LETTER
Practical Synthesis of N-Cyclopropylanilines
2141
(6) The crude 3¢ also contains 2–3% of 1,1-dianilinocyclo-
propane as a by-product, which is readily dealkoxylated to
give 4 and equimolar amount of aniline (see Scheme 3).
Before we began our study, Gillaspy et al. had reported
the one-pot N-cyclopropylations of amines using large ex-
cess of 1 (4.0–6.0 equiv more than amines) followed by
the reduction by sodium cyanoborohydride. But this
method produced biscyclopropylamines in higher
amounts than monocyclopropylamines.7
BH3·THF
NHAr
NHAr
NHAr
ArNH2
+
In our reaction condition, most of the ethoxy group of [(1-
ethoxycyclopropyl)oxy]trimethylsilane (1) was substitut-
ed by the alkoxy group derived from alcohol used as the
solvent, to give the mixture 3¢ which was isolated as crude
oil by removal of the solvent and the organic acid in vac-
uo. The nature of alkoxy group such as methoxy, ethoxy,
isopropoxy, did not cause significant effect on the yields
of the final product 4.11 The reductive dealkoxylation of 3¢
was effectively performed by BH3·THF complex prepared
from BF3·THF (or BF3·Et2O) and NaBH4. In most of our
experiments, excess of Lewis acids and NaBH4 were used
(theoretical molar ratio of BF3 and NaBH4 to 2 is 0.5 and
0.375 respectively when 3¢ is formed quantitatively). But
we could reduce these values.8 The reduction using
NaBH4/AlCl3 also gives the desired N-cyclopropyl-
anilines 4. The reduction by BH3·THF complex gave bet-
Scheme 3
(7) Gillaspy, M. L.; Lefker, B. A.; Hada, W. A.; Hoover, D. J.
Tetrahedron Lett. 1995, 36, 7399.
(8) For example, the molar ratio of BF3·THF and NaBH4 to 2g
could be reduced to 1.05 each without decreasing the yield
of 4g.
(9) When NaBH4/AlCl3 was used in Step 2, the reaction
proceeded smoothly, but 14% of N,N-dicyclopropyl-3,4-
difluoro-2-methoxyaniline was found in the product (entry
9). The reason why such large amount of this by-product was
formed is not clear (normally <1%). But intermolecular
migration of cyclopropyl group seemed have occurred in this
reagent system.
(10) Typical reaction procedure (entry 7) is as follows: Into a 200
mL four-necked flask fitted with a reflux condenser, a
magnetic stirrer and a thermometer were fed 3,4-difluoro-2-
methoxyaniline (2g, 7.96 g, 50 mmol), AcOH (12.0 g, 200
mL) and MeOH (50 mL). After [(1-ethoxycyclo-
9
ter yield and less by-products than that by NaBH4/AlCl3.
These results are summarized in Table 1.10
propyl)oxy]-trimethylsilane (1, 10.0 g, 57.4 mmol) was
added dropwise at r.t., the reaction mixture was refluxed at
67–69 °C for 3 h under N2 atmosphere. Then the mixture was
concentrated in vacuo using a rotary evaporator, to obtain
crude oil 3g¢ (11.52 g) which contained 90.1% of N-(1¢-
methoxy)cyclopropyl-3,4-dimethoxyaniline and 2.7% of
N-(1¢-ethoxy)cyclopropyl-3,4-dimethoxyaniline, analyzed
by GC and GC–MS.
Into a 200 mL four-necked flask fitted with a reflux
condenser, a mechanical stirrer and a thermometer were fed
NaBH4 (3.78 g, 100 mmol) and anhyd THF (50 mL). After
cooling to 5 °C and adding BF3·Et2O complex (14.19 g, 100
mmol) dropwise, the mixture was stirred under N2
atmosphere for 1 h at 5 °C. Then, crude 3g¢ dissolved in THF
(25 mL) was added dropwise at 5–10 °C in a time period of
20 min. After stirring at r.t. for 5 h, at reflux temperature for
2 h, and recovering THF (60 mL) by distillation, the mixture
was cooled to r.t. and poured into water (300 mL). Then the
resulting mixture was extracted with Et2O (2 × 100 mL). The
Et2O layer was washed with water (2 × 100 mL) and dried
over anhyd Na2SO4 followed by the removal of Et2O by a
rotary evaporator, to obtain a crude oil (10.56 g). The oil was
subjected to distillation under reduced pressure to give N-
cyclopropyl-3,4-difluoro-2-methoxy-aniline (4g) (8.83g,
89%); bp 71–73 °C/0.53 kPa.
These reactions somewhat depend on the substituents of
aromatic ring. The ortho-substituted anilines gave better
yields and fewer by-products. However in the case of no
substituents at the ortho position, the yield was relatively
low (entries 1 vs 2–4 and entry 4 vs 5, 8). We considered
that stability of the intermediate, 1-alkoxy-1-anilinocy-
clopropane (3¢), during solvent evaporation in acidic me-
dia is increased by existence of the ortho-substituents,
such as, chloro, fluoro, methyl or methoxy group. But the
condensation reaction of ortho-substituted anilines pro-
ceeds relatively slow compared to those with substituents
at other positions. Especially, 2-nitroaniline reacted ex-
tremely slowly because of the strong electron-withdraw-
ing group.12
In summary, we have found a very convenient method for
the synthesis of N-cyclopropylaniline derivatives; the in-
termediates of quinolinecarboxylic acids useful as syn-
thetic antimicrobial agents and this method is applicable
for the laboratory scale to the industrial preparation scale.
This method should find utility in medicinal and synthetic
organic chemistry.
(11) The effect of alcohol used in Step1 was studied in the
following manner: 3,4-Difluoro-2-methoxyaniline(2g) was
employed to react in both EtOH and i-PrOH. The molar ratio
of reagents and volume of alcohol (L/mol) was identical with
that in entry 7. After refluxing in EtOH for 5 h, 87.1% of N-
(1¢-ethoxy) cyclopropyl-3,4-dimethoxyaniline was formed
(determined by GC). Employing the conditions described in
Step 2 (2g:NaBH4:BF3·THF = 1:1.2:1.2) at r.t. for 1 h and at
60 °C for 6 h, N-cyclopropyl-3,4-difluoro-2-methoxy-
aniline (4g) was isolated in 83% yield. On the other hand,
Step 1 in i-PrOH required higher temperature and
substitution of the ethoxy group in 3¢ by i-PrOH proceeded
rather sluggishly. Thus after reaction at 70 °C for 4 h and
refluxing (87 °C) for 4 h, 62.8% of N-(1¢-isopropoxy)-
cyclopropyl-3,4-dimethoxyaniline and 14.5% of N-(1¢-
References
(1) (a) Michael, S.; Uwe, P.; Klaus, G.; Andreas, K. EP 0332033
A2, 1989. (b) Mcguirk, P. R.; Conn, L. US Pat 5039682 A,
1991.
(2) (a) Rühlmann, K. Synthesis 1971, 236. (b) Salaün, J.;
Marguerite, J. Org. Synth. 1985, 63, 147.
(3) Kang, J.; Kim, K. S. J. Chem. Soc., Chem. Commun. 1987,
897.
(4) Miller, S. A.; Gadwood, R. C. Org. Synth. 1989, 67, 210.
(5) Employing AcOH and formic acid seemed to give the same
results, but we used AcOH due to better handling and lower
toxicity.
Synlett 2003, No. 14, 2139–2142 © Thieme Stuttgart · New York