Catalytic methylation of aromatic amines with formic acid as the unique carbon and hydrogen source

Article abstract of DOI:10.1039/c4cc05908e

A novel methodology is presented for the direct methylation of amines, using formic acid as a unique source of carbon and hydrogen. Based on ruthenium(II) catalysts, the formation of the N - CH₃group proceeds via an efficient formylation/transfer hydrogenation pathway.

Full text of DOI:10.1039/c4cc05908e

ChemComm
View Article Online
View Journal
COMMUNICATION
Catalytic methylation of aromatic amines
with formic acid as the unique carbon and
Cite this:DOI: 10.1039/c4cc05908e
hydrogen source†
Received 29th July 2014,
Solene Savourey, Guillaume Lefevre, Jean-Claude Berthet and Thibault Cantat*
`
`
Accepted 23rd September 2014
DOI: 10.1039/c4cc05908e
www.rsc.org/chemcomm
A novel methodology is presented for the direct methylation of
amines, using formic acid as a unique source of carbon and hydrogen.
Based on ruthenium(II) catalysts, the formation of the N–CH3 group
proceeds via an efficient formylation/transfer hydrogenation pathway.
The use of CO2 as a building block for the production of value-added
chemicals has recently attracted interest as it is a cheap and
renewable resource. While CO2 is already used for the industrial
production of urea (Bosch-Meiser process), CO2 conversion to
methylamines has only been developed since 2013, to by-pass the
use of formaldehyde or toxic methylating reagents such as methyl
iodide, dimethyl sulfate or diazomethane.1 The methylation of
amines with CO2 has first been unveiled, in parallel by our group
Scheme 1 Strategies for the methylation of amines with CO2 and HCOOH.
and the Beller group, using hydrosilanes as reductants (Scheme 1).2 methylation of amines with HCOOH remains unknown to date.
Shortly afterwards, Klankermayer et al. and Beller et al. described the The closest example to such a reaction is represented by the recent
hydrogen version of this reaction.3,4 Notably, H2 could be considered utilization of HCOOH as a carbon source for the methylation of
as a renewable reductant, if it is produced by carbon-free (photo)- amines, with hydrosilanes as sacrificial reductants.6,7
electro-reduction of water, and it advantageously circumvents the
RuII complexes are potent hydrogenation catalysts and they have
formation of siloxanes by-products resulting from the oxidation of been successfully utilized in CO2 hydrogenation.3,4 In addition, we
hydrosilanes reductants. Nonetheless, the utilization of H2 comes have recently shown that Ru(COD)(methylallyl)2, associated with
with a kinetic price and the methylation of amines with CO2/H2 still CH3C(CH2PPh2)3(triphos), could efficiently catalyze the dispropor-
requires a high pressure of H2 which results in a low hydrogen yield tionation of HCOOH to methanol in up to 50% yield.8 Because this
and, hence, a low faradaic efficiency.
catalytic system is also able to promote the methylation of amines
From another standpoint, efficient electrocatalysts have been with CO2/H2, we investigated its reactivity in the presence of amines
developed over the past decade to promote the 2-electron reduction and HCOOH. To our delight, we observed that heating a THF
of CO2 to formic acid (HCOOH), in an electrochemical cell, and this solution of aniline 1a with 3 equiv. HCOOH in the presence of
technology is becoming mature.5 In this context, an appealing 1.0 mol% Ru(COD)(methylallyl)2, 1.0 mol% triphos and 1.5 mol%
strategy could emerge by utilizing HCOOH as a unique carbon MSA (methanesulfonic acid) led to the complete consumption of
and hydrogen source for the methylation of amines. This approach HCOOH and 43% conversion of aniline 1a to N-methylaniline 2a
would thus benefit from the low bond dissociation energy (41% yield) and N,N-dimethylaniline 3a (2% yield), after 17 h in a
(BDE) of 91 kcal molÀ1 for the C–H bond in HCOOH (vs. 104 sealed autoclave at 150 1C (entry 2, Table 1). 1H and 13C NMR
and 92 kcal molÀ1 for the H–H and Si–H bonds, respectively), while monitoring of the crude mixture revealed the formation of two side-
producing only H2O and CO2 as by-products. Yet, the direct products, in addition to the expected CO2 and water: methanol
(o5%) which results from the disproportionation of HCOOH,
and H2 which results from its dehydrogenation. Similarly to the
CEA, IRAMIS, NIMBE, CNRS UMR 3299, CEA/Saclay, 91191 Gif-sur-Yvette, France.
ruthenium-catalyzed methylation of amines with CO2/H2,3a the
E-mail: thibault.cantat@cea.fr; Fax: +33 1 6908 6640; Tel: +33 1 6908 4338
presence of an acid promoter, in addition to the ruthenium
precursor and phosphine ligand, is crucial to ensure the
† Electronic supplementary Information (ESI) available: Synthetic procedures
and experimental and DFT data. See DOI: 10.1039/c4cc05908e
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun.

View Article Online
Communication
ChemComm
Table 1 Ruthenium-catalyzed methylation of 1a and 2a with HCOOHa
to the methylation products 2 and 3 ranging from 51 to 88%,
after 24 h at 150 1C with 1.0 mol% [Ru(COD)(methylallyl)2
+
triphos] and 1.5 mol% MSA. Interestingly, the selectivity of
the mono- vs. bis-methylation of primary anilines depends on
the electronic nature of the substituents on the aryl ring.
With strong electronic withdrawing groups (characterized by
a Hammett constant s 4 0.2), the selective formation of 2 is
favored and 3 was obtained in low yield (o9%, for 3e–g and 3j)
(Table 2). The bulky aniline 1c gave 2c in a low 13% yield (entry 3,
Table 2) and the formamide derivative was identified as the major
product in this reaction (traces of the iminium product were also
detected by GC/MS chromatography).9 While the ester group in 1j is
found unaffected, methylation of 1l is accompanied with the
complete reduction of the nitro group to afford 4-aminoaniline in
77% yield and 4-amino-N-methylaniline in 23% yield (entry 12 in
Table 2).7c Additionally, keto, cyano and non-conjugated CQC
groups are not well tolerated in the present methodology, whereas
amide functions are compatible with the methylation of an aromatic
–NH2 group (Table S2, ESI†). Basic amines such as aliphatic amines
were shown to exhibit a lower reactivity in the methylation strategies
utilizing CO2 with PhSiH3 or H2.2,3a,4 This trend is also marked in
the present methylation of amines with HCOOH and, for example,
methylation of benzylamine 1k was found unproductive (entry 11
and Table S2, ESI†). Nonetheless, modest to good yields were also
obtained for the methylation of secondary anilines with HNTf2
Yield (%)
Cat.
Triphos
Conv.
(%) 2a 3a
Entry
R
(mol%) (mol%) Additive
n
1
2
3
4
5
6
H
H
H
H
H
H
H
H
1.0
1.0
2.5
1.0
1.0
1.0
1.0
1.0
1.0
0.8
1.0
1.0
2.5
1.0
1.0
1.0
1.0
1.0
1.0
0.8
—
—
3.0
3.0
3.0
6.0
6.0
36
43
40
88
79
70
69
88
2
41
40
71
19
23
23
61
o1
MSA
MSA
MSA
HNTf2
2
o1
17
40
47
46
22
HNTf2 b 9.0
7
HNTf2
MSA
HNTf2
HNTf2
MSA
6.0
6.0
6.0
8c
9
Me
Me
85 o1
85
10d
11
12
6.0 499 o1 499
H or Me 1.0
H or Me
6.0 499 o1
6.0 499 o1
o1
o1
—
—
—
a
Reaction conditions: substrate (8.3 mmol), Ru(COD)(methylallyl)2,
triphos, formic acid (n equiv.), additive (1.5 mol%), 150 1C, 17 h. Yield
determined by GC/MS using hexamethylbenzene as an internal stan-
b
c
dard, after calibration. HNTf2 (3.0 mol%). Reaction carried out at
d
80 1C. Substrate 2a (0.4 mmol) in a sapphire tube, yield determined by
1H NMR spectroscopy with mesitylene as an internal standard.
catalytic activity and, in the absence of MSA, only 2% 2a were (entries 13 and 17–19). Indole 1n gave 3n in a low 2% yield without
observed (entries 1, 11 and 12, Table 1). Increasing the HCOOH hydrogenation of the CQC double bond (entry 16).
loading to 6 equiv. facilitated the formation of N–CH3 groups
Beyond the proof of concept, the methylation of amines with
and 2a and 3a were obtained in 71 and 17% yield, respectively HCOOH still suffers from a limited scope and we therefore inves-
(entry 4), while no improvement was observed with 9 equiv. HCOOH tigated the mechanism of this novel reaction so as to guide the
nor by increasing the catalyst loading from 1 to 2.5 mol% (entries 3 design of future catalysts. Based on the organic species detected in
and 6, Table 1 and Table S1, ESI†). Importantly, while the methyla- solution (formamide and iminium intermediates, methanol and
tion of 1a is efficient at 150 1C, it also proceeds well at 80 1C (entry 8). CO2), a plausible pathway for the methylation of the N–H bond with
Interestingly, the more acidic HNTf2 additive increases the activity of HCOOH involves the formation of a formamide intermediate which
the catalytic system and favors the bis-methylation of aniline 1a is reduced to an iminium species, prior to its reduction to a N–CH3
(entries 4 and 5 in Table 1). With 3.0 mol% HNTf2, the methylation group (Scheme 2). In fact, formylation of 2a is thermally available
of 1a with 6 equiv. HCOOH provided the bismethylated product 3a and formamide 4a was obtained in quantitative yield after 1 h at
in 46% yield and 2a in 23% yield (entry 7, Table 1). As such, 57% of 150 1C.10 Subsequent reduction of formamide 4a afforded 67% of 3a
the C–H bonds in HCOOH are efficiently converted to C–H bonds (Fig. S3, ESI†). A control reaction confirmed that methanol, issued
in the N–CH3 products, while the remaining 43% of the C–H from the disproportionation of HCOOH, is not a methylating agent,
bonds mainly evolved into H2. Consequently, the methylation of since no methylation of 2a was observed with Ru(COD)(methylallyl)2/
the secondary amine 2a is more efficient with HNTf2 (Table 1 and triphos + MSA and methanol after 24 h at 150 1C. Monitoring the
1
Table S1, ESI†). Based on these findings, the efficient methyla- products distribution over time by H NMR spectroscopy revealed
tion of 2a was achieved on a 0.4 mmol scale, in 17 h in a sealed that HCOOH undergoes dehydrogenation at the earlier stages of the
sapphire NMR tube, with 6 equiv. HCOOH and 1.0 mol% methylation of 2a and serves in parallel as a formylation agent to
Ru(COD)(methylallyl)2/triphos + HNTf2 (1.5 mol%), yielding 3a in yield 4a (Fig. S4, ESI†). HCOOH is then fully consumed and
quantitative yield (entry 10, Table 1). This result corresponds to a the quantity of H2 in solution decreases while 3a is produced,
50% faradaic efficiency and to a catalyst turnover number (TON) of suggesting that the reduction of 4a proceeds both via transfer
100 (TOF 5.9 hÀ1). In comparison, similar TONs and TOFs were hydrogenation (from HCOOH) and hydrogenation. Competition
obtained for the methylation of amines with H2 and CO2 with between the methylation of 2a, the dehydrogenation of HCOOH
Ru(COD)(methylallyl)2 + triphos after 24 h at 150 1C, lower faradaic and its disproportionation to MeOH has been investigated using
efficiencies were obtained, ranging from 0.4 (ref. 4) to 28%.3a
DFT calculations, with the simplified CH3C(CH2PMe2)3 ligand in
The methylation of N–H bonds in a variety of amines was place of triphos. A schematic summary of the results is presented in
then carried out to explore the potential of this novel catalytic Scheme 2 and the computed potential energy surface is given in the
transformation (Table 2). Using 6 equiv. HCOOH, the methyla- ESI† (Fig. S5). In the presence of an acid promoter, such as MSA or
tion of primary anilines 1a–j is efficient with cumulative yields HNTf2, protonation of the reactive Ru(triphos)(k1–OCHO)(k2–OCHO)
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2014

View Article Online
ChemComm
Communication
Table 2 Ruthenium-catalyzed methylation of substituted amines with
formic acida
Substrate
Products distribution (%)
Entry
Conversion (%)
1a 100
1
2
2a, 71 (58) 3a, 17 (12)
1b 84
1c 88
2b, 56
2c, 13
3b, 19
3
3c, o1
4
5
1d 86
1e 51
2d, 63
2e, 51
3d, 23
3e, o1
6
7
1f
70
2f, 51
2g, 50
3f, 9
Scheme 2 Computed (DFT) pathways for the methylation of 2a to 3a.
1g 51
1h 80
3g, o1
complex is expected to form 5. The activation energy associated
with the decarboxylation of 5 was computed at 23.3 kcal molÀ1
to yield hydride complex 6. In agreement with our previous
findings on the disproportionation of HCOOH, generation of
the reactive hydride intermediate is the rate determining step,
meaning that the selectivity of the reaction is mostly under
thermodynamic control. 6 is able to promote either the
reduction of formamide 4a (en route to the methylation of 2a)
or a second molecule of HCOOH (leading to the disproportio-
nation pathway). Alternatively, the Ru–H function can be
quenched by the acidic proton of HCOOH to yield H2 and
complete the dehydrogenation of HCOOH. The three divergent
routes present different thermodynamic and kinetic character-
istics. From 6, release of H2 is essentially barrier less. However,
8
2h, 62 (53) 3h, 17 (10)
9
1i
1j
86
65
2i, 57c
2j, 54
3i, 29c
3j, o1
10
11
1k 37
1l 100
2a 85
2k, o1
2l, 23
3k, o1
12
13b
3a, 85 (69)
14b
2m
2n
9
2
3m, 9
the dehydrogenation of HCOOH has
a low exergonicity
(À9.9 kcal molÀ1 and À29.7 kcal molÀ1 for the dehydrogenation
of 3HCOOH) and it is therefore reversible under the applied
conditions. H2 can thus lead to the re-formation of 6 and, in
turn, be utilized for the reduction of 4a. In contrast, conversion
of 6 to the hemiaminal complex 14 (ESI†) requires an activation
energy of 17.2 kcal molÀ1 and it is irreversible, yielding the
methylamine product 3a, with an overall energy balance of
À30.4 kcal molÀ1. This mechanism is thus in agreement with
the experimental results pointing to a convergent reduction of
4a via both transfer hydrogenation from HCOOH and hydro-
genation. Importantly, this mechanism also shows that the
disproportionation of HCOOH to methanol is less favored
than the reduction of 4a as it requires an activation energy of
20.8 kcal molÀ1 for an exergonicity of À26.1 kcal molÀ1. Never-
theless, methanol formation is unproductive in the methylation
of 2a because the energy barrier required to regenerate 6 from
formaldehyde exceeds 24.8 kcal molÀ1 (Fig. S5, ESI†). Finally, it
16b
17b
18b
3n, 2
2o 66
3o, 66 (52)
2p 51
2q 77
3p, 44 (33)
3q, 77c
19b
a
Reaction conditions: substrate (8.3 mmol), Ru(COD)(methylallyl)2
(1.0 mol%), triphos (1.0 mol%), MSA (1.5 mol%), formic acid (6 equiv.),
150 1C, 24 h. Yield determined by GC/MS chromatography using
hexamethylbenzene as an internal standard, after calibration. Isolated
b
yields are given in parenthesis. MSA was replaced with HNTf2 (1.5 mol%).
c
Reaction carried out in a sapphire tube: substrate (0.4 mmol); yield
determined by 1H NMR spectroscopy. Note: unless otherwise noted,
formamide derivatives were observed as the only side-products when
the yields of 2 and 3 don’t add up to the conversion of 1.
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun.

View Article Online
Communication
ChemComm
see: (b) S. Wesselbaum, T. vom Stein, J. Klankermayer and
W. Leitner, Angew. Chem., Int. Ed., 2012, 51, 7499.
4 Y. Li, I. Sorribes, T. Yan, K. Junge and M. Beller, Angew. Chem., Int.
Ed., 2013, 52, 12156.
5 For a review, see H.-R. Jhong, S. Ma and P. J. A. Kenis, Curr. Opin.
Chem. Eng., 2013, 2, 191.
6 I. Sorribes, K. Junge and M. Beller, Chem. – Eur. J., 2014, 20, 7878.
7 For examples of ruthenium catalyzed hydrogenation reactions with
formic acid see: (a) Y. Watanabe, T. Ohta and Y. Tsuji, Chem. Lett.,
1980, 12, 1585; (b) Y. Watanabe, T. Ohta and Y. Tsuji, Bull. Chem.
Soc. Jpn., 1982, 55, 2441; (c) Y. Watanabe, T. Ohta, Y. Tsuji and
T. Hiyoshi, Bull. Chem. Soc. Jpn., 1984, 57, 2440; (d) N. Menashe,
E. Salant and Y. Shvo, J. Organomet. Chem., 1996, 514, 97.
is remarkable that the mechanism of this unprecedented
methylation of amines with HCOOH differs completely from
the classical Eschweiler–Clarke reaction, which relies on the
condensation of an amine substrate onto formaldehyde and
subsequent reduction of the resulting imine with HCOOH.11
The authors gratefully acknowledge support of this work by
CEA, CNRS, CINES (for computer time, project no. c2014086494),
the European Research Council (ERC starting grant agreement
no. 336467) and the PhosAgro/UNESCO/IUPAC program for
Green Chemistry. T.C. thanks the Fondation Louis D.—Institut
`
´
8 S. Savourey, G. Lefevre, J. C. Berthet, P. Thuery, C. Genre and
T. Cantat, Angew. Chem., Int. Ed., 2014, 53, 10466.
¨
de France for its support. Dr Laurent El Kaım (ENSTA) and Dr
Patrick Berthault (CEA) are thanked for the generous loan of an
autoclave and a sapphire NMR tube.
9 For examples of CO2 reduction to imines, see: S. Bontemps,
L. Vendier and C. Sabo-Etienne, J. Am. Chem. Soc., 2014, 136, 4419.
10 (a) L. F. Frieser and J. E. Jones, Org. Synth., 1955, III, 590;
(b) S. H. Jung, J. H. Ahn, S. K. Park and J. K. Choi, Bull. Korean
Chem. Soc., 2002, 23, 149; (c) A. K. Bose, S. N. Ganguly, M. S. Manhas,
A. Guha and E. Pombo-Villars, Tetrahedron Lett., 2006, 47, 4605;
(d) M. Hosseini-Sarvari and H. Sharghi, J. Org. Chem., 2006, 71, 6652;
(e) C. L. Allen and J. M. J. Williams, Chem. Soc. Rev., 2011, 40, 3405;
( f ) B. A. Aleiwi, K. Mitachi and M. Kurosu, Tetrahedron Lett., 2013,
54, 2077.
Notes and references
1 M. B. Smith and J. March, Advanced Organic Chemistry, Wiley-Inter-
science, New York, 5th edn, 2001.
2 (a) O. Jacquet, X. Frogneux, C. D. N. Gomes and T. Cantat, Chem.
Sci., 2013, 4, 2127; (b) Y. Li, X. Fang, K. Junge and M. Beller, Angew.
Chem., Int. Ed., 2013, 52, 9568.
11 H. T. Clarke, H. B. Gillepsie and S. Z. Weisshaus, J. Am. Chem. Soc.,
1933, 55, 4571.
3 (a) K. Beydoun, T. vom Stein, J. Klankermayer and W. Leitner,
Angew. Chem., Int. Ed., 2013, 52, 9554 for hydrogenation of CO2,
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2014