Relay propagation of crowding: The trifluoromethyl group as both an emitter and transmitter of steric pressure

Article abstract of DOI:10.1002/ejoc.200500728

Whereas lithium 2,2,6,6-tetramethylpiperidide (LITMP) abstracts a proton exclusively from the 4-position of 3-bromobenzotrifluoride, the same base attacks selectively the 2-position when employed in the presence of N,N,N′,N″,N″-pentamethyldiethylenetriami

Full text of DOI:10.1002/ejoc.200500728

FULL PAPER
DOI: 10.1002/ejoc.200500728
Relay Propagation of Crowding: The Trifluoromethyl Group as Both an
Emitter and Transmitter of Steric Pressure
Manfred Schlosser,*[a,b] Fabrice Cottet,[a] Christophe Heiss,[a] Olivier Lefebvre,[a]
Marc Marull,[a] Eric Masson,[a] and Rosario Scopelliti[a]
Keywords: Buttressing / Halogen substituents / Metalation / Pyridines / Quinolines / Steric effects / Trifluoromethyl groups
Whereas lithium 2,2,6,6-tetramethylpiperidide (LITMP) ab-
stracts a proton exclusively from the 4-position of 3-bromo-
tions. However, the buttressing caused by the introduction of
a methoxy group at the peri-(5-)position impedes the depro-
benzotrifluoride, the same base attacks selectively the 2-po- tonation of either bromo(trifluoromethyl)quinoline. Com-
sition when employed in the presence of N,N,NЈ,NЈЈ,NЈЈ-
pentamethyldiethylenetriamine and potassium tert-butoxide
(“Faigl mix”). 1-Bromo-3,5-bis(trifluoromethyl)benzene also
undergoes metalation at the 2-position but [2-bromo-4-(tri-
fluoromethyl)phenyl]silane does not react at all, evidently
locked up by a C–SiR3/C–Br buttressing effect. 2-Bromo-4-
(trifluoromethyl)pyridine, aza-analogous to the parent model
arene 3-bromobenzotrifluoride, and both the benzo-aza-
analogous 2-bromo-4-(trifluoromethyl)quinoline and its re-
gioinverted isomer 4-bromo-2-(trifluoromethyl)quinoline are
again readily deprotonated at the Br- and CF3-flanked posi-
pared with methoxy, a peri-methyl substituent (to be assimi-
lated to an ortho-tert-butyl substituent) exerts a smaller but-
tressing effect. Although 4-bromo-5,7-dimethoxy-4-(trifluoro-
methyl)quinoline proves to be again totally inert towards
bases,
4-bromo-5,7-dimethyl-4-(trifluoromethyl)quinoline
can be lithiated at the 3-position. Obviously, methoxy is more
powerful as an emitter of steric pressure than methyl is and
bromine is superior to trifluoromethyl as a transmitter of ste-
ric pressure.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2006)
clusively from the 4-position if used alone and exclusively
from the 2-position when in the presence of stoichiometric
amounts of potassium tert-butoxide. With tert-butyllithium
in tetrahydropyran metalation occurs concomitantly, in
equal proportions, at the 4- and 5-position.[1] Finally,
LITMP promotes the deprotonation of 1,2,4-tris(trifluoro-
methyl)benzene at the 5- and 6-position in the ratios of 25:1
and 8:1 when the reaction is accomplished in diethyl ether
(at –25 °C) and tetrahydrofuran (at –75 °C), respectively.[1,6]
Introduction
The trifluoromethyl group exerts one of the strongest in-
ductively electron-withdrawing effects so far recognized.[1,2]
At the same time, it is a relatively bulky substituent.[3] Early
attempts to metalate (trifluoromethyl)benzene with butyl-
lithium in refluxing diethyl ether proved regiochemically
unclean, affording, after trapping with dry ice, a 100:40:1
mixture of 2-, 3- and 4-(trifluoromethyl)benzoic acid in
moderate yield (33–48%).[4,5] However, (trifluoromethyl)-
benzene can be effectively and selectively metalated at the
ortho-position when superbasic mixed-metal reagents such
as methyllithium in the presence of potassium tert-butoxide
are employed.[1] 1,4-Bis(trifluoromethyl)benzene undergoes
almost quantitative deprotonation when lithium 2,2,6,6-tet-
ramethylpiperidide (LITMP) serves as the base.[1] However,
the isomeric 1,2-bis(trifluoromethyl)benzene causes the first
surprise. After consecutive treatment with LITMP and car-
bon dioxide the 2,3- and 3,4-bis(trifluoromethyl)benzoic ac-
ids are isolated in a 2:1 ratio and a total yield of 80%.[1]
1,3-Bis(trifluoromethyl)benzene is an even more capricious
substrate. LITMP in tetrahydrofuran abstracts a proton ex-
[a] Institute of Chemical Sciences and Engineering, Ecole Polytech-
nique Fédérale, BCh
1015 Lausanne, Switzerland
E-mail: manfred.schlosser@epfl.ch
[b] Faculté des Sciences, Université, BCh
1015 Lausanne, Switzerland
Eur. J. Org. Chem. 2006, 729–734
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
729

M. Schlosser et al.
FULL PAPER
No low molecular weight products are obtained at all, when
this substrate is treated with alkyllithium or superbasic
mixed-metal reagents as nucleophilic aromatic addition su-
persedes metalation.
This last example is most instructive as it simultaneously
reveals two anomalies. The pronounced regiodiscrimination
against an attack at the 6-position must be imputed to a
“buttressing effect”.[7,8] The same holds for the regiochem-
ically unselective metalation of 1,2-bis(trifluoromethyl)ben-
zene. The vicinal proximity of two trifluoromethyl groups
deprives each of them of flexibility. Held tightly at their
respective places they cannot get out of the trajectory of the
approaching base and thus must interfere with the proton
transfer process if taking place at a neighboring position.
On the other hand, also non-vicinal trifluoromethyl groups
can give rise to steric hindrance as illustrated by the case
of 1,3-bis(trifluoromethyl)benzene mentioned above. Bulky
reagents like tert-butyllithium or LITMP, unless activated
by potassium tert-butoxide, are unable to enter the con-
gested region flanked by the two trifluoromethyl substitu-
ents and to accomplish there a permutational hydrogen/me-
tal interconversion.
A seemingly straightforward way to direct the base to the
2-position of 3-bromobenzotrifluoride should be the block-
ing of the 4-position by a trialkylsilyl group, one of the
favorite “toolbox methods”.[16] However, this attempt failed
completely, [2-bromo-4-(trifluoromethyl)phenyl]triethylsi-
lane (3) proving totally inert toward LITMP or the Faigl
mix at –100 °C and –75 °C. Obviously due to the but-
tressing effect exerted by the bulky silyl substituent and
transmitted by the bromine atom, proton abstraction from
the 2-position was definitively suppressed.
The 3-halobenzotrifluoride series offers another impres-
sive illustration of how sensitive bases monitor steric screen-
ing. Both 3-fluoro- and 3-chlorobenzotrifluoride react with
butyllithium solely at the doubly activated 2-position
whereas bulkier reagents such as sec-butyllithium, alone or
complexed to N,N,NЈ,NЈЈ,NЈЈ-pentamethylethylenediamine
(PMDTA), metalate preferentially or exclusively the 4-posi-
tion.[9,10] No such “optional site selectivity”[11–14] was found
so far with 3-bromobenzotrifluoride as the substrate,
mainly because the choice of bases must be kept restricted
to metal dialkylamides as organolithium reagents would
give rise to permutational halogen/metal interconversions.
Another simple way to prevent proton abstraction from
the 4-position is to replace the CH unit by a nitrogen atom.
The latter might at the same time facilitate the deproton-
ation by virtue of its electron-withdrawing effect. 2-Bromo-
4-(trifluoromethyl)pyridine and even 2-iodo-4-(trifluoro-
methyl)pyridine had been found to react indeed smoothly
with amide-type bases such as lithium diisopropylamide to
furnish the 2-bromo- and 2-iodo-4-(trifluoromethyl)pyr-
idine-3-carboxylic acids (54 and 51%, respectively) after
carboxylation.[17] In the same way, 2-bromo-4-(trifluoro-
methyl)quinoline[18] and 4-bromo-2-(trifluoromethyl)quino-
line[19] have been successfully subjected to the lithiation/car-
boxylation sequence.
Results and Discussion
We wanted first to clarify whether it was really imposs-
ible to abstract a proton from the doubly activated, though
crowded position flanked by the trifluoromethyl group and
the bromine atom. In fact, when the three-component base
(“Faigl mix”[15]) consisting of LITMP, PMDTA and potas-
sium tert-butoxide was used, only the 2-position and no
longer the 4-position was attacked. However, the yield of
the 2-bromo-6-(trifluoromethyl)benzoic acid (1) isolated af-
ter carboxylation remained poor (24%). In contrast, 1-
bromo-3,5-bis(trifluoromethyl)benzene, a more acidic sub-
However, 5-fluoro- and 5-methoxy-substituted 4-bromo-
strate, reacted even with unactivated LITMP smoothly to 2-(trifluoromethyl)quinolines (4 and 5) were reported to re-
afford almost quantitatively the 2-bromo-4,6-bis(trifluoro- sist the attack of all amide bases.[19] Meanwhile we have
methyl)benzoic acid (2; 83% of purified product) after car- extended the investigation to additional model compounds.
boxylation.
4-Bromo-5,7-dimethyl-2-(trifluoromethyl)quinoline (6) and
730
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 729–734

The Trifluoromethyl Group as Both an Emitter and Transmitter of Steric Pressure
FULL PAPER
2-bromo-5,7-dimethoxy-4-(trifluoromethyl)quinoline
(8) pattern. 4-Bromo-2,7-bis(trifluoromethyl)quinolines (14), 4-
were also found to be totally inert toward lithium diisopro- bromo-8-fluoro-2-(trifluoromethyl)quinolines (15) and 2-
pylamide or LITMP. On the other hand, 4-chloro-5-fluoro- bromo-8-fluoro-2-(trifluoromethyl)quinolines (16) were se-
2-(trifluoromethyl)quinoline (7) and 2-bromo-5,7-dimethyl- lected as sterically unbiased reference compounds. As this
4-(trifluoromethyl)quinoline (9) did react under such condi- is the case with any ordinary quinoline derivative (e.g.,
tions. Upon carboxylation, the 4-chloro-5-fluoro-2-(trifluo- quinoline[21] itself) too, all bond angles approximate 120°
romethyl)quinoline-3-carboxylic acid (10; 61%) and 2- except those at the junction, where the C(4),C(4a),C(5) an-
bromo-5,7-dimethyl-4-(trifluoromethyl)quinoline-3-carbox- gle is widened to 125° and the C(8),C(8a),N angle is short-
ylic acid (11; 26%) were formed.
ened to 118°, and except for the C(9),N,C(1) angle which
decreases to 117°. The introduction of a substituent into
the 5-position inevitably causes severe skeletal distortions
in the vicinity. The peri-methyl group in 2-bromo-5,7-di-
methyl-4-(trifluoromethyl)quinolines (9) forces the trifluo-
romethyl neighbor to make room for it by expanding
the C(4a),C(4),C(CF3) angle from 119° to 123° at the
expense of the C(CF3),C(4),C(3) angle which shrinks from
120° to 116°. The peri-methoxy group in 2-bromo-5,7-
dimethoxy-4-(trifluoromethyl)quinolines has an even more
pronounced effect and, still slightly more, in 4-bromo-5-
methoxy-2-(trifluoromethyl)quinoline (5), enlarging the
C(4a),C(4),C(CF3) angles to 125° and compressing the
C(CF3),C(4),C(5) angles to 116° and 115°, respectively. In
all these compounds the C(5),C(4a),C(4) angle is widened
to 127° or 128°. The most striking difference between the
5-methyl and 5-methoxy series is the orientation of the peri
substituent. Steric repulsion causes the methyl group in
On the basis of such findings we dare to make a counter-
intuitive prediction. Although 4-bromo-2-(trifluoromethyl)-
pyridine (12) is readily deprotonated by LIDA or LITMP
at the 3-position, we expect any attempt to metalate 4-
bromo-2,5-bis(trifluoromethyl)pyridine (13) to be doomed
to failure despite the increased acidity of the latter sub-
strate.
quinolines
9 to bend away from its CF3 neighbor
[C(4),C(5),C(CH3) angle 125°, C(CH3),C(5),C(6) angle
116°]. In contrast, the methoxy group leans over to get
closer to the CF3 substituent [C(4),C(5),C(CH3) angles of
116° and 115°, C(OCH3),C(5),C(6) angles of 124° and 126°
in compounds 8 and 5, respectively].
When evaluating the role of substituents in the given con-
text one has to differentiate between emitters and transmit-
ters of buttressing effects. For example, the perturbation af-
fecting 4-bromo-5-fluoro-2-(trifluoromethyl)quinoline (4)
[19]
originates from the fluorine atom and is propagated by
the bromine atom which, being buttressed, cannot move
away when the base tries to approach and abstract a proton
from the 3-position. Comparing the various substrates 4–
11 for their propensity to undergo metalation at the 3-posi-
tion and combining this with knowledge derived from other
sources[8,17,20] one can rank typical substituents according
to their emitter and transmitter aptitudes.
Emitter: R3Si Ͼ F3C Ϸ H3CO Ͼ H3C Ͼ H
Transmitter: I Ͼ Br Ͼ F3C Ϸ Cl ϾϾ F Ͼ H
To probe such emitter and transmitter effects not only
on reactivity but also on structure we have carried out a
On the whole, the comparison between the ground-state
crystallographic study of three 4-bromo-2-(trifluoromethyl)- structures of the quinolines 5, 8 and 9 and those of the peri-
quinolines and equally of three 2-bromo-4-(trifluoromethyl)- unsubstituted model compounds 14, 15 and 16 does not
quinolines, each one harboring a characteristic substituent reveal conspicuous features that would immediately forecast
Eur. J. Org. Chem. 2006, 729–734
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
731

M. Schlosser et al.
FULL PAPER
the residue distilled; colorless liquid; b.p. 68–69 °C/2 Torr; yield:
11.2 g (75%). H NMR: δ = 7.77 (s, 1 H), 7.55 (d, J = 8.0 Hz, 1
trouble with the deprotonation of the peri-substituted sub-
strates. The CF3–C(4)/C(2)–Br “yawning angle” averages
116° or 114° depending on whether a peri substituent is
absent or present. It is true, this difference is more pro-
nounced in the regioisomeric series, the Br–C(4)/C(2)–CF3
angle decreases from 121° to approximately 112° upon the
introduction of a peri substituent. However, there are no
steric constraints that should definitively prevent a base
from frontally approaching the C(3)–H bond. Therefore, we
maintain our view[20] that electronic reasons force the sub-
stituents surrounding a center undergoing metalation to
step away from the trajectory of the reagent. This dodging
movement is blocked if there are buttressing substituents.
1
H), 7.51 (d, J = 8.0 Hz, 1 H) ppm. 13C NMR: δ = 146.1, 136.4,
132.7 (q, J = 33 Hz), 130.4, 129.3 (q, J = 4 Hz), 123.2 (q, J =
273 Hz), 123.0 (q, J = 4 Hz), –0.9 ppm. MS (c.i.): m/z (%) = 315
(0) [M+ + NH4], 297 (3) [M+], 281 (49), 203 (29), 140 (100).
C10H12BrF3Si (297.19): calcd. C 40.42, H 4.07; found C 40.34, H
4.00.
4,4,4-Trifluoro-N-(3,5-dimethoxyphenyl)-3-oxobutanamide: Ethyl 2-
oxo-1,1,1-trifluorobutanoate (58 mL, 74 g, 0.40 mol) and 3,5-di-
methoxyaniline (31 g, 0.20 mol) were heated to 130 °C. A small
amount of water (4 mL) was added every 30 min for the duration
of 6 h. Upon cooling in an ice bath, the anilide precipitated. It
was collected by filtration and washed with hexanes before being
recrystallized from ethyl acetate and hexanes; colorless prisms;
1
yield: 46.1 g (79%). H NMR*: δ = 6.41 (d, J = 2.2 Hz, 1 H), 6.32
(d, J = 2.3 Hz, 1 H), 6.07 (s, 1 H), 3.99 (s, 3 H), 3.82 (s, 3 H), 2.97
(d, J = 16.8 Hz, 1 H), 2.85 (d, J = 16.8 Hz, 1 H) ppm. 13C NMR*:
δ = 165.9, 162.1, 160.1, 140.5, 126.2 (q, J = 289 Hz), 99.5, 94.4,
93.4, 73.6 (q, J = 30 Hz), 56.0, 54.8, 39.0 ppm. MS (c.i.): m/z (%)
= 309 (0) [M+ + NH4], 292 (22) [M+ + 1], 291 (11) [M+], 274 (4),
222 (100). C12H12F3NO4 (291.23): calcd. C 49.49, H 4.15; found C
49.52, H 4.13.
Conclusions
Success and failure alternate when CF3-bearing model
compounds of the arene and quinoline type are exposed to
a variety of metalating agents. Although the introduction
of an additional electron-withdrawing substituent (such as
F, CF3 or OCH3) increases the (thermodynamic) acidity of
the substrate, it does not necessarily facilitate its (kinetic)
deprotonation and may even prevent it. This counterintu-
itive conclusion can be rationalized by taking buttressing
effects into account.
Buttressing effects have been ignored for too long time.
We believe the scope of this newly recognized phenomenon
to be far-reaching. It may help to rationalize anomalies
which otherwise remain unexplained.
4,4,4-Trifluoro-N-(3,5-dimethylphenyl)-3-oxobutanamide: Prepared
analogously from 3,5-dimethylaniline (25 mL, 24 g, 0.20 mol); col-
orless prisms; yield: 47.3 g (90%). 1H NMR: δ = 7.43 (s, 1 H), 7.10
(s, 2 H), 6.83 (s, 1 H), 2.77 (s, 2 H), 2.28 (s, 6 H) ppm. 13C NMR*:
δ = 170.1, 139.2, 138.8, 126.9, 124.0 (q, J = 286 Hz), 118.6, 93.9
(q, J = 32 Hz), 39.2, 21.4 ppm. MS (c.i.): m/z (%) = 277 (25) [M+
+ NH4], 260 (100) [M+ + 1], 259 (25) [M+], 190 (2), 121 (31).
C12H12F3NO2 (259.22): calcd. C 55.60, H 4.67; found C 55.48, H
4.51.
5,7-Dimethoxy-4-(trifluoromethyl)quinolin-2(1H)one:
4,4,4-Tri-
fluoro-N-(3,5-dimethoxyphenyl)-3-oxobutanamide (29 g, 0.10 mol)
was added to 75% aqueous sulfuric acid (0.10 L). The mixture was
heated at 90 °C for 45 min and poured onto crushed ice (0.25 L).
The precipitate formed was collected by filtration and crystallized
from ethanol; colorless needles; m.p. 105–106 °C; yield 23 g (83%).
MS (c.i.): m/z (%) = 292 (0) [M+ + NH4], 273 (100) [M+], 244 (11).
C12H10F3NO3 (273.21): calcd. C 52.76, H 3.69; found C 52.75, H
3.66.
Experimental Section
Generalities: Working practices and abbreviations are specified in
previous articles from this laboratory.[22–24] 1H and (1H-decoupled)
13C NMR spectra were recorded at 400 and 101 MHz, respectively,
relative to the internal standard tetramethylsilane (chemical shift δ
= 0.00 ppm). The samples were dissolved in deuteriochloroform or,
if marked by an asterisk, in hexadeuterioacetone. Mass spectra
were obtained at 70 eV ionization potential while a source tempera-
ture of 200 °C was maintained. Whenever no molecular peak was
observed under such conditions, chemical ionization (“c.i.”) in an
ammonia atmosphere at 100 °C source temperature was applied.
To avoid redundancy, in all cases only the [35Cl] and [79Br] frag-
ments and not the [37Cl] and [81Br] isotopomers are listed.
5,7-Dimethyl-4-(trifluoromethyl)quinolin-2(1H)-one: Prepared anal-
ogously starting from 4,4,4-trifluoro-N-(3,5-dimethylphenyl)-3-
oxobutanamide (26 g, 0.10 mol); colorless needles; m.p. 202–
204 °C; yield: 21 g (85%). 1H NMR (D3CSOCD3): δ = 7.13 (s, 1
H), 7.01 (s, 1 H), 6.97 (s, 1 H), 2.54 (s, 3 H), 2.33 (s, 3 H) ppm.
13C NMR (D3CSOCD3): δ = 159.7, 141.9, 141.5, 136.3 (q, J =
34 Hz), 134.2, 129.5, 123.3 (q, J = 274 Hz), 123.0 (q, J = 8.0 Hz),
115.5, 111.4, 22.5 (q, J = 8 Hz), 21.1 ppm. MS (c.i.): m/z (%) = 259
(0) [M+ + NH4], 242 (64) [M+ + 1], 241 (100) [M+], 226 (19), 198
(15). C12H10F3NO (241.21): calcd. C 59.75, H 4.18; found C 59.74,
H 3.98.
1. Starting Materials
4-Bromo-5-fluoro-2-(trifluoromethyl)quinoline (4) and 4-bromo-5-
methoxy-2-(trifluoromethyl)quinoline (5) have already been re-
ported.[19] The quinolines 6–9 were prepared according to well-es-
tablished procedures as described below.
5,7-Dimethyl-2-(trifluoromethyl)quinolin-4(1H)-one: Prepared from
3,5-dimethylaniline (1, 31 mL, 30 g, 0.25 mol) and ethyl 4,4,4-tri-
fluoro-3-oxobutanoate (38 mL, 50 g, 0.25 mol) as described in
analogous cases;[19] m.p. 251–253 °C (reprod.; colorless needles
from ethanol); yield: 31.4 g (52%). 1H NMR (D3CSOCD3): δ =
7.51 (s, 1 H), 7.11 (s, 1 H), 6.83 (s, 1 H), 2.81 (s, 3 H), 2.41 (s, 3
H) ppm. 13C NMR (D3CSOCD3): δ = 170.0, 147.2, 143.1 (q, J =
34 Hz), 141.0, 136.7, 130.3, 122.9, 121.2 (q, J = 275 Hz), 119.8,
103.1, 23.5, 21.0 ppm. MS (c.i.): m/z (%) = 260 (0) [M+ +NH4],
1-[2-Bromo-4-(trifluoromethyl)phenyl]trimethylsilane
(3):
At
–100 °C 2,2,6,6-tetramethylpiperidine (8.4 mL, 7.1 g, 50 mmol), 1-
bromo-3-(trifluoromethyl)benzene (11 g, 50 mmol) and chlorotri-
methylsilane (6.3 mL, 5.4 g, 50 mmol) were added consecutively to
butyllithium (50 mmol) in tetrahydrofuran (65 mL) and hexanes
(35 mL). After 2 h at –100 °C, the solvents were evaporated and
732
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 729–734

The Trifluoromethyl Group as Both an Emitter and Transmitter of Steric Pressure
FULL PAPER
242 (61) [M+ + 1], 241 (100) [M+], 193 (9), 91 (3). C12H10F3NO
(241.21): calcd. C 59.75, H 4.18; found C 59.70, H 4.08.
(CDCl3): δ = 164.1 (d, J = 255 Hz), 149.0 (d, J = 13 Hz), 148.9 (q,
J = 35 Hz), 144.5, 126.4 (d, J = 10 Hz), 124.0, 120.6 (q, J =
276 Hz), 120.1 (d, J = 26 Hz), 116.5 (quint, J = 2 Hz), 114.0 (d, J
= 21 Hz) ppm. MS (c.i.): m/z (%) = 267 (0) [M+ + NH4], 250 (100)
[M+ + 1], 249 (86) [M+], 180 (45), 145 (18). C10H4ClF4N (249.59):
calcd. C 48.12, H 1.62; found C 48.39, H 1.39.
4-Bromo-5,7-dimethyl-2-(trifluoromethyl)quinoline (6): Prepared
from 5,7-dimethyl-2-(trifluoromethyl)quinolin-4(1H)-one (24 g,
0.10 mol) and phosphoric tribromide (29 g, 0.10 mol) as described
in analogous cases;[19] m.p. 172–174 °C (reprod.; colorless needles
1
from methanol); yield: 23.4 g (77%). H NMR: δ = 7.94 (s, 1 H),
2. Metalation and Subsequent Carboxylation of Trifluoromethyl-
Substituted Benzenes and Quinolines
7.89 (s, 1 H), 7.35 (s, 1 H), 3.06 (s, 3 H), 2.51 (s, 3 H) ppm. 13C
NMR: δ = 146.9 (q, J = 35 Hz), 145.3, 141.0, 132.2, 131.6, 127.1,
126.8, 124.2, 119.0 (q, J = 275 Hz), 118.6, 21.2, 18.4 ppm. MS (c.i.):
m/z (%) = 321 (0) [M+ + NH4], 304 (100) [M+ + 1], 303 (41) [M+],
260 (15), 127 (3). C12H9BrF3NO (304.11): calcd. C 47.39, H 2.98;
found C 47.50, H 2.92.
2-Bromo-6-(trifluoromethyl)benzoic Acid (1): Potassium tert-butox-
ide (2.8 g, 25 mmol) and N,N,NЈ,NЈ,NЈЈ-pentamethyldiethylenetria-
mine (3.7 mL, 2.9 g, 25 mmol) were added consecutively to the
solution prepared from 2,2,6,6-tetramethylpiperidine (4.2 mL,
3.5 g, 25 mmol) and butyllithium (25 mmol) in tetrahydrofuran
(50 mL) and hexanes (16 mL), cooled in a dry ice/methanol bath.
After the dissolution of the potassium tert-butoxide, the mixture
was diluted with diethyl ether (15 mL) and pentanes (15 mL). At
–115 °C, 3-bromo(trifluoromethyl)benzene (5.6 g, 25 mmol) was
added. After 45 min at –115 °C, the reaction mixture was poured
onto an excess of freshly crushed dry ice covered with tetra-
hydrofuran (25 mL). It was later extracted with 1.0 m aqueous so-
dium hydroxide (3×50 mL). The combined aqueous layers were
washed with diethyl ether (2×50 mL) and acidified to pH = 1 with
concentrated hydrochloric acid. Upon extraction with diethyl ether
(2×50 mL), drying and concentration, a viscous brown residue was
left behind. It was extracted with boiling hexanes (3×75 mL).
Upon concentration and cooling, slightly greenish needles formed;
m.p. 129–130 °C (ref.[9] m.p. 131–133 °C); 1.62 g (24%).
2-Bromo-5,7-dimethoxy-4-(trifluoromethyl)quinoline (8): Prepared
analogously
from
5,7-dimethoxy-4-(trifluoromethyl)quinolin-
2(1H)-one (14 g, 50 mmol); colorless needles from methanol; m.p.
107–109 °C; yield: 8.7 g (52%). 1H NMR: δ = 7.70 (s, 1 H), 7.08
(d, J = 2.2 Hz, 1 H), 6.66 (d, J = 2.2 Hz, 1 H), 3.96 (s, 3 H), 3.94
(s, 3 H) ppm. 13C NMR: δ = 162.3, 155.8, 152.6, 141.8, 135.6 (q,
J = 34 Hz), 124.4, 122.4 (q, J = 274 Hz), 121.4 (q, J = 8 Hz), 101.2,
100.9, 56.2, 55.8 ppm. MS (c.i.): m/z (%) = 353 (0) [M+ + NH4],
336 (66) [M+ + 1], 335 (42) [M+], 226 (37), 185 (46), 86 (100).
However, the product proved contaminated from the beginning by
some 10% of 2,7-dibromo-5-methoxy-4-(trifluoromethyl)quinoline
[1H NMR: δ = 7.89 (s, 1 H), 7.85 (s, 1 H), 7.14 (s, 1 H), 4.00 (s, 3
H) ppm]. After repetitive recrystallization from methanol, a 1:1
complex (“molecular compound”) composed of this by-product
and quinoline 8 was obtained; yellowish prisms; m.p. 103–107 °C.
C23H15Br3F6N2O3 (721.08): calcd. C 38.31, H 2.10; found C 38.69,
H 2.48.
2-Bromo-4,6-bis(trifluoromethyl)benzoic Acid (2): 3,5-Bis(trifluoro-
methyl)bromobenzene (4.3 mL, 7.3 g, 25 mmol) was added to the
solution prepared from 2,2,6,6-tetramethylpiperidine (4.2 mL,
3.5 g, 25 mmol) and butyllithium (25 mmol) in tetrahydrofuran
(50 mL) and hexanes (15 mL), kept in a dry ice/methanol bath.
After 45 min at –75 °C, the reaction mixture was poured onto an
excess of freshly crushed dry ice covered with tetrahydrofuran
(25 mL). The volatiles were evaporated and the residue partitioned
between diethyl ether (100 mL) and 6.0 m hydrochloric acid
(50 mL). Upon drying and evaporation, a slightly yellow solid was
obtained (7.81 g, m.p. 161–165 °C). Direct sublimation afforded
colorless needles; 7.00 g (83%); m.p. 174–175 °C. 1H NMR: δ =
8.12 (s, 1 H), 7.94 (s, 1 H) ppm. 13C NMR: δ = 169.5, 136.1, 133.7
(q, J = 34 Hz), 133.5 (q, J = 3 Hz), 130.2 (q, J = 34 Hz), 122.5
(hept, J = 4 Hz), 122.0 (q, J = 274 Hz), 121.9 (q, J = 275 Hz),
121.4 ppm. C9H3BrF6O2 (337.01): calcd. C 32.08, H 0.90, found C
31.98, H 0.89.
2-Bromo-5,7-dimethyl-4-(trifluoromethyl)quinoline (9): Analogously
starting from 5,7-dimethyl-4-(trifluoromethyl)quinolin-2(1H)-one
(12 g, 50 mmol); colorless needles from methanol; m.p. 60–62 °C;
1
yield: 11 g (74%). H NMR: δ = 7.82 (s, 1 H), 7.75 (s, 1 H), 7.36
(s, 1 H), 2.74 (s, 3 H), 2.49 (s, 3 H) ppm. 13C NMR: δ = 151.6;
141.1, 140.0, 135.4 (q, J = 33 Hz), 135.2, 133.6, 128.2, 123.7 (q, J
= 8 Hz), 122.8 (q, J = 276 Hz), 120.4, 22.8 (q, J = 8 Hz), 21.2 ppm.
MS (c.i.): m/z (%) = 312 (0) [M+ + NH4], 304 (100) [M+ + 1], 303
(73) [M+]. C12H9BrF3N (304.11): calcd. C 47.39, H 2.98; found C
46.96, H 2.55.
4-Chloro-5-fluoro-2-(trifluoromethyl)quinoline (7) and 4-Chloro-7-
fluoro-2-(trifluoromethyl)quinoline: At 125 °C, a mixture of 5-fluo-
ro- and 7-fluoroquinolin-4(1H)-ones[19] (23 g, 0.10 mol) was slowly
added to phosphoric trichloride (18 mL, 31 g, 0.20 mol). After 2 h
at this temperature, the mixture was poured into ice/water (0.40 L).
The insoluble material was collected. A 28:72 mixture of 4-chloro-
5-fluoro- and 4-chloro-7-fluoro-2-(trifluoromethyl)quinoline was
obtained according to gas chromatography (30 m, DB 1701,
110 °C; 30 m, DB 210, 100 °C) and separated by elution from silica
gel (0.60 L) with hexanes. 4-Chloro-5-fluoro-2-(trifluoromethyl)quin-
4-Chloro-5-fluoro-2-(trifluoromethyl)quinoline-3-carboxylic
Acid
(10): Diisopropylamine (3.5 mL, 2.5 g, 25 mmol) and 4-chloro-5-
fluoro-2-(trifluoromethyl)quinoline (7; 6.2 g, 25 mmol) were added
consecutively to a solution of butyllithium (25 mmol) in tetra-
hydrofuran (30 mL) and hexanes (15 mL), kept in a dry ice/meth-
anol bath. After 2 h at –75 °C, the mixture was poured onto an
1
oline (7): M.p. 31–33 °C; yield: 6.08 g (24%). H NMR (CDCl3): δ
= 8.08 (d, J = 8.6 Hz, 1 H), 7.81 (m, 2 H), 7.42 (dd, J = 11.8, excess of freshly crushed dry ice. A 1.8 m ethereal solution (15 mL)
8.0 Hz, 1 H) ppm. 13C NMR (CDCl3): δ = 157.3 (d, J = 262 Hz),
of hydrochloric acid was added. The solvents were evaporated and
149.5, 148.3 (q, J = 35 Hz), 141.8 (d, J = 2 Hz), 131.0 (d, J = the residue was crystallized from a 4:1 (v/v) mixture of hexanes and
10 Hz), 127.0 (d, J = 5 Hz), 120.7 (q, J = 276 Hz), 119.3 (quint, J
= 2 Hz), 118.0 (d, J = 9 Hz), 114.6 (d, J = 22 Hz) ppm. MS (c.i.):
ethyl acetate; colorless prisms; m.p. 225–227 °C (decomp.); yield:
1
4.48 g (61%). H NMR*: δ = 8.14 (d, J = 8.0 Hz, 1 H), 8.07 (ddd,
m/z (%) = 267 (0) [M+ + NH4], 250 (100) [M+ + 1], 249 (80) [M+], J = 8.3, 7.7, 4.8 Hz, 1 H), 7.72 (ddd, J = 12.5, 7.7, 1.3 Hz, 1 H)
180 (8), 99 (4). C10H4ClF4N (249.59): calcd. C 48.12, H 1.62; found
ppm. 13C NMR*: δ = 164.9, 158.5 (d, J = 261 Hz), 148.8, 144.6
(qd, J = 35, 2 Hz), 139.0 (d, J = 3 Hz), 133.4 (d, J = 10 Hz), 127.8,
C 47.83, H 1.44. 4-Chloro-7-fluoro-2-(trifluoromethyl)quinoline:
M.p. 37–38 °C; yield: 15.5 g (62%). 1H NMR (CDCl3): δ = 8.34 128.0 (d, J = 5 Hz), 120.8 (q, J = 276 Hz), 118.0 (d, J = 8 Hz),
(dd, J = 9.3, 5.8 Hz, 1 H), 7.89 (dd, J = 9.6, 2.6 Hz, 1 H), 7.80 (s,
117.0 (d, J = 22 Hz) ppm. MS (c.i.): m/z (%) = 311 (0) [M+ + NH4],
294 (100) [M+ 1], 293 (87) [M+], 276 (50), 194 (29).
1 H), 7.68 (ddd, J = 10.6, 8.0, 2.6 Hz, 1 H) ppm. 13C NMR
+
Eur. J. Org. Chem. 2006, 729–734
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
733

M. Schlosser et al.
FULL PAPER
C11H4ClF4NO2 (293.60): calcd. C 45.00, H 1.37; found C 44.79, H
1.12.
Acknowledgments
This work was financially supported by the Schweizerische Natio-
nalfonds zur Förderung der wissenschaftlichen Forschung, Bern
(grant 20-100Ј336-02). The authors wish also to express their grati-
tude to the Ishihara Sangyo Kaisha (ISK) Company, Tokyo, for a
generous gift of polyhalogenated pyridines.
2-Bromo-5,7-dimethyl-4-(trifluoromethyl)quinoline-3-carboxylic
Acid (11): Prepared analogously from 2-bromo-5,7-dimethyl-4-(tri-
fluoromethyl)quinoline (9; 4.6 g, 15 mmol). The product was parti-
tioned between water (50 mL) and diethyl ether (3×50 mL). The
combined organic layers were concentrated and the residue was
crystallized from ethanol; colorless prisms; m.p. 164–165 °C (de-
1
comp.); yield: 1.36 g (26%). H NMR*: δ = 7.79 (s, 1 H), 7.64 (s,
[1] M. Schlosser, F. Mongin, J. Porwisiak, W. Dmowski, H. H.
Büker, N. M. M. Nibbering, Chem. Eur. J. 1998, 4, 1281–1286.
[2] B. Kovacˇevic´, Z. B. Maksic´, M. Primorac, Eur. J. Org. Chem.
2003, 3777–3783.
[3] B. E. Smart, in: Organofluorine Chemistry: Principles and Com-
mercial Applications (Eds.: R. E. Banks, B. E. Smart, J. C. Tat-
low), Plenum Press, New York, 1994, pp. 57–88, spec. 81.
[4] J. D. Roberts, D. Y. Curtin, J. Am. Chem. Soc. 1946, 68, 1658–
1660.
1 H), 2.78 (q, J = 2.9 Hz, 3 H), 2.57 (s, 3 H) ppm. 13C NMR*: δ
= 166.6, 151.0, 143.1, 139.0, 137.2, 134.9, 131.9 (q, J = 32 Hz),
130.9, 128.2, 123.7 (q, J = 276 Hz), 120.7, 23.2 (q, J = 9 Hz),
21.1 ppm. MS (c.i.): m/z (%) = 363 (0) [M+ + NH4], 348 (100) [M+
+ 1], 347 (74) [M+], 303 (14). C13H9BrF3NO2 (348.12): calcd. C
44.85, H 2.61; found C 44.79, H 1.12.
No trace of an acid was obtained when 4-bromo-5-fluoro-2-(tri-
[5] D. A. Shirley, J. R. Johnson, J. P. Hendrix, J. Organomet. Chem.
1968, 11, 209–216.
fluoromethyl)quinoline
(4),
4-bromo-5-methoxy-2-(trifluoro-
methyl)quinoline (5), 4-bromo-5,7-dimethyl-2-(trifluoromethyl)-
quinoline (6) and 2-bromo-5,7-dimethoxy-4-(trifluoromethyl)quin-
oline (8) were subjected to the same and similar reaction condi-
tions.
[6] M. Schlosser, J. Porwisiak, F. Mongin, Tetrahedron 1998, 54,
895–900.
[7] F. Mongin, E. Marzi, M. Schlosser, Chimia 1996, 50, 650–652.
[8] C. Heiss, E. Marzi, M. Schlosser, Eur. J. Org. Chem. 2003,
4625–4629.
3. Crystallography
[9] F. Mongin, O. Desponds, M. Schlosser, Tetrahedron Lett. 1996,
4-Bromo-5-methoxy-2-(trifluoromethyl)quinoline (5),[19] 4-bromo-
2,7-bis(trifluoromethyl)quinoline (14),[25] 4-bromo-8-fluoro-2-(tri-
fluoromethyl)quinoline (15)[19] and 2-bromo-6-trifluoromethoxy-4-
(trifluoromethyl)quinoline (16),[26] were available from previous
work. The preparation of compounds 8 and 9 is described above.
As explained in Subsection 2, quinoline 8 crystallized as a 1:1 mol-
ecular complex together with 2,7-dibromo-5-methoxy-4-(trifluoro-
methyl)quinoline, accidentally formed as a by-product. Although
this contamination certainly constitutes a flaw, it did not compro-
mise the ultimate goal. As the nature of the substituent at the 7-
position does not affect the structure of the heterocyclic ring nor of
the peri-(5-)position, the pertinent data could be acquired without
complication. X-ray diffraction data were collected using the Mo-
Kα (0.71073 Å) radiation and a low-temperature device [T =
140(2) K] on a four-circle kappa goniometer, equipped with an Ox-
ford Diffraction KM4 Sapphire CCD in the case of compounds 8,
9, 14, and 16, whereas for compounds 5 and 15 a marresearch
mar345 IPDS was used. All data reductions were performed with
the CrysAlis RED 1.7.0 software (Oxford Diffraction, Abingdon,
UK, 2003). Absorption correction has been applied to all data sets
except of compound 14. The structures were refined using the full-
matrix least squares on F2 with all non-hydrogen atoms anisotropi-
cally defined. The hydrogen atoms were placed in calculated posi-
tions using the “riding model”. Structure refinements and geomet-
rical calculations were carried out with the SHELXTL software
(University of Göttingen, 1997; Bruker AXS, Madison, 1997). All
crystallographic data the present manuscript refers to are contained
in the CCDC-275711 to -275716 files. They can be downloaded
free of charge from the Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif.
37, 2767–2770.
[10] M. Schlosser, F. Mongin, A. Tognini, unpublished results; see:
A. Tognini, Diploma Thesis, Université de Lausanne, 1996, pp.
7, 8, 20.
[11] M. Schlosser, G. Katsoulos, S. Takagishi, Synlett 1990, 747–
748.
[12] F. Mongin, R. Maggi, M. Schlosser, Chimia 1996, 50, 650–652.
[13] M. Schlosser, Eur. J. Org. Chem. 2001, 3975–3984.
[14] M. Schlosser, in Organometallics in Synthesis: A Manual (Ed.:
M. Schlosser), 2nd ed., Wiley, Chichester, 2002, pp. 1–352,
spec. 253–259.
[15] F. Faigl, E. Marzi, M. Schlosser, Chem. Eur. J. 2000, 6, 771–
777.
[16] M. Schlosser, Angew. Chem. 2005, 117, 380–398; Angew. Chem.
Int. Ed. 2005, 44, 376–393.
[17] F. Cottet, M. Marull, F. Mongin, D. Espinosa, M. Schlosser,
Synthesis 2004, 1619–1624.
[18] O. Lefebvre, M. Marull, M. Schlosser, Eur. J. Org. Chem. 2003,
2115–2121.
[19] M. Marull, M. Schlosser, Eur. J. Org. Chem. 2003, 1576–1588.
[20] J. Gorecka, C. Heiss, R. Scopelliti, M. Schlosser, Org. Lett.
2004, 6, 4591–4593.
[21] J. E. Davies, A. D. Bond, Acta Crystallogr. Sect. E (Struct.
Rep. Online) 2001, 57, o947.
[22] C. Bobbio, M. Schlosser, Eur. J. Org. Chem. 2001, 4533–4536.
[23] C. Heiss, M. Schlosser, Eur. J. Org. Chem. 2003, 447–451.
[24] M. Schlosser, M. Marull, Eur. J. Org. Chem. 2003, 1569–1575.
[25] M. Schlosser, O. Lefebvre, F. Leroux, unpublished results,
2002.
[26] M. Marull, O. Lefebvre, M. Schlosser, Eur. J. Org. Chem. 2004,
54–63.
Received: September 27, 2005
Published Online: November 30, 2005
734
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 729–734