Full Paper
(hydrodefluorination), in which CꢀC bonds are not formed.[10]
This might be partially due to the fact that CꢀF bonds are ex-
ceptionally strong (DH298 CH3ꢀF=481 kJmolꢀ1, DH298 PhꢀF=
Table 1. Screening of reaction conditions for the cyclization of 2-allyl-3-
(trifluoromethyl)phenols.
532 kJmolꢀ1 [11]
and unreactive. In general, highly reactive cata-
)
lysts that are usually not compatible with other functional
groups are required for the activation of CꢀF bonds.[12] Howev-
er, allylic trifluoromethyl[13] and difluoromethyl groups[14] dis-
play a lower activation barrier and can undergo substitution
with carbon and nitrogen nucleophiles through a SN2’ mecha-
nism. For instance, difluoroallenes[15] and b,b-difluorostyrenes[16]
were reported to be versatile substrates for the preparation of
fluorine-substituted PAHs and 3-fluorinated isoquinolines.
The application of the CF3 group in medicinal chemistry was
first reported in seminal work by Kiselyov and Strekowski
et al.[17] This enabled the synthesis of aromatic (e.g. naphtha-
lene and aniline)[18] and heteroaromatic (e.g. quinoline, isoqui-
noline, and quinolinone)[19] compounds. Activation of the tri-
fluoromethyl group could be accomplished in ethereal solvents
(tetrahydrofuran and diethyl ether) by using non-nucleophilic
bases (e.g. lithium N,N-diisopropylamide, lithium hexamethyldi-
silazane, and potassium tert-butoxide) at temperatures typically
ranging between ꢀ78 and 238C. In many cases, displacement
of the postulated ortho-difluoroquinone methide intermediate
preceded intramolecular cyclization to afford the heteroatom-
substituted arenes. In a few cases, the fluorinated product
could be isolated, but no correlation between substitution pat-
tern and reaction pathway could be found. With the aim to
identify a substrate class that exclusively leads to the fluorine-
Solvent[a] Base
Equiv T [8C] t [h][c] 10 [%][c] 11 [%][c]
1
2
3
4
5
6
7
8
9
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
KOtBu
5.0
KHMDS 5.0
CsOtBu 5.0
120
120
120
120
120
120
120
23
1.0
1.0
1.0
0.5
0.5
0.5
1.0
6.0
58
55
43
13
14
–
–
–
15
62
55
–
66
79
58
–
LiOtBu
NaOtBu 5.0
KDMSO 5.0
Cs2CO3
KOtBu
KOtBu
KOtBu
KOtBu
KOtBu
5.0
5.0
5.0
5.0
5.0
3.0
2.5
–
–
60
60
12
44
22
39
58
31
37
–
10 DMSO
11 DMSO
12 DMSO
160
120
120
0.5
1.0
6.0
7.0
3.0
0.5
0.5
0.5
0.5
0.5
1.0
–
24
–
13 sulfolane KOtBu
7.0[b] 120
14 DMF
15 HMPA
16 NMP
17 DMPU
18 dioxane
19 toluene
20 tBuOH
KOtBu
KOtBu
KOtBu
KOtBu
KOtBu
KOtBu
KOtBu
5.0
5.0
5.0
5.0
5.0
5.0
5.0
120
120
120
120
120
120
120
–
2
11
28
–
–
30
11
–
–
55
[a] 1m solution. [b] One equivalent of base was added per hour. [c] Isolat-
ed yields.
containing product, we decided to investigate this transforma- was observed (entry 6). Using inorganic bases, such as cesium
tion in more detail.
carbonate (entry 7), or Schwesinger’s P1-tBu base[24] (not
shown) led to partial isomerization of the double bond, but no
product was obtained. We then examined the influence of
temperature, time, and substrate to base ratio. When the reac-
Results and Discussion
Herein, we report the use of readily available 2-allyl-3-(trifluoro- tion was performed at 238C (entry 8), no cyclization product
methyl)phenols for the synthesis of 5-fluoronaphthalen-1-ols. was formed, whereas only trace amounts were detected at
At the outset of this study, model substrate 9, prepared from 608C (entry 9). At higher temperatures (1608C), the reaction
inexpensive 4-(trifluoromethyl)salicylic acid in three steps, was was less clean and the yield decreased to 44% (entry 10). Reac-
exposed to various reaction conditions beginning with a set of tions were generally complete in one hour for 9. Reducing the
different bases (Table 1). Upon treating a solution of 9 in di- equivalents of base (entry 11 and 12) resulted in incomplete
methyl sulfoxide (DMSO, 1m) with an excess of potassium tert- conversion even after prolonged reaction times (entry 12).
butoxide (entry 1) or potassium hexamethyldisilazane (KHDMS,
A solvent screen using the model system, 9, revealed that
entry 2) at 1208C, naphthol 10 could be isolated in 58 and the reaction proceeded most efficiently in dry dimethyl sulfox-
55% yield, respectively. In this context, we also observed ide or sulfolane (entry 13).[25,26] A sluggish reaction was ob-
a strong counterion effect, with potassium and cesium ions served when the dipolar, non-protic solvents N,N-dimethylfor-
(entry 3) being more efficient than lithium (entry 4) and mamide (DMF, entry 14), hexamethylphosporamide (HMPA,
sodium ions (entry 5). Previous studies on carbanion formation entry 15), N-methyl-2-pyrrolidone (NMP, entry 16), and N,N’-di-
in dimethyl sulfoxide by using alkali metal tert-butoxides methylpropyleneurea (DMPU, entry 17) were used, though
showed a similar overall order of reactivity.[20,21] Sublimed po- some of these solvents are known to break down the oligo-
tassium tert-butoxide, which exists in a cubane-like tetrameric meric structures of potassium tert-butoxide to produce
structure [KOtBu]4, promoted the reaction with equal efficien- a highly basic media.[26c] Toluene (entry 18) and dioxane
cy.[22] Nelson et al. reported that a suspension of potassium (entry 19) also gave lower yields, and the polar protic solvent
tert-butoxide in dimethyl sulfoxide contains low concentrations tert-butanol (entry 20) only led to partial isomerization (55%
of potassium dimsyl (KDMSO).[23] On the basis of these results, yield).
we speculated that potassium dimsyl might be the active pro-
When samples were taken out of the reaction mixtures after
moter. However, when 9 was exposed to a suspension of po- 6, 12, 18, and 24 min, and analyzed by 19F NMR spectroscopy
tassium dimsyl, freshly prepared from dimethyl sulfoxide and (1m in CDCl3, C6F6 internal standard), we noticed that 9 (d=
potassium hydride, only decomposition of the starting material ꢀ62.9 ppm) was immediately consumed, and smoothly con-
&
&
Chem. Eur. J. 2014, 20, 1 – 7
2
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!