Noncatalytic mono-N-methylation of aniline in supercritical methanol: The kinetics and acid/base effect

Article abstract of DOI:10.1039/b504050g

Aniline is easily N-methylated in supercritical methanol without catalyst at 350°C and 0.237 g cm^-3 to give mono-N-methylaniline with high selectivity, and the reaction rate is increased by a small amount of base (LiOH, NaOH, KOH, and CH₃ONa), indicating a difference in the reaction mechanism from the ordinary acid-catalyzed one. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b504050g

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Noncatalytic mono-N-methylation of aniline in supercritical methanol:
the kinetics and acid/base effect{
Yoshihiro Takebayashi,*a Yoshinori Morita,b Hideki Sakai,b Masahiko Abe,b Satoshi Yoda,a
Takeshi Furuya,a Tsutomu Sugetaa and Katsuto Otakea
Received (in Cambridge, UK) 22nd March 2005, Accepted 27th June 2005
First published as an Advance Article on the web 7th July 2005
DOI: 10.1039/b504050g
of acid catalyst. We thus try to elucidate the reaction mechanism
by focusing on the difference between the noncatalytic
N-methylation in supercritical methanol and the ordinary acid-
catalyzed one. In this paper, we perform a kinetic analysis for the
supercritical reaction in order to explain the high selectivity of
NMA. We also report an interesting acid/base effect on the
reaction rate: the supercritical N-methylation is accelerated not by
acid but by base. A new reaction mechanism is proposed that can
account for the experimental results.
Aniline is easily N-methylated in supercritical methanol without
catalyst at 350 uC and 0.237
cm23 to give mono-
N-methylaniline with high selectivity, and the reaction rate is
increased by a small amount of base (LiOH, NaOH, KOH,
and CH3ONa), indicating a difference in the reaction mechan-
ism from the ordinary acid-catalyzed one.
g
N-methylanilines are useful chemical intermediates for dyes and
pharmaceuticals, and are industrially synthesized by a vapor-phase
methylation of aniline with methanol.1,2 In the conventional
method, however, strong acid catalyst is indispensable for the
production of methyl cation as an electrophile, and the selectivity
of mono-N-methylaniline (NMA) is significantly lowered by the
side-reactions, a further N-methylation to N,N-dimethylaniline
(NNDMA) and a C-methylation to toluidines, as shown in
Scheme 1.
Methanol solution of aniline was prepared at a concentration of
50 mM (M ; mol dm23). The solution (3.18 cm3) was packed into
a SUS-316 stainless batch reactor with an inner volume of
10.6 cm3. The air in the reactor was purged with argon. The
sample was heated in a molten salt bath (Thomas Kagaku,
Celsius-600H) at 350 uC. At the supercritical condition, the density
of the sample solution is 0.237 g cm23, and the pressure is
estimated to be 22.8 MPa from the PVT data of pure methanol.4
After a given reaction time (20 min to 6.0 h), the sample was
quickly cooled in a water bath. The yields and selectivities of the
products were determined by GC-MS and GC-FID (Hewlett
Packard, HP-6890). The selectivity of NMA is defined as:
(selectivity of NMA) 5 (yield of NMA)/(total yield of NMA
and NNDMA) 6 100 (%).
In supercritical methanol (TC 5 239 uC, PC 5 8.1 MPa, and
rC 5 0.272 g cm23), by contrast, aniline is readily N-methylated
without catalyst to give NMA with high selectivity.3 The
supercritical N-methylation can provide a simple and clean process
by eliminating the problems associated with the use of acid catalyst
and the separation of the byproducts. It has been rarely explored,
however, why the supercritical methylation occurs in the absence
Fig. 1 shows the reaction time dependence of the concentrations
of aniline, NMA, and NNDMA in the supercritical N-methylation
without catalyst. Here we note that no toluidines were formed in
this reaction. The concentration of aniline monotonically decreases
with reaction time down to 1 mM at 4 h, while the yield of NMA
shows a rapid increase with reaction time up to 78% at 2 h. Since
Scheme 1 Methylation of aniline and the rate constants involved.
aNanotechnology Research Institute, National Institute of Advanced
Industrial Science and Technology (AIST), 1-1-1, Higashi, Tsukuba,
Ibaraki, 305-8565, Japan. E-mail: chikulin@ni.aist.go.jp;
Fax: (+81) 298 61 4726; Tel: (+81) 298 61 9272
bGraduate School of Science and Technology, Tokyo University of
Science, 2641, Yamazaki, Noda, Chiba, 278-8510, Japan
{ Electronic supplementary information (ESI) available: experimental
data in detail for (1) acid/base effect, (2) concentration effect of base, (3)
salt effect, (4) density effect of methanol, and (5) temperature effect. See
http://dx.doi.org/10.1039/b504050g
Fig. 1 Reaction time dependence of the concentrations of aniline ($),
N-methylaniline (&), and N,N-dimethylaniline (r) in supercritical
methanol without catalyst at 350 uC and 0.237 g cm23
.
This journal is ß The Royal Society of Chemistry 2005
Chem. Commun., 2005, 3965–3967 | 3965

View Article Online
the yield of NNDMA at 2 h is only 3%, the selectivity of NMA at
the reaction time is higher than 95%. At reaction times longer than
4 h, however, the yield of NMA shows a slow decrease because of
the further N-methylation to NNDMA, resulting in a decrease in
the NMA selectivity with reaction time.
(see ESI for more and detailed data). Surprisingly, the rate
constant k1 is increased by base,13 whereas it is unaffected or rather
reduced by acid, indicating a difference in the reaction mechanism
from the ordinary acid-catalyzed one. The selectivity of NMA is
also enhanced by base, and no toluidines are produced in the base-
catalyzed reaction. A small amount of base, therefore, can improve
the highly selective mono-N-methylation of aniline in supercritical
methanol. We further found, as shown in the ESI in detail, that the
higher the concentration of base, the larger the k1, at least up to
4.8 h21 at 2.5 mM with negligible decrease in the NMA selectivity.
In the presence of acid, on the other hand, the selectivity of NMA
is reduced, and toluidines are by-produced, as in conventional
acid-catalyzed reactions.
In Fig. 2, the selectivity of NMA in the supercritical methylation
is compared with those in acid-catalyzed reactions in the
literature.2,5–10 The NMA selectivity is plotted against the
conversion of aniline. The higher the aniline conversion, the lower
the NMA selectivity in general. In the supercritical N-methylation,
however, both the aniline conversion and the NMA selectivity are
sufficiently high when the reaction time is 2 to 4 h, in comparison
with those in the conventional reactions with acid catalyst.11
The high selectivity of NMA indicates that the N-methylation
rate constant of aniline to NMA (k1 in Scheme 1) is much larger
than that of NMA to NNDMA (k2). The pseudo-first-order rate
constants k1 and k2 were determined from the least-squares fitting
of the time profile shown in Fig. 1. The kinetic analysis revealed
that k1 5 1.00 ¡ 0.07 h21 is 28 times as large as k2 5 0.036 ¡
0.06 h21. The k1/k2 ratio in the supercritical reaction (28) is
markedly larger than those in acid-catalyzed reactions (typically
ca. 3, and 10 at the largest, as shown in Fig. 2).12
For explanation of the results above, i.e., (i) noncatalytic
reaction, (ii) no toluidine formation, (iii) high NMA selectivity,
and (iv) acceleration by base instead of acid, we propose a new
reaction mechanism for the supercritical N-methylation, as
described in Fig. 3 in comparison with that for the acid-catalyzed
one. The main difference between the two mechanisms is ‘‘which
of the two elementary processes plays an important role in the
N-methylation: an electrophilic attack of a methyl group or a
dissociation of an N–H bond’’. In the conventional reaction
mechanism, acid catalyst is necessary for the production of methyl
cation as an electrophile.14 The methyl cation attacks not only the
amino nitrogen but also the aromatic ring of aniline, yielding
toluidines as well as NMA. Under the electrophilic substitution
mechanism, NMA is readily N-methylated to NNDMA, because
an electron-donating ability of the methyl group of NMA
promotes a further electrophilic attack onto the amino nitrogen,
leading to the low selectivity of NMA. On the other hand, the
supercritical reaction is triggered by the dissociation of an N–H
bond, where a methanol molecule abstracts the amino proton
from aniline. In the noncatalytic reaction mechanism, the aniline
molecule itself behaves as an acid. The presence of base makes
easier the abstraction of an amino proton; the polarization of the
Nd2–Hd+ bond of aniline and the Od2–Hd+ bond of methanol can
be enhanced by a large negative charge of OH2 through hydrogen
bond formation with their protons. The reaction proceeds locally
around the amino group, thus yielding no toluidines. Two
methanol molecules can participate in the reaction to form a six-
membered ring structure, as shown in Fig. 3b, where a proton
transfer occurs between the methanol molecules along their
hydrogen bonding. Similar reaction mechanisms involving a cyclic
transition state are often proposed for N-methylation of aniline on
a zeolite surface15 and for noncatalytic organic reactions in
supercritical water, e.g., a decarboxylation of formic acid and a
We further investigate the acid/base effect on the supercritical
N-methylation. Here we added HCl, H2SO4, LiOH, NaOH,
KOH, and NaOCH3 into the sample solution at a concentration
of 0.5 mM, where a molar ratio of acid/base to aniline is 1 : 100.
The effects on the rate constant k1, the selectivity of NMA at 1 h,
and the total yield of toluidines at 1 h are summarized in Table 1
Fig. 2 Selectivity of N-methylaniline vs. conversion of aniline: the values
for the noncatalytic supercritical N-methylation at various reaction times
($) are compared with those for acid-catalyzed reactions in the literature
(#). The dotted lines are the theoretical curves at k1/k2 5 10 and 3.
Table 1 Acid/base effects (0.5 mM) on the rate constant k1, the
selectivity of N-methylaniline at 1 h, and the yield of toluidines at 1 h
N-methylaniline
selectivity (%)
Toluidines
yield (%)
Acid/base
k1 (h21
)
HCl
1.10 ¡ 0.02
0.85 ¡ 0.01
1.03 ¡ 0.03
2.29 ¡ 0.05
2.23 ¡ 0.16
2.08 ¡ 0.05
2.34 ¡ 0.10
89.8
93.0
96.8
98.7
98.6
97.8
98.2
1.4
1.7
0
0
0
H2SO4
No catalyst
LiOH
NaOH
KOH
NaOCH3
Fig. 3 Comparison of the reaction mechanism between (a) the acid-
catalyzed methylation of aniline and (b) the supercritical N-methylation
without catalyst.
0
0
3966 | Chem. Commun., 2005, 3965–3967
This journal is ß The Royal Society of Chemistry 2005

View Article Online
dehydrogenation of ethanol.16 The small k2 in the supercritical
reaction can be explained by the two reasons: (i) the dissociation of
an amino proton from NMA is hindered by the electron-donating
ability of the methyl group of NMA, and (ii) the methanol
molecule(s) involved in the noncatalytic reaction mechanism has a
steric hindrance larger than that of methyl cation in the acid-
catalyzed one, preventing further methylation of NMA. The
noncatalytic N-methylation is never specific to the methanol
density studied here (see the ESI for the density and temperature
dependence of k1). In the vapor phase, however, the noncatalytic
pathway has a negligible contribution, because methanol density in
the gas phase is much lower than that in the supercritical state.
In conclusion, we propose a new reaction mechanism for the
noncatalytic N-methylation of aniline in supercritical methanol to
explain the high selectivity of NMA and the acid/base effect on the
reaction rate. For a better understanding of the reaction
mechanism, it is essential to investigate both experimentally and
theoretically the three roles of methanol – a solvent, a reactant,
and a proton-transfer catalyst – as well as the acid–base equilibria
in supercritical methanol, which will be studied in our subsequent
work.
2 14303 no kagaku shouhin (in Japanese), The Chemical Daily, Tokyo,
2003, pp. 760–761.
3 (a) T. Sako, Supercritical fluids: Molecular interactions, physical
properties, and new applications, ed. Y. Arai, T. Sako, and
Y. Takebayashi, Springer, Berlin, 2002, pp. 362–364; (b) Y. Horikawa,
Y. Uchino and T. Sako, Chem. Lett., 2003, 32, 232–233.
4 R. D. Goodwin, J. Phys. Chem. Ref. Data, 1987, 16, 799–892.
5 F. Valot, F. Fache, R. Jacquot, M. Spagnol and M. Lemaire,
Tetrahedron Lett., 1999, 40, 3689–3692.
6 K. Sreekumar and S. Sugunan, Appl. Catal., A, 2002, 230, 245–251.
7 B. L. Su and D. Barthomeuf, Appl. Catal., A, 1995, 124, 73–80.
8 Y. K. Park, K. Y. Park and S. I. Woo, Catal. Lett., 1994, 26, 169–180.
9 S. P. Elangovan, C. Kannan, B. Arabindoo and V. Murugesan, Appl.
Catal., A, 1998, 174, 213–219.
10 N. Nagaraju and G. Kuriakose, New J. Chem., 2003, 27, 765–768.
11 Solid acid catalysts, such as metal oxides and zeolites, are often used in
the laboratory (Refs. 5–10), whereas a homogeneous catalyst, PCl3, is
used in a commercial process (Ref. 2). PCl3 is highly toxic and corrosive,
and is methanolysed to produce HCl and CH3Cl.
12 No C-methylation was taken into account in the calculation of the
theoretical curves shown in Fig. 2.
13 A comparative experiment using NaCl and NaBr was performed to
verify that the reaction is accelerated mainly by the basicity of the bases
instead of the countercations (e.g., Na+) involved, as described in the ESI
in detail.
14 G. A. Olah, Friedel–Crafts and related reactions, Vol. 1, Interscience
Publishers, New York, 1963.
15 I. I. Ivanova, E. B. Pomakhina, A. I. Rebrov, M. Hunger,
Y. G. Kolyagin and J. Weitkamp, J. Catal., 2001, 203, 375–381.
16 (a) T. Yagasaki, S. Saito and I. Ohmine, J. Chem. Phys., 2002, 117,
7631–7639; (b) H. Takahashi, H. Hashimoto and T. Nitta, J. Chem.
Phys., 2003, 119, 7964–7971 and references cited therein.
Notes and references
1 S. Narayanan and K. Deshpande, Appl. Catal., A, 2000, 199, 1–31 and
references cited therein.
This journal is ß The Royal Society of Chemistry 2005
Chem. Commun., 2005, 3965–3967 | 3967