSR 105 (1955) 100–102; Chem. Abstr. 50
(1955) 11270b.
(b) G.A. Olah, T. Ohyama, Synthesis (1976) 319–320;
(c) S.L. Taylor, J.C. Martin, J. Org. Chem. 52 (1987) 4147–4156, and references
therein.
[40] J.R. Falck, A. He, H. Fukui, H. Tsutsui, A. Radha, Angew. Chem. Int. Ed. 46 (2007)
4527–4529.
[41] G. Yin, M. Gao, N. She, S. Hu, A. Wu, Y. Pan, Synthesis (2007) 3113–3116.
[42] (a) G.L. Lange, C. Gottardo, Synth. Commun. 20 (1990) 1473–1479;
(b) G. Anilkumar, H. Nambu, Y. Kita, Org. Process Res. Dev. 6 (2002) 190–191.
[43] (a) J.B. Hendrickson, R. Bergeron, A. Giga, D. Sternbach, J. Am. Chem. Soc. 95
(1973) 3412–3413;
(b) R.S. Glass, Chem. Commun. (1971) 1546–1547;
(c) P. Mu¨ ller, N.T.M. Phuong, Helvetica Chimica Acta 62 (1979) 1485–1492.
(b) L.M. Yagupolskii, V.I. Troitskaya, J. Gen. Chem. USSR 31 (1961) 845–852;
(c) L.M. Yagupolskii, V.V. Orda, J. Gen. Chem. USSR 34 (1964) 1994–1998;