N-Fluoro-bis[(trifluoromethyl)sulfonyl]imide: Interesting reactions involving nucleophilic attack on sulfonyl groups

Article abstract of DOI:10.1016/S0022-1139(99)00231-6

N-Fluoro-bis[(trifluoromethyl)sulfonyl]imide (1) undergoes reaction with a variety of nucleophiles to give a number of interesting products. Under suitable conditions, the sulfonyl group on 1 is subject to nucleophilic attack forming N-fluoro-trifluoromethylsulfonyl amide (3) and Nu-SO₂CF₃. Surprisingly, 3 can also function as a weak nucleophile in the absence of other stronger bases.

Full text of DOI:10.1016/S0022-1139(99)00231-6

Journal of Fluorine Chemistry 102 (2000) 135±139
N-Fluoro-bis[(tri¯uoromethyl)sulfonyl]imide: interesting reactions
involving nucleophilic attack on sulfonyl groups
*
Weiwen Ying, Darryl D. DesMarteau , Ze-Qi Xu, Michael Witz
Howard L. Hunter Chemistry Laboratory, Clemson University, Clemson, SC 29634-1905, USA
Received 22 June 1999; received in revised form 12 July 1999; accepted 2 August 1999
Dedicated to Professor Paul Tarrant on the occasion of his 85th birthday
Abstract
N-Fluoro-bis[(tri¯uoromethyl)sulfonyl]imide (1) undergoes reaction with a variety of nucleophiles to give a number of interesting
products. Under suitable conditions, the sulfonyl group on 1 is subject to nucleophilic attack forming N-¯uoro-tri¯uoromethylsulfonyl
2 3
amide (3) and Nu-SO CF . Surprisingly, 3 can also function as a weak nucleophile in the absence of other stronger bases. # 2000 Elsevier
Science S.A. All rights reserved.
Keywords: Selective ¯uorination; N-¯uorosulfonylimides and amides; Nucleophilic reactions; Metal ¯uorides
1
. Introduction
tions by 1 in polar solvents, the presence of small amounts of
water in the solvent indicated 1 was reactive with water.
In the reaction of 1 with water in acetonitrile, water is an
apparently good nucleophile to attack the sulfonyl group to
give both tri¯ic acid (2) and N-¯uoro tri¯uoromethylsulfo-
nyl amide (3) [7,8]. Neat 1 is insoluble and unreactive in
water. However in a solvent in which both 1 and water are
soluble, reaction of 1 with water is facilitated.
The novel N-¯uoro-bis[(per¯uoroalkyl)sulfonyl]imides
were ®rst synthesized in 1987 and found to have remarkable
¯
¯
¯
published [2]. Interest in bis[(per¯uoroalkyl)sulfonyl)]i-
mides in general has grown as the useful properties of the
class of compounds have emerged. From electrochemical
applications [3] to catalysis (for a selection of examples, see
uorinating capability [1]. A number of studies using N-
uoro-bis[(tri¯uoromethyl)sulfonyl]imide (1) for selective
uorination of a variety of organic compounds have been
CH CN
3
ꢀ
CF SO † NF ! CF SO H ‡ CF SO NHF
3
2 2
3
3
2 3
3
2
H2O
1
1
‡ CH3CH2OH ! 3 ‡ CH3CH2OSO2CF3
As expected, alcohols undergo the same reaction with 1 to
[
4±6]), their remarkable properties are evident. While the
parent acids, metal salts and many derivatives are robust
compounds, in the course of our extensive research on these
imides we have observed some unusual reactions of 1. Here
we describe reactions of 1 with some nucleophiles.
give alkyl tri¯ates and 3 [7]. However, bulky alkyl tri¯ates
may not be stable and can be transformed to a variety of
products. For example, when the reaction between tert-butyl
alcohol and 1 was carried out in anhydrous CH Cl or
2
2
CDCl3, 3,3-di¯uoro-2-methylpropene (5) and N-tert-
butyl-N-¯uoro-tri¯uoromethylsulfonyl amide (6) were
obtained as well as a mixture of mono- and di¯uoro com-
pounds with low boiling points.
2
. Results and discussion
The earliest recognition that 1 reacts with nucleophiles
came from the puri®cation of 1 from HF over metal ¯uor-
ides. At times on long contact with NaF, CF SO F was
3
2
observed. Later we found that the formation of CF SO F
3
2
was much more rapid with KF or CsF. Subsequently we
found that alcohols destroyed neat 1 and could lead to the
explosive decomposition of 1. Then in selective ¯uorina-
*
Corresponding author. Tel.: ‡1-864-656-4705; fax: ‡1-864-656-0627.
E-mail address: fluorin@clemson.edu (D.D. DesMarteau).
0022-1139/00/$ ± see front matter # 2000 Elsevier Science S.A. All rights reserved.
PII: S 0 0 2 2 - 1 1 3 9 ( 9 9 ) 0 0 2 3 1 - 6

1
36
W. Ying et al. / Journal of Fluorine Chemistry 102 (2000) 135±139
The reaction was carried out at � 508C and then warmed
to 238C. After the reaction was completed, 6 was isolated by
trap to trap fractionation under high vacuum. The solvent
19
fraction contained other low-boiling-point compounds.
F
NMR of this fraction showed that the mixture contained
both mono- and di¯uoro products. 3,3-Di¯uoro-2-methyl-
propene (5) was separated from other products after several
trap to trap fractionations and characterized as a mixture
with CDCl by ¯uorine NMR, proton NMR and IR. A small
3
amount of non-volatile residue left in the reaction container
is believed to be polymeric material resulting from acid
catalyzed polymerization of alkenes. A proposed mechan-
ism to account for the observed products is shown in
Scheme 1.
The unexpected formation of N-¯uoro amide 6 is appar-
ently a result of the N-¯uoro amide 3 acting as nucleophile
toward the stabilized tert-butyl cation. N-¯uoro amide 3
should be a very weak nucleophile due to the presence of
both ¯uorine and tri¯uoromethylsulfonyl group. However
because there is no other nucleophile stronger than 3 present
in the reaction mixture, the tert-butyl cation captures 3 to
furnish compound 6 in addition to other products. 6 is a
colorless liquid which slowly decomposes at 228C.
As proposed in Scheme 1, the main products are formed
through the intermediate 2-methylpropene followed by
¯
uorination. Due to the lack of good nucleophile in the
reaction mixture, 2-methyl propene is produced in favor
of products from nucleophilic substitution. Interestingly
more 3,3-di¯uoromethylpropene 5 was obtained than
other mono¯uoro products. This implies that geminal
hydrogen on ¯uoromethyl group of 2-(2-¯uoromethyl)pro-
pyl cation 7 is more acidic than the hydrogen of the tert-
butyl cation, and 1-¯uoro-2-methylpropene is formed more
rapidly than 2-methylpropene. Previous studies have shown
that the reaction between alkenes and 1 is fast [8], so the
distribution of products is determined by the rate of forma-
tion of the alkenes by elimination and then subsequent
reaction with 1.
Scheme 1.
cation is a well-studied intermediate in organomercuric
chemistry, [10] and we suggest that the formation of mer-
curic cation is initiated by the presence of a trace amount of
acid in 1. Since one equivalent of methanol results, nucleo-
philic attack of methanol on 1 is also observed. As suggested
in the reaction between tert-butyl alcohol and 1, N-¯uoro
amide 3 undergoes electrophilic substitution on the mercu-
ric intermediate to give the ®nal product with full retention
of con®guration. The proposed mechanism is shown in
Scheme 2.
Mono¯uoro ethers were successfully synthesized using
nonaromatic mercury derivatives and acetyl hypo¯uorite
0
[
curic chloride (8) using 1, only a small amount of ¯uor-
9]. In an attempt to ¯uorinate 2 -methoxycyclohexyl mer-
1
ination product was obtained. Instead, a similar electrophilic
0
In order to develop a more effective method to prepare
and isolate the N-¯uoro amide 3, reactions of 1 with
tetrabutylammonium ¯uoride and cesium ¯uoride were
carried out. Both ¯uorides react with 1, but tetrabutylam-
monium ¯uoride led to a cleaner reaction.
substitution product N-¯uoro-2 -chloromercuriocyclohexyl
tri¯uoromethylsulfonyl amide (9) by N-¯uoro amide 3 was
obtained as the major product.
‡
!
�
Bu4N F
� ‡
CF SO NF Bu N
3 2 4
1
CF SO F
2
‡
3
1
1
H2SO4
1
1 ! CF SO NHF
3
2
A mechanism involving the three-membered mercuric
intermediate is suggested. A three-membered mercuric
Fluoride ion attacks 1, leading to the formation of tri-
uoromethylsulfonyl ¯uoride and tetrabutylammonium N-
uoro tri¯uoromethylsulfonyl amide (11). The products can
¯
¯
1
The results of reactions of 1 with other mercury derivatives will be
described in a forthcoming publication.
be easily separated because tri¯uoromethylsulfonyl ¯uoride

W. Ying et al. / Journal of Fluorine Chemistry 102 (2000) 135±139
137
Scheme 2.
is gas at room temperature. Acidi®cation of the amide salt
gave the desired N-¯uoro amide 3 which was puri®ed by
trap to trap fractionation. This route proved to be an
excellent method to prepare the N-¯uoro amide 3. Unfortu-
nately all attempts to obtain suitable crystals of 11 for X-ray
structural analysis have so far been unsuccessful.
Scheme 3.
Previous studies have shown that a variety of enols and
enolates can be successfully ¯uorinated using 1 and other
electrophilic ¯uorinating reagents [11,12]. However in sev-
eral cases, we found that nucleophilic substitution on 1 was
favored over ¯uorination. For instance, the methylene group
in 3-methyl-1-phenyl-2-pyrazoline (12) is very reactive
considering the facile formation of the enol, but in this case
the major product from the reaction between 1 and pyrazo-
line 12 is 3-methyl-1-phenyl-5-tri¯uoromethylsulfonylox-
ypyrazole (13). The enol form of 12 can be seen as an
aromatic system, and the hydroxyl group attacks 1 to give
the corresponding tri¯ate. The suggested mechanism for
formation of products is shown in Scheme 3.
Scheme 4.
As for the product 4,4-di¯uoro-3-methyl-1-phenyl pyr-
azoline (14), an alternative route to the same product was
found.
enolate 16 was treated with 1, only the enol tri¯ate 17 was
formed in 38% yield. The presumed coproduct FN(Na)-
SO CF was not isolated.
2
3
In conclusion, while 1 provides a convenient and effective
source of electrophilic ¯uorine which is of considerable
utility in the selective ¯uorination of organic molecules, 1
also has excellent utility in other ways. Useful sulfonyl
derivatives have been illustrated in this work and the results
suggest that a chiral ¯uorinating agent might be prepared
through the reaction of 1 with a chiral tertiary alcohol.
Although only a carbocation mechanism is discussed in this
paper, a possible single electron transfer process leading to
the formation of carbocation in some reactions cannot be
ruled out [8]. Ongoing research on further application of 1
The reaction of 1 with carbon±carbon double bonds has
been investigated in some detail [8]. Considering that anti-
pyrene contains a double bond at the 3,4-position, 1 can
easily react with antipyrene. However since there is no good
nucleophile present to trap the intermediate cation, the 2-
methyl group is lost in the second ¯uorination forming
pyrazoline 14. Scheme 4 shows the proposed mechanism.
Another related example of nucleophilic substitution on 1
instead of ¯uorination is found in attempts to ¯uorinate the
sodium enolate of ethyl acetoacetate (16). When sodium

1
38
W. Ying et al. / Journal of Fluorine Chemistry 102 (2000) 135±139
1
9
will provide more information for the mechanistic study.
Finally, this work provides a useful guide to some of the
limitations for the use of 1 in selective ¯uorination.
3,3-Di¯uoro-2-methylpropene (5) [16]: F NMR d
1
� 117.0 (d, J ˆ 55.9 Hz); H NMR d 1.8 (s, 3H), 5.27
�
1
(m, 1H), 5.35 (m, 1H), 5.96 (t, J ˆ 56.0 Hz); IR (cm
)
(
gas in mixture with CDCl ) 1476, 1398, 1029.
3
0
Reaction of 1 with 2 -methoxycyclohexyl mercuric chlor-
ide (8): A solution of 8 (0.17 g, 0.5 mmol) in CH CN was
3
. Experimental section
3
added to the solution of 1 (0.18 g, 0.6 mmol) at � 408C. The
mixture was warmed to 238C and stirred for 1 h. All
volatiles were collected through a glass vacuum line. A
small amount of the ¯uorination product 1-methoxy-2-
¯uorocyclohexane (10) was found in the mixture with
CH CN. CH Cl (20 ml) was added to dissolve the residue
All reactions were performed in glass apparatus. All
Ê
solvents were dried over 3 A molecular sieves or distilled
from P O before use. Commercially available starting
4
10
materials 4, 10, 12, 15 (from Aldrich Chemical) were used
as received. 8 was prepared as reported [13]. 1 was prepared
by the literature method [14]. Caution! 1 is a strong oxidizer
and must be handled with care. Neat 1 should not be allowed
to contact strong bases or easily oxidized substrates. Also 3
may be explosive under certain conditions, although we
encountered no problems in handling small amounts of 3.
Flash column chromatography on silica gel was performed
3
2
2
and the insoluble compounds were removed through a short
silica gel column. 9 was isolated by column chromatogra-
phy using CH Cl as the eluant. The yield was 51% (0.12 g).
2
2
19
1-Methoxy-2-¯uorocyclohexane (10) [17]: F NMR d
� 180.5 (ddt, J ˆ 50.8, 35.0, 17.7 Hz).
19
0
N-Fluoro-2 -chloromercuriocyclohexyl tri¯uoromethyl-
1
9
1
13
as described in the original paper [15]. F, H and C NMR
were recorded in ppm positive down®eld from internal
sulfonyl amide (9): F NMR d � 69.27 (d-br., J ˆ 36.7 Hz,
1
1F), � 75.10 (d, J ˆ 5.7 Hz, 3F); H NMR d 1.1±2.5 (m,
1
9
1
13
CFCl for F NMR and from TMS for H and C NMR
3
8H), 2.75 (dt, J ˆ 3.7, 13.1 Hz, 1H), 4.22 (dtd, J ˆ 37.6,
13
at 188.31, 200.13, and 75.47 MHz, respectively. Chemical
shifts were reported for CDCl solution unless otherwise
11.3, 3.6 Hz, 1H); C NMR d 24.6, 27.0 (t, J ˆ 117.8 Hz),
31.1 (dt, J ˆ 82.4, 7.6 Hz), 32.9 (t, J ˆ 39.0 Hz), 54.0 (t,
J ˆ 840.6 Hz), 68.4 (dt, J ˆ 14.2, 50.1 Hz), 118.4 (q,
3
noted. Mass spectra were measured at 70 eV for electron
impact (EI) and with methane for chemical ionization (CI).
Reaction of 1 with tert-butyl alcohol (4): A solution of N-
‡
J ˆ 321.0 Hz); MS (EI) m/e 437 (M � Cl, 0.22), 272
‡
‡
� 1
(M � Hg, 54), 69 (CF , 100); IR (cm ) (nujol oil)
3
¯
uoro-bis[(tri¯uoromethyl)sulfonyl]
imide
(0.3 g,
.0 mmol) in anhydrous CH Cl (2 ml) was added to the
1481, 1376, 1206, 1128.
1
solution of tert-butyl alcohol (0.1 g, 1.4 mmol) in CH Cl2
Preparation of N-¯uorotri¯uoromethylsulfonylamide (3).
Tetrabutylammonium N-Fluoro Tri¯uoromethylsulfonyl
amide (11): A solution of 1 (0.45 g, 1.5 mmol) in CH Cl
2
2
2
(
2 ml) at � 508C in 10 min. The reaction mixture was stirred
2
2
at � 508C for 1 h and then warmed up to 238C. The colorless
(5 ml) was added dropwise into a 1.5 M solution of tetra-
butylammonium ¯uoride (1.5 ml) in THF at � 408C under
N2. After 30 min, the mixture was warmed to 08C, and then
quenched with water. The CH Cl solution was separated
19
solution turned brown after being stirred overnight.
NMR of the mixture showed the products were composed
F
of N-¯uoro-N-tert-butyl tri¯uoromethylsulfonyl amide
(
2
2
40%), 3,3-di¯uoro-2-methylpropene (25%) and other gas-
and washed with water again. Evaporation of solvent gave
0.72 g white solid (m.p. 96±988C) which contained a small
amount of unidenti®ed impurities. Attempts at further pur-
eous and polymeric products. Vacuum fractional condensa-
tion was carried out and pure N-¯uoro-N-tert-butyl
tri¯uoromethylsulfonyl amide (6) was isolated in a
1
9
i®cation failed.
F NMR (CD CN) d � 76.50 (d,
3
1
�
308C trap. Some 6 was also collected in a 08C trap and
J ˆ 5.7 Hz, 3F), � 131.07 (q, J ˆ 5.7 Hz, 1F); H NMR
some was carried over with the large amount of CH Cl
(CD CN) d 0.99 (t, J ˆ 7.0 Hz, 12H), 1.38 (q, J ˆ 7.0 Hz,
2
2
3
‡
solvent to the next trap (� 608C). Other low-boiling-point
compounds including 3,3-di¯uoro-2-methylpropene (5)
were collected in � 608C trap. 6 has a vapor pressure of
ca. 4 torr at 238C. After standing for 2 h at 238C, the original
colorless liquid 6 developed a purple color.
8H), 1.58 (m, 8H), 3.20 (m, 8H); MS (EI) m/e: 242 (Bu N ,
4
‡
11), 142 [(Bu N=CH ) , 100].
2
2
N-Fluorotri¯uoromethylsulfonylamide (3): A mixture of
11 (0.60 g, 1.47 mmol) and H SO (98%) (4 ml) was stirred
2
4
for 10 min at 238C, and the product was isolated by trap to
trap fractionation under high vacuum. 3 was collected at
� 808C trap in 41% yield (100 mg). 3 has a vapor pressure of
N-Fluoro-N-tert-butyl tri¯uoromethylsulfonyl amide (6):
1
1
9
F NMR d � 58.9 (br., 1F), � 76.5 (d, J ˆ 8.47 Hz, 3F); H
1
3
19
NMR d 1.53 (s); C NMR d 27.25 (d, J ˆ 5.6 Hz), 69.34 (d,
about 8 torr at 238C. F NMR d � 73.73 (d, J ˆ 5.8 Hz, 3F),
1
J ˆ 12.3 Hz), 118.47 (q, J ˆ 321.8 Hz); MS (EI) m/e 224
� 91.32 (d-br., J ˆ 21.0 Hz, 1F); H NMR d 9.28±9.88 (br.);
‡
‡
� 1
‡
‡
� 1
(
[
M ‡ 1, 0.08), 208 (M � CH3, 0.75), 69 (CF , 53), 57
MS (CI) m/e 168 (M ‡ 1, 100); IR (cm ) (gas) 3656,
3
(CH ) C , 100]; IR (cm ) (gas) 2993.5, 1420.3, 1379.1,
‡
3
3
3570 (N±H), 1447, 1232, 1137.
1
226.0, 1137.5, 1048.0.
A similar reaction was carried out using anhydrous
Reaction of 1 with antipyrene (15): N-Fluoro bis[(tri¯uor-
omethyl)sulfonyl] imide (0.34 g, 1.13 mmol) was added to a
solution of antipyrine (0.19 g, 1.0 mmol) in chloroform
CDCl as solvent. After removal of 6, careful trap to trap
3
fractionation of the remaining volatiles yielded a solution of
5
1
9
(5 ml). The mixture was stirred at 238C for 2 days.
NMR showed only the di¯uorination products. Another
F
in CDCl3.

W. Ying et al. / Journal of Fluorine Chemistry 102 (2000) 135±139
139
1
9
1
portion of N-¯uoro bis[(tri¯uoromethyl)sulfonyl] imide
340 mg, 1.13 mmol) was added. The stirring was continued
(0.15 g). F NMR d � 75.17 (s); H NMR d 1.30 (t,
(
J ˆ 7.1 Hz, 3H), 2.17 (s, 3H), 4.25 (q, J ˆ 7.1 Hz, 2H),
‡
for another 3 days. The products were extracted by chloro-
form (3 Â 50 ml). The combined extract was dried over
anhydrous sodium sulfate. The main product 4,4-di¯uoro-3-
methyl-1-phenyl-2-pyrazolin-5-one (14) [18] was isolated
by column chromatography using hexane/ethyl acetate (4:1)
5.76 (p, J ˆ 0.9 Hz, 1H); MS (EI) m/e 263 (M ‡ 1, 5), 262
‡
‡
� 1
(M , 4), 69 (CF3 , 100); IR (cm ) (neat) 1725, 1686, 1424.
Acknowledgements
1
9
as eluant. The yield was 38% (80 mg). F NMR d � 121.6
1
(
7
1
q, J ˆ 1.6 Hz); H NMR d 2.28 (t, J ˆ 1.6 Hz, 3H), 7.23±
The authors are grateful to the National Science Founda-
tion for support of this research.
1
3
.84 (m, 5H); C NMR d 11.5, 107.8 (t, J ˆ 255.1 Hz),
18.3, 126.0, 128.9, 136.4, 151.6 (t, J ˆ 22.7 Hz), 158.9 (t,
J ˆ 29.5 Hz).
Reaction of 1 with 3-methyl-1-phenyl-2-pyrazoline (12):
The reaction was run in a similar fashion as above. After 2 h,
the products 13 (yield 90%) and 14 (6%) were isolated by
preparative TLC using ethyl acetate-hexane as developing
solvent.
References
[1] S. Singh, D.D. DesMarteau, S.S. Zuberi, M. Witz, H.-N. Huang, J.
Am. Chem. Soc. 104 (1987) 443.
[
[
[
[
2] W. Ying, D.D. DesMarteau, Y. Gotoh, Tetrahedron 52 (1996) 15.
3] D.D. DesMarteau, J. Fluorine Chem. 72 (1995) 203.
4] K. Ishihara, M. Kubota, H. Yamamoto, Synlett (1996) 1045.
5] K. Ishihara, Y. Karumi, M. Kubota, H. Yamamoto, Synlett (1996)
839.
3
-Methyl-1-phenyl-5-tri¯uoromethylsulfonyloxypyrazole
1
9
1
(
13): F NMR d � 73.1 (s); H NMR d 2.28 (s, 3H), 6.08 (s,
13
1
H), 7.36±7.45 (m, 5H); C NMR d 14.5, 96.4, 118.4 (q,
[
6] H. Kobayashi, J. Nie, T. SoInada, Chem. Lett. (1995) 307.
J ˆ 322 Hz), 123.6, 128.3, 129.4, 136.7, 140.9, 149.8; MS
19
‡
‡
3
[7] Products were detected by F NMR as a mixture in CD CN.
(
(
EI) m/e 306 (M , 10), 173 (M � CF SO , 95), 77
� 1
6 5
3
2
[
8] D.D. DesMarteau, Z.-Q. Xu, M. Witz, M. J. Org. Chem. 57 (1992)
29.
9] D. Hebel, S. Rozen, J. Org. Chem. 52 (1987) 2588.
‡
C H , 100); IR (cm ) (neat) 1599, 1558, 1504, 1435,
6
1
385, 1368, 1244, 1219, 1132.
Reaction of 1 with ethyl acetoacetate, sodium salt (16). A
[
[10] C.A. McAuliffe, The Chemistry of Mercury, MacMillan, London,
977.
1
solution of 16 in anhydrous THF [prepared by adding NaH
powder (0.08 g, 3.3 mmol) to ethyl acetoacetate (0.20 g,
[
[
[
[
11] G.S. Lal, G.P. Pez, R.G. Syvret, Chem. Rev. 96 (1996).
12] F.A. Davis, P.V.N. Kasu, Org. Prep. Proceed. Int. 312 (1999) 125.
13] J. Chatt, Chem. Rev. 48 (1951) 7.
1
.5 mmol) in THF (5 ml) followed by ®ltration of excessive
NaH] was added to a stirred solution of 1 in THF (2 ml)
under N2 at � 508C. After 1 h, usual work-up was performed
and the crude product was puri®ed by column chromato-
graphy using CH Cl as eluant. Ethyl 3-tri¯uoromethane-
14] D.D. DesMarteau, M. Witz, J. Fluorine Chem. 52 (1991) 7.
[15] W.C. Still, H. Kanh, A. Mitra, J. Org. Chem. 43 (1978) 2934.
[
[
[
16] F.J. Weigert, J. Org. Chem. 45 (1980) 3476.
17] A. Bakloutt, M.M. Chaabouni, J. Fluorine Chem. 18 (1981) 45.
18] F.A. Davis, W. Han, C.K. Murphy, J. Org. Chem. 60 (1995) 4730.
2
2
sulfonyloxy-2-crotonate (17) was obtained in 38% yield