A novel aromatic alkylation of anilines with cyclic and acyclic ketones under hydrothermal conditions

Article abstract of DOI:10.1016/j.tetlet.2005.07.165

A novel aromatic ring-alkylation was achieved by condensation between aniline-HCl salts and cyclic or acyclic ketones under hydrothermal conditions.

Full text of DOI:10.1016/j.tetlet.2005.07.165

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 6953–6956
A novel aromatic alkylation of anilines with cyclic and
acyclic ketones under hydrothermal conditions
Barun K. Mehta,a Koji Kumamoto,b Kazumichi Yanagisawaa and Hiyoshizo Kotsukib,*
aResearch Laboratory of Hydrothermal Chemistry, Faculty of Science, Kochi University, Akebono-cho, Kochi 780-8520, Japan
bLaboratory of Natural Products Synthesis, Faculty of Science, Kochi University, Akebono-cho,
Kochi 780-8520, Japan
Received 24June 2005; revised 11 July 2005; accepted 22 July 2005
Abstract—A novel aromatic ring-alkylation was achieved by condensation between anilineÆHCl salts and cyclic or acyclic ketones
under hydrothermal conditions.
ꢀ 2005 Elsevier Ltd. All rights reserved.
Over the past decade, a great deal of attention has been
focused on the development of new synthetic reactions
in aqueous media, mainly due to the increasing impor-
tance of ꢀgreen chemical processesꢁ in modern organic
synthesis.1 In our own efforts in this field, we recently re-
ported a novel synthesis of phenanthridine derivatives
that involves the efficient condensation of o-phenylani-
lineÆHCl with cyclic ketones under hydrothermal condi-
tions (Scheme 1).2 In this work, in addition to the
targeted 6-alkyl-substituted phenanthridine derivatives,
we noted that unexpected byproducts with a cycloalkyl
functionality at the 2-position were formed in the reac-
tion mixture. We considered that the latter compounds
might be accessed through aromatic ring-alkylation
followed by phenanthridine core-construction.3
matic ring-alkylation and to demonstrate the power of
the hydrothermal technique in the development of envi-
ronment-friendly chemical transformations. The results
are summarized in Table 1.5
When a mixture of anilineÆHCl (1a, 2 mmol)6 and cyclo-
hexanone (2a, 7 mmol) in 10 mL of H2O was reacted at
250 ꢁC for 24h, p-cyclohexylaniline (3) was isolated in
29% yield along with tetrahydrophenanthridine deriva-
tives 4 (36%) and 5 (17%) (entry 1).7 It is conceivable that
1a was first condensed with 2a at both the ortho- (prod-
uct A) and para-positions (product D), and, because of
its favorable 6p-electronic system (B!C), the intermedi-
ate ortho-adduct A was converted immediately to 4 by
condensation with another equivalent of 2a (Scheme 2).
Following an identical process, 5 should be derived from
3. Accordingly, the only problematic pathway in this
reaction sequence is the formation of 3 from D. One pos-
sible explanation for this step is that ionic hydrogenation
of the benzylic carbocation intermediate E occurs,8,9
since under hydrothermal conditions the acidic reaction
A survey of the literature revealed that there are no
precedents for this type of alkylation, except for the re-
lated work on the alkylation of pyrrole with ketones in a
hot-water system.4 The purpose of this study was to
clarify the general features of the above mentioned aro-
(
)
n
250-300 ˚C
+
(
)
+
O
n
H2O
NH3+Cl-
N
N
(
)
( )
n
n
Scheme 1.
Keywords: Hydrothermal reactions; Anilines; Aromatic ring-alkylation; Ketones.
*
Corresponding author. Tel.: +81 888 44 8298; fax: +81 888 44 8359; e-mail: kotsuki@cc.kochi-u.ac.jp
0040-4039/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.07.165

6954
B. K. Mehta et al. / Tetrahedron Letters 46 (2005) 6953–6956
Table 1. Reaction of anilineÆHCl salts (1) with ketones (2) under hydrothermal conditions (250 ꢁC, 24h) a
Entry
1
AnilineÆHCl (1)
Ketone (2)
Product(s) (yield, %)b
NH2
H2N
H2N
O
5 (17)
C5H11
4 (36)
2a
3 (29)
6 (38)
9 (21)
12 (3)
N
N
C5H11
N
1a
O
2
1a
7 (19)
C6H13
8 (7)
C6H13
2b
N
H2N
H2N
O
3c
1a
1a
10 (33)
C4H9
11 (14)
2c
N
N
C4H9
C2H5
C2H5
O
4
2d
N
N
14 (trace)
13 (48)
5
6
1a
1a
3-Pentanone (2e)
3-Heptanone (2f)
15 (27)
16 (31)
H2N
H2N
CH(C2H5)2
CH(C3H7)2
H2N
7
8
2a
17 (71)
C5H11
1b
N
R
Me2N
2a
2a
1c
R = NMe2: 18 (19)
R = NHMe: 19 (18)
N
N
20 (9)
21 (10)
NH
2
9
H2N
22 (67)
1d
O
10
11
1d
H2N
23 (48)
2g
O
O
1d
H2N
H2N
OH
2h
24 (11)
25 (6)
Cl
Cl
NH2
Cl
Cl
12
2a
H2N
26 (47)
1e
a All reactions were performed with 1 (2.0 mmol) and 2 (7.0 mmol) in water (10 mL).
b Isolated yields after silica gel column chromatography.
c 2-Cyclopentylidenecyclopentan-1-one was also isolated in 7% yield.
media might be considerably altered, leading to a
superacidic-like environment.10
Under similar conditions, cycloheptanone (2b), cyclo-
pentanone (2c), and cyclobutanone (2d) reacted with

B. K. Mehta et al. / Tetrahedron Letters 46 (2005) 6953–6956
6955
H
O
H+
2a
NH3+Cl-
N
C5H11
NH2
+
N
N
B
E
4
A
C
2a
(-H2O)
1a
H+
NH2
2a
[H]
NH2
NH2
N
C5H11
D
5
3
Scheme 2.
1a to give adducts 6–14, following the same reaction
course as above (entries 2–4). By analogy with our pre-
vious observations,2 the formation of quinolines 13 and
14 can be interpreted as the result of a sequential
mechanistic relay of (i) 6p-electrocyclic ring closure,
(ii) spontaneous [1,2]-migrative ring-expansion, and
(iii) cyclobutane ring-opening (Scheme 3). The present
method was also effective for reactions using acyclic
ketones such as 3-pentanone (2e) and 4-heptanone
(2f), and para-alkylated compounds 15 and 16 were
obtained in respective yields of 27% and 31% (entries 5
and 6).
11), suggesting that the alkylation–aromatization pro-
cess is much faster than the double alkylation–hydroge-
nation process. Finally, the reaction between 2,6-
dichloroaniline (1e) and 2a also proceeded effectively
to afford the desired adduct 26 in 47% yield without loss
of the chlorine atoms (entry 12).
In conclusion, we have described a novel pathway for
the aromatic ring-alkylation of anilines with a variety
of cyclic and acyclic ketones under hydrothermal condi-
tions.12 Although the yields are not always so high, we
believe that the method provides a potential utility of
hydrothermal chemistry in organic synthesis. Further
studies to explore the feasibility of hydrothermal reac-
tions in other types of transformation are now in
progress.
O
-H2O
+
NH3+Cl-
NH2
1a
Acknowledgments
-H2O
O
+
H
This work was supported in part by a Grant-in-Aid for
Scientific Research from the Japan Society for the
Promotion of Science, by a Research Grant from the
Research Institute of Innovative Technology for
the Earth (RITE) (H.K.), and by a Special Research
Grant for Green Science from Kochi University. We
also thank Prof. Y. Fukuyama of Tokushima Bunri
University for MS/HRMS measurements.
13
N
H
N
Scheme 3.
The use of 4-methylanilineÆHCl (1b) as the substrate
proceeded cleanly to provide tetrahydrophenanthridine
17 in reasonably good yield (71%) (entry 7). On the
other hand, N,N-dimethylaniline (1c) was less reactive
than aniline itself: compounds 18–21 were obtained in
a combined yield of 56% and in a ratio of 34:32:16:18
(entry 8). These results imply that the presence of a free
NH function is essential for the desired aromatic ring-
alkylation.11
References and notes
1. Green Chemistry. Designing Chemistry for the Environ-
ment; Anastas, P. T., Williamson, T. C., Eds.; American
Chemical Society: Washington, DC, 1996; Li, C.-J.; Chan,
T.-H. Organic Reactions in Aqueous Media; Wiley: New
York, 1997; Organic Synthesis in Water; Grieco, P. A.,
Ed.; Blackie Academic & Professional: London, 1998;
Green Chemistry: Frontiers in Benign Chemical Syntheses
and Processes; Anastas, P. T., Williamson, T. C., Eds.;
Oxford University Press: New York, 1999; Clean Solvents.
Alternative Media for Chemical Reactions and Processing;
Abraham, M. A., Moens, L., Eds.; American Chemical
Society: Washington, DC, 2002.
2. Mehta, B. K.; Yanagisawa, K.; Shiro, M.; Kotsuki, H.
Org. Lett. 2003, 6, 1605–1608.
3. In a separate experiment, we confirmed that cyclohexa-
none did not react with 6-pentyl-phenanthridine under the
similar conditions.
We also examined the reactions using 2,6-disubstituted
anilines such as 1d and 1e to protect against concomi-
tant ortho-alkenylation. As expected, the reaction be-
tween 1d and 2a gave 22 as a single product in 67%
yield (entry 9). Interestingly, when 4-tert-butylcyclo-
hexanone (2g) was used as a ketone component, the
thermodynamically stable trans-isomer 23 was only iso-
lated in moderate yield (48%) (entry 10). The use of
diketone 2h caused no dialkylation, and instead arylated
compounds 24 and 25 were obtained in low yields (entry

6956
B. K. Mehta et al. / Tetrahedron Letters 46 (2005) 6953–6956
4. Sobral, A. J. F. N.; Rebanda, N. G. C. L.; da Silva, M.;
8. For reviews on hot-water reactions see: Siskin, M.;
Katritzky, A. R. Science 1991, 254, 231–237; Katritzky,
A. R.; Allin, S. M.; Siskin, M. Acc. Chem. Res. 1996, 29,
399–406; Bro¨ll, D.; Kaul, C.; Kra¨mer, A.; Krammer, P.;
Richter, T.; Jung, M.; Vogel, H.; Zehner, P. Angew.
Chem., Int. Ed. 1999, 38, 2998–3014; Savage, P. E. Chem.
Rev. 1999, 99, 603–621; Akiya, N.; Savage, P. E. Chem.
Rev. 2002, 102, 2725–2750.
9. For a similar work on the transformation of o-cyclopent-
enylanilines to p-cyclopentylanilines in hot aniline, see:
Gataullin, R. R.; Kazhanova, T. V.; Fatykhov, A. A.;
Spirikhin, L. V.; Abdrakhmanov, I. B. Russ. Chem. Bull.
2000, 49, 174–176.
Lampreia, S. H.; Silva, M. R.; Beja, A. M.; Paixao, J. A.;
Rocha Gonsalves, A. M. dꢁA. Tetrahedron Lett. 2003, 44,
3971–3973.
5. All reactions were conducted in a Teflon autoclave
reaction vessel (for higher temperature reactions a Haste-
loy-C reaction vessel was used) with cone and thread
fittings and an internal volume of 20 mL, designed to
withstand temperatures up to 250 ꢁC.
Typical experimental procedure for the reaction of 1a with
2a (entry 1). A mixture of anilineÆHCl (1a; 260 mg,
2 mmol) and cyclohexanone (2a; 687 mg, 7 mmol) in
10 mL of H2O was placed in an autoclave reaction vessel
and allowed to react at 250 ꢁC for 24h. After basification
with saturated NaHCO3, the crude mixture was extracted
with ethyl acetate. Purification by silica gel column
chromatography gave 3 (102 mg, 29%), 4 (182 mg, 36%),
and 5 (112 mg, 17%).
To confirm the proposed mechanism, compound D was
prepared independently by Suzuki-Miyaura cross-cou-
pling of cyclohexenyl triflate with p-NH(Boc)C6H4B(OH)2
followed by deprotection of the Boc group. When this
substrate, as its HCl salt, in H2O was reacted at 250ꢁC for
24h, the formation of 3 and 4 was observed in addition to
a significant amount of aniline (1a).
Compound 3: Rf = 0.15 (hexane/AcOEt = 9:1); FTIR
(neat) m 3353, 1620, 1516, 1449, 1275 cmꢀ1 1H NMR
;
(400 MHz, CDCl3) d 1.15–1.26 (1H, m), 1.30–1.40 (4H,
m), 1.68–1.75 (1H, m), 1.77–1.84 (4H, m), 2.35–2.42 (1H,
m), 3.53 (2H, br), 6.63, 7.00 (each 2H, dm, J = 8.3 Hz);
13C NMR (100 MHz, CDCl3) d 26.22, 27.01 (·2), 34.75
(·2), 43.69, 115.21 (·2), 127.52 (·2), 138.55, 144.14; MS
m/z 175 (M+).
250 ˚C, 24 h
H2O (10 mL)
NH3+Cl-
D (1 mmol)
Compound 4: Rf = 0.41 (hexane/AcOEt = 9:1); FTIR
(neat) m 1589, 1499, 1458 cmꢀ1 1H NMR (400 MHz,
;
1a (68%)
+ 3 (14%) + 4 (5%) + 5 (trace)
CDCl3) d 0.92 (3H, t, J = 7.2 Hz), 1.34–1.49 (4H, m),
1.72–1.80 (2H, m), 1.87–1.96 (4H, m), 2.85 (2H, br t,
J = 5.6 Hz), 2.91 (2H, m), 3.12 (2H, br t, J = 5.6 Hz), 7.45,
7.59 (each 1H, ddd, J = 8.3, 6.8, 1.0 Hz), 7.89, 7.99 (each
1H, dd, J = 8.3, 1.0 Hz); 13C NMR (100 MHz, CDCl3) d
14.10, 22.12, 22.66, 22.69, 25.64, 26.37, 28.75, 32.24, 36.14,
122.36, 125.32, 126.59, 127.83, 128.38, 129.20, 141.10,
145.59, 162.21; HRMS m/z 253.1854, calcd for C18H23N:
253.1831.
10. For similar reactions using cycloalkanes in the presence of
superacids, see: Koltunov, K. Yu.; Surya Prakash, G. K.;
Rasul, G.; Olah, G. A. J. Org. Chem. 2002, 67, 4330–4336;
Koltunov, K. Yu.; Surya Prakash, G. K.; Rasul, G.; Olah,
G. A. Tetrahedron 2002, 58, 5423–5426, and references
cited therein. In accordance with this postulation no
reaction was observed when aromatic ketones such as
benzophenone was used.
11. The formation of 6-unsubstituted products 20 and 21
indicates that the reaction may proceed via the methylene
imine intermediate F formed by a loss of methane from
initial adducts.
Compound 5: Rf = 0.46 (hexane/AcOEt = 9:1); FTIR
(neat) m 1588, 1499, 1449 cmꢀ1 1H NMR (400 MHz,
;
CDCl3) d 0.91 (3H, t, J = 7.3 Hz), 1.24–1.57 (8H, m), 1.69–
1.81 (4H, m), 1.86–1.96 (8H, m), 2.68 (1H, tt, J = 11.5,
3.2 Hz), 2.84(2H, t, J = 5.8 Hz), 2.89 (2H, m), 3.12 (2H, t,
J = 5.8 Hz), 7.49 (1H, dd, J = 8.6, 2.0 Hz), 7.66 (1H, d,
J = 2.0 Hz), 7.92 (1H, d, J = 8.6 Hz); 13C NMR (100 MHz,
CDCl3) d 14.10, 22.19, 22.67, 22.76, 25.71, 26.22, 26.41,
26.96 (·2), 28.91, 32.23, 34.64 (·2), 36.08, 44.96, 119.15,
126.45, 127.83, 128.18, 128.95, 140.77, 144.41, 145.11,
161.29; HRMS m/z 335.2606, calcd for C24H33N: 335.2613.
6. All anilines were used as their hydrochloric salts. The use
of free anilines under neutral conditions resulted in no
product formation.
R
F: R = H or C6H11
N
12. We briefly examined the applicability of this method to
phenols, anisole, and pyridine under acidic conditions, but
the reactions completely failed. Although further studies
are necessary, at present we can conclude that this method
is limited to anilineÆHCl salts. See also Ref. 3.
7. The use of a large excess of 2a gave only a complex
mixture of products.