A very simple copper-catalyzed synthesis of anilines by employing aqueous ammonia

Article abstract of DOI:10.1002/anie.200802569

(Figure Presented) Phasing up to matters: The use of aqueous ammonia in a biphasic system with a copper catalyst and diketone supporting ligands allows the catalytic amination of both activated and unactivated aryl and heteroaryl iodides and bromides under very mild conditions (see scheme; R=electron-donating or -withdrawing group; acac=acetylacetonate).

Full text of DOI:10.1002/anie.200802569

Angewandte
Chemie
DOI: 10.1002/anie.200802569
Homogeneous Catalysis (2)
AVery Simple Copper-Catalyzed Synthesis of Anilines by Employing
Aqueous Ammonia**
Ning Xia and Marc Taillefer*
Ammonia is one of the most abundant and least expensive
synthetic inorganic chemicals.[1] It is made industrially from
dinitrogen and dihydrogen under harsh conditions (up to
5508C and 350 atm) by the catalytic Haber–Bosch process.[2]
Production, which requires 1% of the energy consumed
annually by humans, reaches 108 tons per annum.[3,4b] This
scale is equivalent to that of the biological reduction of
dinitrogen to ammonia by nitrogenase enzymes, which
represents the main source of nitrogen for living organisms.[4]
Ammonia is also the common nitrogen source for the
industrial production of fertilizers and organic amino com-
pounds.[1] It is therefore evident that the development of new
processes that use NH3 as a feedstock constitutes a very
ation of NH4Cl or aqueous NH3 was reported. The system is
very efficient for aryl iodides, but the only example involving
more challenging non-activated aryl bromides shows low
reactivity even in the presence of 20% of CuI at 808C.[13]
As part of our studies on the copper-catalyzed arylation of
nucleophiles,[14] we now report that the amination of both
activated and unactivated aryl and heteroaryl bromides and
iodides can be achieved under very mild conditions by using
aqueous ammonia (a convenient and cheap source of NH3) in
a catalytic system in which the less toxic and inexpensive
metal copper is used together with simple diketone ligands.[15]
We initially selected 4-bromobiphenyl as a model sub-
strate for optimization of the reaction conditions. In our
preliminary studies we were pleased to find that, with DMFas
the solvent, this aryl bromide can be successfully coupled with
NH3·H2O at 908C in the presence of catalytic amounts of CuI
and the ligand 2,4-pentadione (1) to afford 4-aminobiphenyl
in a yield of 76% (Table 1, entry 1).
À
important research target. Activation of N H bonds by
transition metals or singlet carbenes has attracted increasing
attention, but remains difficult.[5,6] The other way to activate
NH3 involves the well-known formation of Lewis acid/base
adducts between transition metals and ammonia (Werner-
type complexes). However, catalytic reactions that consume
NH3 are extremely rare.[7] One such reported example of
copper-catalyzed amination of aryl halides suffers from
several drawbacks: reactions are performed under pressure
(3–4 bars) using liquid ammonia and activated aromatic or
heteroaromatic bromides; moreover, the method encounters
It is worth noting that the reaction is carried out at
approximately atmospheric pressure (1.3 bars), which permits
Table 1: Amination of 4-bromobiphenyl using diketone ligands L.[a]
À
problems with selectivity because of competitive C O
arylation of the ethylene glycol solvent.[8] Recently, the
palladium-catalyzed amination of non-activated aryl halides
with ammonia has been reported by Shen and Hartwig[7] as
well as by Surry and Buchwald[9] in which bulky ferrocene or
electron-rich phosphine ligands were used.[10] These methods
provide an interesting route to anilines, which are key
intermediates in aromatic processes[1,11] and can be carried
out without the ammonia surrogates traditionally required
with palladium catalysts.[12] Here too there are drawbacks: the
necessary use of ammonia pressure in one of the two methods
(5–6 bars),[7] as well as the use of strong bases, a toxic and
expensive metal, and sophisticated supporting ligands. During
the submission process, a copper–l-proline catalyzed aryl-
Ligands L
Yield [%][b]
76
Selectivity [%][c]
92
1
2
1
2
24
92
3
4
3
4
7
–
29
91
5
5
35
99
[*] N. Xia, Dr. M. Taillefer
CNRS, UMR 5253
6
7
6
7
79
7
99
–
Institut Charles Gerhardt Montpellier AM2N, ENSCMontpellier
8, rue de l’Ecole Normale, 34296 Montpellier Cedex 5 (France)
Fax: (+33)467-144-319
E-mail: marc.taillefer@enscm.fr
[**] We are grateful to the CNRS and the region of Languedoc Roussillon
for a PhD grant and financial support.
8
8
0
–
Supporting information for this article (general procedures for the
isolation and full characterization of all the aniline derivatives) is
available on the WWW under http://dx.doi.org/10.1002/anie.
200802569.
[a] L (0.6 equiv) and commercial 28% aqueous NH3 (5 equiv) were used.
[b] GC yield determined with 1,3-dimethoxybenzene as an internal
À
standard. [c] Selectivity/C C coupling between L and 4-bromobiphenyl.
Angew. Chem. Int. Ed. 2009, 48, 337 –339
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
337

Communications
the use of common glass reactors. Concerning the selectivity,
no trace of di- or triarylamines—by-products frequently
formed with palladium catalysis—was observed.
to establish a protocol that permitted the amination of 4-
bromobiphenyl with ammonia in excellent yield. Of the
solvents, bases, additives, and copper sources tested (Table 2),
DMF, caesium carbonate, [Cu(acac)2] (10%), and ligand 1
(0.4 equiv) were found to be the best combination (93% yield
of 4-aminobiphenyl; Table 2, entry 18). It is noteworthy that
the key factor for the success of our system is its biphasic
character.[17] On heating, two phases are present and this
seems necessary, as attested to by the absence of reactivity in
either neat water (Table 2, entry 14) or neat DMF saturated
with NH3 (Table 2, entry 15). We can reasonably assume that
the reaction takes place in the organic phase and that the
aqueous phase plays the role of a reservoir for NH3 (which is
only sparingly soluble in DMF) and copper complexes such as
[Cu(NH3)4]2+. These observations are very preliminary, but
work is currently in progress to better understand how this
biphasic catalytic system functions.
Next we explored the scope of this new method. Our
catalytic system efficiently promotes cross-coupling reactions
between aqueous ammonia and aryl iodides with either
electron-donating or electron-withdrawing substituents
(Table 3, entries 1–4) at 908C. The reaction is even possible
at lower temperatures (608C) with activated substrates such
as p-iodocyanobenzene (Table 3, entry 3).[18] A poor yield is
obtained with 2-iodobenzoic ester because of partial hydrol-
ysis of the ester group (Table 3, entry 5). Furthermore, the
yield of the aniline derivative from the highly activated 4-
nitroiodobenzene (63%; Table 3, entry 4) was slightly low-
ered because of disubstitution at the nitrogen atom.
À
However, small amounts of the C C coupling
product (ca. 5%), resulting from arylation of the
diketone, was detected (see above structure).
We then tested other diketone ligands, and
observed that 1 gives yields almost as high as the
electron-rich, but more expensive 2,2,6,6-tetra-
methyl-3,5-heptanedione (6). All other ligands tested
(Table 1, entries 2–5, 7, and 8) gave only poor yields. The
relationship between the structure of the ligand and reactivity
is still not clear. We presume that there is a compromise
between the solubility of the ligand in the water phase and its
electron-donating ability, which may explain the similar yields
observed in the presence of ligands 1 and 6 (the former is
more soluble in water but less electron-donating than the
latter). We presume that the low reactivity observed with
ligands 3 and 5 results from the presence of a substituent in
the middle position of the diketone. Several copper sources
were screened using the same model substrate and supporting
ligand 1.[16] The presence of copper, ligand, and base are all
necessary in this protocol (Table 2, entries 1–3). Different
forms of Cu (oxidation state 0, I, or II) gave quite similar
results (63 to 79% yield of product) in the presence of
0.6 equivalents of 1 (Table 2, entries 4–8), or with 0.4 equiv-
alents of 1 in the case of [Cu(acac)2] (Table 2, entry 10). By a
systematic variation of the reaction parameters we were able
Although the aryl iodides are interesting substrates, we
focused our attention on arylation with aryl bromides, which
are less reactive electrophiles but of much greater interest for
industrial applications. Our procedure is in this case also
compatible with a wide range of substituents. Various aniline
derivatives were obtained with both unreactive and reactive
aryl bromides including PhBr, bromonaphthyl, 3-bromoani-
sole, 4-bromotoluene, 4-bromobiphenyl, 4-bromocyanoben-
zene, and 4-acetobromobenzene. In most cases, good to
excellent yields of the isolated products were obtained when
the reactions were carried out at 908C and, as in the case of
the corresponding iodides, excellent yields were obtained
even at 608C with activated aryl bromides (Table 3, entries 11
and 12). The reaction with the electron-rich 4-bromotoluene
was slower, but the product could be isolated in good yield by
extending the reaction time to 36 h (Table 3, entry 9). In the
case of 1,4-dibromobenzene, 4-bromoaniline was isolated in
only 41% yield because of the formation of disubstituted
product (Table 3, entry 13). We were successful in extending
the procedure to include heterocyclic aromatic bromides (2-
or 3-bromopyridines; Table 3, entries 14 and 15), which are
important pharmaceutical intermediates.[19]
Table 2: Amination of 4-bromobiphenyl using various copper sources
(0.1 equiv) in the presence of supporting ligand 2,4-pentadione (1).[a]
[Cu]
Ligand 1 (equiv)
Solvent
Yield [%][a]
1
2
3
4
5
6
7
8
–
CuI
CuI
CuI
Cu
CuO
Cu(OAc)2
Cu2O
[Cu(acac)2]
[Cu(acac)2]
[Cu(acac)2]
[Cu(acac)2]
[Cu(acac)2]
[Cu(acac)2]
[Cu(acac)2]
[Cu(acac)2]
[Cu(acac)2]
[Cu(acac)2]
0.6
–
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMSO
CH3CN
NMP
H2O
0
0[b]
2[c]
76
0.6
0.6
0.6
0.6
0.6
0.6
–
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
68
79
73
63
23
76
18
34
9
10
11
12
13
14
15
16
17
18
50
0
DMF
DMF
DMF
DMF
6[d]
45[e]
45[f]
93,[g] 20[h]
In conclusion, we have discovered a general, practical,
economical, and efficient method for transforming aryl
iodides and bromides into aniline derivatives in one step.
This biphasic catalytic procedure is unusual, and work is
currently in progress to elucidate exactly how it functions. The
very low operational pressure allows this reaction to be
performed without autoclaves, which are required for work-
ing with liquid ammonia, and this should be an enormous
[a] GC yield determined using 1,3-dimethoxybenzene as an internal
standard. [b] Same result at 1408C. [c] Without base. [d] DMF presatu-
rated with gaseous ammonia. [e] Addition of 0.5 equiv NBu4Br. [f] K2CO3
was used instead of Cs2CO3. [g] Reaction time 24 h. [h] Reaction time
24 h, 3% of [Cu(acac)2]was used. acac=acetylacetanoate.
338
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 337 –339

Angewandte
Chemie
Table 3: Copper/ligand-catalyzed amination of aryl halides.
Keywords: amination · ammonia · anilines · copper ·
homogeneous catalysis
.
[1] a) K. Weissermel, H. J. Arpe, Industry Organic Chemistry,
Wiley-VCH, Weinheim, 1997; b) D. M. Roundhill, Chem. Rev.
1992, 92, 1.
[2] Catalytic Ammonia Synthesis (Ed.: J. R. Jennings), New York,
Plenum, 1991.
[3] P. Avenier, M. Taoufik, A. Lesage, X. Solans-Monfort, A.
Baudouin, A. de Mallmann, L. Veyre, J. M. Basset, O. Eisen-
stein, L. Emsley, E. A. Quadrelli, Science 2007, 317, 1056.
[4] a) A. V. Yandulov, R. R. Schrock, Science 2003, 301, 76; b) R. R.
Schrock, Proc. Natl. Acad. Sci. USA 2006, 103, 17087.
[5] G. D. Frey, V. Lavallo, B. Donnadieu, W. W. Scholler, G.
Bertrand, Science 2007, 316, 439.
ArX
ArNH2
Yield [%][a]
90
1
2
3
4
80
94,93[b] 99[c]
63
[6] Y. Nakajima, H. Kameo, H. Suzuki, Angew. Chem. 2006, 118,
964; Angew. Chem. Int. Ed. 2006, 45, 950.
[7] Q. Shen, J. F. Hartwig, J. Am. Chem. Soc. 2006, 128, 10028.
[8] a) F. Lang, D. Zewge, I. N. Houpis, R. P. Volante, Tetrahedron
Lett. 2001, 42, 3251; b) S. Gaillard, M. K. Elmkaddem, C.
Fischmeister, C. M. Thomas, J.-L. Renaud, Tetrahedron Lett.
2008, 49, 3471.
[9] D. S. Surry, S. L. Buchwald, J. Am. Chem. Soc. 2007, 129, 10354.
[10] M. C. Willis, Angew. Chem. 2007, 119, 3470; Angew. Chem. Int.
Ed. 2007, 46, 3402.
5
6
23
78
PhBr
PhNH2
7
8
88
85
[11] S. A. Lawrence, Amines: Synthesis Properties and Applications,
Cambridge University Press, Cambridge, 2004.
[12] a) D. Y. Lee, J. F. Hartwig, Org. Lett. 2005, 7, 1169; b) X. H.
Huang, K. W. Anderson, D. Zim, L. Jiang, A. Klapars, S. L.
Buchwald, J. Am. Chem. Soc. 2003, 125, 6653; c) J. Barluenga, F.
Aznar, C. Valdes, Angew. Chem. 2004, 116, 347; Angew. Chem.
Int. Ed. 2004, 43, 343; d) G. A. Grasa, M. S. Viciu, J. K. Huang,
S. P. Nolan, J. Org. Chem. 2001, 66, 7729.
9
65, 79[d]
90,98[c]
92, 82[b]
92, 84[b]
41
10
11
12
13
14
[13] J. Kim, S. Chang, Chem. Commun. 2008, 3052.
[14] a) M. Taillefer, N. Xia, Chem. Eur. J. 2008, 14, 6037; b) F.
Monnier, M. Taillefer, Angew. Chem. 2008, 120, 3140 – 3143;
Angew. Chem. Int. Ed. 2008, 47, 3096; c) A. Ouali, J.-F. Spindler,
A. Jutand, M. Taillefer, Adv. Synth. Catal. 2007, 349, 1906; d) A.
Ouali, R. Laurent, A.-M. Caminade, J.-P. Majoral, M. Taillefer, J.
Am. Chem. Soc. 2006, 128, 15990; e) H.-J. Cristau, P. P. Cellier,
J.-F. Spindler, M. Taillefer, Chem. Eur. J. 2004, 10, 5607; f) M.
Taillefer, H.-J. Cristau, P. P. Cellier, J.-F. Spindler, Fr 2833947-
WO 0353225 (Pr. No. Fr 16547), 2001.
84
15
82
[a] At 908C (general case). [b] At 608C. [c] At 908C with ligand 6.
[15] Our results, published in patent form in 2007, appear only now
in the literature due to the time required for patent protec-
tion: Procꢀdꢀ de Synthꢁse dꢂarylamines, M. Taillefer, N. Xia,
Fr 2007 06827 and PCT 2008 051701.
[d] Reaction time, 36 h.
advantage in terms of both cost and safety for scaling-up the
process. The convenience of aqueous ammonia and low cost
of the catalytic copper system make this method readily
adaptable to production on an industrial scale, where safety
and environmental factors are of particular concern. We
believe that the simple procedure outlined here is extremely
competitive with existing palladium-based catalysts and that
it will be rapidly adopted in custom synthesis and industrial
processes.[15]
[16] Although slightly less selective than 6, 2,4-pentadione (1) was
chosen because it is about 200-times less expensive, an important
factor from an industrial perspective.
[17] Under the conditions of Table 2, entry 18, the optimized ratio of
DMF and aqueous ammonia is 20:3 by volume.
[18] Note that at 908C the reaction is fully selective (99%) in the
presence of supporting ligand 6 (Table 3, entries 3 and 10).
[19] An additional finding of our general study is that the demanding
iron-mediated arylation of ammonia with aryl iodides can also
be achieved. Preliminary results show that in a biphasic/[Fe-
(acac)3] (0.2 equiv)/diketone 6 (0.8 equiv) system, the amination
of iodobenzene with NH3 is possible, although low yields of the
aniline derivatives (20–30%) were obtained.[15]
Received: June 2, 2008
Published online: September 30, 2008
Angew. Chem. Int. Ed. 2009, 48, 337 –339
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
339