Effective monoallylation of anilines catalyzed by supported KF

Article abstract of DOI:10.1021/ol070890o

Equation Presented A mild and straightforward monoallylation procedure for different anilines is described using the efficient, inexpensive, noncorrosive, and environmentally friendly reagent KF-Celite. By using only a 1/1.2 stoichiometric ratio of electrophilic reagent to aniline, in very short reaction times, the monoallylated products are obtained in high isolated yields via this procedure, which works very effectively regardless of the electronic nature of the substituent on the ring, although electron withdrawing groups make the reactions go even faster.

Full text of DOI:10.1021/ol070890o

ORGANIC
LETTERS
2007
Vol. 9, No. 14
2661-2664
Effective Monoallylation of Anilines
Catalyzed by Supported KF
Vittorio Pace, Fernando Mart´ınez, Mar´ıa Ferna´ndez, Jose V. Sinisterra, and
Andre´s R. Alca´ntara*
Organic and Pharmaceutical Chemistry Department, Pharmacy Faculty,
Complutense UniVersity, 28040-Madrid, Spain
andresr@farm.ucm.es
Received April 17, 2007
ABSTRACT
A mild and straightforward monoallylation procedure for different anilines is described using the efficient, inexpensive, noncorrosive, and
environmentally friendly reagent KF-Celite. By using only a 1/1.2 stoichiometric ratio of electrophilic reagent to aniline, in very short reaction
times, the monoallylated products are obtained in high isolated yields via this procedure, which works very effectively regardless of the
electronic nature of the substituent on the ring, although electron withdrawing groups make the reactions go even faster.
Allylic secondary amines are useful building blocks in
organic chemistry,1 making improved procedures for their
synthesis or new preparative routes an attractive research
field.
In this sense, the transition-metal catalyzed allylation of
anilines with allylic alcohols constitutes an interesting
procedure for the synthesis of N-allylanilines. In this regard,
significant progress has been made using transition-metal
catalysts prepared from palladium2 or titanium.3 On the other
hand, the synthesis of aromatic secondary amines through
classical methodologies4 is generally complicated by the
formation of significant amounts of N-dialkylation products,
which lowers the chemoselectivity and the economy of the
process.5 To overcome this problem, the use of inorganic
catalysts such as zeolites6 has been reported as a methodology
that provides selectively N-allylaniline derivatives, even when
deactivating groups are present in the aromatic moiety.
Nevertheless, an excess of amine (to prevent overallylation)
and extended reaction times are demanded, because of the
low nucleophilicity of the aniline nitrogen.
(4) For a review, see: Salvatore, R. N.; Yoon, C. H.; Jung, K. W.
Tetrahedron 2001, 57, 7785 and references therein.
(5) (a) Fellows, I. M.; Kaelin, D. E.; Martin, S. F. J. Am. Chem. Soc.
2000, 122, 10781. (b) D’Amico, J. J.; Harman, M. W.; Cooper, R. H. J.
Am. Chem. Soc. 1957, 79, 5270. (c) Mori, M.; Chiba, K.; Okita, M.; Kayo,
I.; Ban, Y. Tetrahedron 1985, 41, 375. (d) Pollard, C. B.; Parcell, R. F. J.
Am. Chem. Soc. 1951, 73, 2925. (e) Barluenga, J.; Canteli, R. M.; Flo´rez,
J. J. Org. Chem. 1996, 61, 3753. (f) Capella, L.; Montevecchi, P. C.;
Navacchia, M. L. J. Org. Chem. 1995, 60, 7424. (g) Barluenga, J.; Foubelo,
F.; Fan˜ana´s, F. J.; Yus, M. J. Chem. Soc., Perkin Trans. 1 1989, 553. (h)
Oh, H. K.; Koh, H. J.; Lee, I. J. Chem. Soc., Perkin Trans. 2 1991, 1981.
(i) Oh, H. K.; Shin, C. H.; Lee, I. J. Phys. Org. Chem. 1992, 5, 731. (j)
Lee, I.; Kang, H. K.; Lee, H. W. J. Am. Chem. Soc. 1987, 109, 7472. (k)
Oh, H. K.; Jeong, E.-M.; Lee, I. Bull. Korean Chem. Soc. 1998, 19, 1334.
(l) Nazarov, S. I.; Magerramov, M. N. Russ. J. Org. Chem. 1997, 33, 66.
(m) McCall, I.; Proctor, G. R.; Purdie, L. J. Chem. Soc. C 1970, 1126
(6) Onaka, M.; Umezono, A.; Kawai, M.; Izumi, Y. J. Chem. Soc., Chem.
Commun. 1985, 1202.
(1) For a review, see: Johannsen, M.; Jørgens, K. A. Chem. ReV. 1998,
98, 1689.
(2) (a) Ozawa, F.; Okamoto, H.; Kawagishi, S.; Yamamoto, S.; Minami,
T.; Yoshifuji, M. J. Am. Chem. Soc. 2002, 124, 10968. (b) Yang, S.-C.;
Tsai, Y.-C. Organometallics 2001, 20, 763. (c) Hsu, Y.-C.; Gan, K.-H.;
Yang, S.-C. Chem. Pharm. Bull. 2005, 53, 1266. (d) Yang, S.-C.; Hsu,
Y.-C.; Gan, K.-H. Tetrahedron 2006, 62, 3949. (e) Johns, A. M.;
Utsunomiya, M.; Incarvito, C. D.; Hartwig, J. F. J. Am. Chem. Soc. 2006,
128, 1828. (f) Tsuji, J. Organic Synthesis with Palladium Compounds;
Springer: Heidelberg, Germany, 1980. (g) Yang, S.-C.; Hung, C.-W. J.
Org. Chem. 1999, 64, 5000. (h) Moreno-Man˜as, M.; Morral, L.; Pleixats,
R. J. Org. Chem. 1998, 63, 6160.
(3) Ramanathan, B.; Odom, A. L. J. Am. Chem. Soc. 2006, 128, 9344.
10.1021/ol070890o CCC: $37.00
© 2007 American Chemical Society
Published on Web 06/13/2007

Alkaline fluorides have shown their efficacy for alkylating
both aliphatic and aromatic amines. The capability of a
fluoride anion to catalyze N-alkylation has been related, as
shown in Figure 1, to its ability to form a hydrogen bond
Table 1. N-Allylation of Aniline Using Different Allyl Halides
Figure 1. Mechanism of KF-catalyzed N-alkylation
with the amine proton, therefore enhancing the Lewis acid
behavior of the metal.7,8
In this paper, an efficient improvement of the selective
N-allylation of anilines using KF-Celite in acetonitrile at
reflux is described. The design of the synthesis is represented
in Scheme 1. In a typical experiment,9 the electrophilic
a Isolated yields. b Temperature ) 70 °C.
Unhindered allyl halides such as allyl iodide (1a), 1-bromo-
3-methylbut-2-ene (1b), and 3-iodo-2-chloropropene (1c) led
to similar or higher chemoselectivities compared to previ-
ously reported methods.2a,d More concretely, we have
improved the stoichiometric ratio between electrophilic
reagent and aniline under our mild reaction conditions. In
fact, we use a 1.2/1 ratio, compared to the previously reported
ratios (4/1) to obtain similar yields in alkylation with KF-
Celite.7a Furthermore, Ramanathan and Odom,3 who use
allylic alcohols, reported only a 51% of isolated yield of 2a
employing a 3/1 ratio of aniline versus allylic alcohol.
Scheme 1. Synthesis of N-Allyl Anilines
The results presented in Table 1 show how the allylation
of aniline was obtained with excellent efficiency using
different allyl halides, which gave rise to good isolated yields
of the corresponding N-monoallyl anilines. When using
halides containing electron-withdrawing groups (entries 3 and
4) the reaction rates were somewhat slower. These differ-
ences in reactivity could be related to deactivation of the
allyl moiety promoted by the substituent.6k However, as
indicated by the results in entry 4, the electronic properties
of these functionalities do not affect reaction yields. Fur-
thermore, no traces of any secondary product were detected
by using this methodology, as shown by the results obtained
when 1b was employed as allylating agent (entry 2),
compared to the undesired allylic rearrangement described
in the synthesis of 2b using allylic alcohols.2d
reagent (1.00 mmol) is added to a suspension of the catalyst
(1.25 mmol) and a small excess of aniline (1.20 mmol) in
acetonitrile. The results obtained, showing the isolated yields
of the corresponding compounds, are shown in Table 1.
As can be seen, even though a small excess of aromatic
amine was used, the corresponding N-allylaniline was always
recovered as the predominant or exclusive product. In fact,
N,N-diallylaniline production decreased as the molecular size
(H < Cl < Br < CH3, entries 1, 3, 4, and 5) of the terminal
vinyl substituent increased and nucleophilic displacement on
2,3-dibromopropene (1d) or on 3-chloro-2-methylpropene
(1e) did not afford any of the tertiary aniline products.
(7) (a) Ando, T.; Yamawaki, J. Chem. Lett. 1979, 45. (b) Hayat, S.;
Rahman, A.; Choudhary, M. I.; Khan, K. M.; Schumann, W.; Bayer, E.
Tetrahedron 2001, 57, 9951. (c) Clark, J. H.; Miller, J. M. J. Am. Chem.
Soc. 1977, 99, 498 and references therein
(8) (a) Kruizinga, W. H.; Kollogg, R. M. J. Am. Chem. Soc. 1981, 103,
5183. (b) Yamada, M.; Yahiro, S.; Yamano, T.; Nakatani, Y.; Ourisson, G.
Bull. Soc. Chim. Fr. 1990, 127, 824.
(9) Typical experimental procedure: To a stirred solution of aniline
derivative (1.20 mmol) and KF-Celite (1.25 mmol) in anhydrous acetonitrile
(3.5 mL), the allyl halide (1.00 mmol) was added. Then, the mixture was
continued for stirring at reflux up to completion of the reaction. The reaction
mixture was filtered to remove catalyst and rinsed several times with
dichloromethane, after the filtrate was evaporated under reduced pression.
The product was purified, whenever necessary, by column chromatography
on silica gel using various solvent systems as eluents.
A more detailed study of the influence of the temperature
on the kinetics of the N-allylation (shown in Table 2) has
indicated that, at the acetonitrile reflux temperature (82 °C),
aniline undergoes N-allylation very smoothly to afford the
best results in terms of both reaction time and chemoselec-
tivity.
It must be pointed that the products could also be obtained
in good yield when the reaction was run at 35 °C (entry 6)
or at 60 °C (entry 7); as can be seen, at 82 °C, the KF-
2662
Org. Lett., Vol. 9, No. 14, 2007

Table 2. Effect of Temperature on Allylation of Aniline
Using 1c
Table 4. N-allylation of Substituted Anilines (4a-c)
temp
°C
time
h
yield of 2c
yield of 3c
ratio
2c/3c
entry
%
%
6
7
8
35
60
82
24
14
5
52
68
83
20
18
8
2.6
3.4
10.4
Celite catalysis gives its highest synthetic performance. Thus,
the regioselectivity of the process is temperature dependent.
To study the influence of the solvent, Table 3 shows the
results obtained with polar aprotic solvents such as DMF,
Table 3. Solvent Effect on KF-Celite Catalyzed Allylation of
Aniline with 1c
time temp yield of 2c yield of 3c ratio
entry solvent
h
°C
%
%
2c/3c
9
DMF
DMSO
THF
18
18
24
24
82
82
66
55
70
60
68
70
13
18
14
19
5.4
3.3
4.8
3.7
10
11
12
Me2CO
DMSO, THF, and acetone. Clearly, the solvent does not alter
the chemoselectivity, but the reaction was more sluggish in
these solvents when compared with acetonitrile (entry 8).
We also studied the influence of different substituents in
the aromatic moiety to study the applicability of this
methodology. The obtained products (some of them described
for the first time) are shown in Table 4.
Allylation of anilines substituted with both electron-
donating (entries 13 to 16) and electron-withdrawing groups
(EWGs) (entries 17 to 20) gives excellent yields of the
corresponding N-monoallylanilines as the predominant or
exclusive reaction products with 2-chloro-3-iodopropene 1c,
in really short reaction times, ranging from 2.5 h (entries 18
and 20) to 4 h (entry 17). When electron-withdrawing groups
are introduced (entries 17-20 and 22-25), the best results
in terms of chemoselectivity, yield, and reaction times are
obtained. The observed differences in reactivity promoted
by the different electronic nature of the substituents could
be related to the acidity of the corresponding anilines: thus,
among nitroanilines, m-nitro (entries 18-19) led to higher
yields of the isolated monoallyl anilines (sole reaction
products) compared to the para regioisomer (entry 17), maybe
Org. Lett., Vol. 9, No. 14, 2007
2663

because of the higher basicity of the meta isomer.10 Under
similar experimental conditions the more acidic ortho isomer
did not react after 48 h. Perhaps this is due to both
intramolecular hydrogen-bond formation11 (which would
obstruct the proton removal by the fluoride) and steric
hindrance. A similar pattern was observed with fluoroanilines
(entries 22-24), where, once again, the best yields were
obtained for the meta isomer. Unlike o-nitro aniline (no
reaction at all after 48 h), o-fluoroaniline was N-allylated,
(entry 25); although the reaction was considerably slower
(36 h), probably because the fluoro substituent does not
introduce such a significant steric hindrance into the molecule
unlike the nitro. However, intramolecular NH-H‚‚‚F hy-
drogen bonding in the ortho position is much weaker12 than
the corresponding NH-H‚‚‚O2N so that the depressing effect
of the o-fluoro group on the overall reaction rate is smaller
compared to the o-nitro. Similarly, the reaction with m-
aminobenzonitrile (entry 20) was also very straightforward.
In any case, compared to reported data on the alkylation of
deactivated anilines,4 our methodology leads to much better
yields and selectivity in the monoalkylation process. In fact,
allylation of deactivated anilines to secondary amines has
only been reported in an alumina or zeolite-catalyzed process
by Onaka et al.5 and in some specific regioisomer of
nitroaniline by Hsu et al.2c
and finally by the m-methoxy (run 3). The mesomeric effect
improves the reaction rate of the allylation of anilines bearing
a p-substituted electron donating group; this fact was
observed also in the morpholinoaniline (entry 16). Among
all m-substituted anilines, 3-chloro-4-methoxy was the only
one (entry 21) that afforded a small amount of diallylated
product.
Comparing various 2-substituted anilines, the steric effect
of the substituent, as well as its electronic nature, plays an
important role in the success of N-allylation via our method.
2-Methoxy and 2,4-difluoro (runs 15 and 24) were the only
reactive anilines with a substituent in the hindered ortho
position that reacted, while a substituent with more steric
hindrance such as 2-nitro did not react under the described
experimental conditions.
To conclude, the synthetic protocol described herein allows
the synthesis of almost exclusively secondary allylic anilines
in short reaction times, improving the stoichiometric relation
between amine and allylating agent. Therefore, we encourage
the use of KF-Celite as an efficient, inexpensive, noncor-
rosive, and environmentally friendly reagent for the synthesis
of allylic anilines.
Acknowledgment. This work was partly supported by a
Research Project of the CAM (Comunidad Auto´noma de
Madrid, QO-UCM, ref S-0505/PPQ/0344). One of the
authors (V. Pace) thanks the MEC (Spanish Ministery of
Education and Science) for a Ph.D. grant (MEC,FPU AP-
2005-5112)
On the other hand, the effect of an electron-donating group
(EDG) such as methoxy (entries 13-15) was predictable on
the basis of position/nucleophility correlation; thus, p-
methoxy, the most nucleophilic, reacted in the shortest time
(entry 1), followed by the o-methoxy regioisomer (run 3),
Supporting Information Available: General experimen-
tal procedures, full spectroscopic data for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(10) Smith, J. W.; Patai, S. In The Chemistry of the Amino Group;
Wiley: New York, 1968; p 161.
(11) (a) Dhaneshwar, N. N.; Tavale, S. S.; Pant, L. M. Acta Crystallogr.,
Sect. B: Struct. Sci. 1978, 34, 2507. (b) Panunto, T. W.; Urba´nczyk-
Lipkowska, Z.; Jonhson, R.; Etter, M. C. J. Am. Chem. Soc. 1987, 109,
7786.
(12) Takahashi, Y.; Higuchi, T.; Sekiguchi, O.; Ubukata, M.; Tajima,
S. Rapid Commun. Mass Spectrom. 1999, 13, 393.
OL070890O
2664
Org. Lett., Vol. 9, No. 14, 2007