A bio-inspired copper catalyst system for practical catalytic cyanation of aryl bromides

Article abstract of DOI:10.1055/s-0028-1083161

A general and environmentally improved protocol for the cyanation of aryl halides with the nontoxic cyanide source potassium hexacyanoferrate(II) {K ₄[Fe(CN)₆]} using copper catalysis and a ligand system based on 1-alkyl-1H-imidazoles is presented. The advantages of this system are a wide substrate range, high selectivity, easy handling, and inexpensive reagents.

Full text of DOI:10.1055/s-0028-1083161

PRACTICAL SYNTHETIC PROCEDURES
3351
A Bio-inspired Copper Catalyst System for Practical Catalytic Cyanation of
Aryl Bromides
C
onvenientCya
h
nation of Ar
o
ylHalides mas Schareina,a Alexander Zapf,a Alain Cotté,b Nikolaus Müller,b Matthias Beller*a
a
Leibniz-Institut für Katalyse e.V. an der Universität Rostock, Albert-Einstein-Strasse 29A, 18059 Rostock, Germany
Saltigo GmbH, Building Q 18–2, 51369 Leverkusen, Germany
b
Fax +49(381)12815000; E-mail: matthias.beller@catalysis.de
Received 4 June 2008; revised 16 July 2008
PSP
No133
Abstract: A general and environmentally improved protocol for the cyanation of aryl halides with the nontoxic cyanide source potassium
hexacyanoferrate(II) {K4[Fe(CN)6]} using copper catalysis and a ligand system based on 1-alkyl-1H-imidazoles is presented. The advan-
tages of this system are a wide substrate range, high selectivity, easy handling, and inexpensive reagents.
Key words: copper(I), cyanation, alkylimidazoles, bromoarenes, nitriles
Br
CN
10% CuI, alkylimidazole
K4[Fe(CN)6]
toluene, 140–180 °C, 16 h
R
R
Scheme 1
ing agent.5 Since then numerous catalysts, based on palla-
dium, nickel, and more recently copper, have been
Introduction
Benzonitriles are of broad synthetic interest as an integral developed for the coupling of aryl halides with cyanide.
part of dyes, herbicides, agrochemicals, pharmaceuticals, Clearly, palladium complexes have so far dominated as
and natural products. In addition, the nitrile group serves catalysts in cyanation reactions because they tolerate a
as precursor for a plethora of other functional groups, such wide variety of functional groups. In addition, they are
as benzoic acid derivatives, benzylamines, benzalde- less sensitive to air and humidity compared to nickel and
hydes, heterocycles, etc.
they are more active than copper catalysts.
Benzonitriles have been prepared in numerous ways. For The general problem with metal-catalyzed cyanations is
more than a century, stoichiometric reactions have pre- the high affinity of cyanide towards the metal center, re-
vailed in laboratory and industry. These methods include sulting in the fast deactivation of the catalytic system by
the Rosenmund–von Braun reaction of aryl halides1 and the formation of stable cyanide complexes. In order to
the diazotization of anilines with subsequent Sandmeyer overcome this problem, typically, solvents are used in
reaction.2 Due to (over)stoichiometric amounts of metal which the standard cyanide sources, like NaCN, KCN and
waste, such processes do not meet today’s criteria of sus- Zn(CN)2,6 have very low solubility.7 We and others have
tainable synthesis. In industry the method of choice is shown that organic (e.g., TMEDA),8 and inorganic (e.g.,
ammoxidation,3 whereby the corresponding toluene de- Zn and Zn salts)9 additives are beneficial for the regener-
rivatives are reacted with oxygen and ammonia at high ation of the active catalyst. Another approach has been the
temperature (300–550 °C) in the presence of heteroge- slow dosage of cyanide, e.g. acetone cyanohydrin10 or tri-
neous fixed-bed catalysts.4 However, lack of functional methylsilyl cyanide,11 which keeps the cyanide concentra-
group tolerance and the harsh reaction conditions make tion low and leads to a higher catalyst productivity. Other
this method less suitable for functionalized benzonitriles. recent developments included microwave activation,12
For such products the transition-metal-catalyzed cyana- and the application of novel catalyst systems.13 However,
tion of aryl halides is probably the most direct and versa- all these developments have drawbacks such as toxicity
tile route. The first palladium-catalyzed cyanation of aryl and the high price of the cyanide source and comparably
halides was introduced in 1973 by Takagi et al., using aryl high catalyst costs (low catalyst productivity).
bromides and iodides with potassium cyanide as cyanat-
In 2004, we described, for the first time, catalytic cyana-
tions
with
potassium
hexacyanoferrate(II)
{K4[Fe(CN)6]},15a which has the advantage of being es-
sentially the least toxic cyanide source conceivable. While
all known other cyanation sources, e.g. KCN [LDLo (oral,
SYNTHESIS 2008, No. 20, pp 3351–3355
x
x.
x
x
.
2
0
0
8
Advanced online publication: 25.09.2008
DOI: 10.1055/s-0028-1083161; Art ID: M02808SS
© Georg Thieme Verlag Stuttgart · New York

3352
T. Schareina et al.
PRACTICAL SYNTHETIC PROCEDURES
human) = 2.86 mg/kg], are highly poisonous, potassium ous types of aryl and heteroaryl bromides.19 Importantly,
hexacyanoferrate(II) is nontoxic [the LD50 of notoriously difficult substrates react in excellent yield and
K4[Fe(CN)6] is lower than that for NaCl!] and it is even selectivity, making the method applicable on a kilogram
used in the food industry for metal precipitation.
scale.
Important for practical applications of potassium hexacy- The idea of using imidazoles to control the stability and
anoferrate(II) is its commercial availability on a ton scale selectivity of copper catalysts developed from the fact that
and its low price; it is usually less expensive than potassi- the most abundant metal-binding amino acid in nature is
um cyanide.14
histidine. In fact, in most metalloenzymes the actual bind-
ing site contains at least one and in some cases up to three
histidine units per metal atom.20 Following this idea
(Scheme 1), in exploratory experiments we used 1-meth-
yl-1H-imidazole as a mimic of histidine due to its reason-
able price and availability.21 To our delight the use of
1-methyl-1H-imidazole as solvent, to ensure sufficient
binding to the metal center, resulted in a good yield (71%)
of the desired product (Table 1, entry 3). Advantageously,
the addition of sodium carbonate, potassium iodide, or
N,N¢-dimethylethylenediamine, which were necessary in
previous cyanation protocols, gave significantly lower
yields (Table 1, entry 2). Further experiments showed that
1-methyl-1H-imidazole can also be used as an additive in
toluene as solvent (Table 1, entry 4). However, without
any 1-methyl-1H-imidazole present no reaction took
Notably, the cyanide ion is strongly bound in the hexacy-
anoferrate(II) anion and substantially more slowly liberat-
ed than from other metal cyanides. The lower basicity and
nucleophilicity of the hexacyanoferrate(II) anion com-
pared to the free cyanide ion may help to prevent side re-
actions.
Our new approach has proven its value in both palla-
dium-15 and copper-catalyzed cyanations,16 and has been
adopted nicely by other groups, e.g., Ozawa,13a Weiss-
man,17a and Gelman17b and co-workers. In two cases, the
comparably slow cyanation reaction was used as the scav-
enging step in interesting domino sequences.18
More recently, we discovered that copper catalysts con-
taining imidazole ligands allow for the cyanation of vari-
Table 1 Copper-Catalyzed Cyanation of 1-Bromo-3,5-bis(trifluoromethyl)benzenea
Br
CN
catalyst
K4[Fe(CN)6]
F3C
Entry
1c,d
2d,e
3e
CF3
F3C
CF3
Ligand
Additive (mol%)
Temp (°C)
140
Conversionb (%) Yieldb (%)
Selectivity (%)
N,N¢-dimethylethylenediamine
N,N¢-dimethylethylenediamine
–
Na2CO3 (20), KI (20)
98
97
73
66
4
15
40
71
65
0
15
41
97
98
0
Na2CO3 (20), KI (20)
140
–
–
–
–
–
–
–
–
–
–
–
–
140
4
1-methyl-1H-imidazole
–
160
5
160
6
1-butyl-1H-imidazole
1-dodecyl-1H-imidazole
1H-imidazole
160
99
97
62
5
98
87
35
3
99
90
56
60
77
44
22
20
0
7
160
8
160
9
(S)-histidine
160
10
11d
12
13
14
pyridine
160
13
16
9
10
7
2,2¢-bipyridine
160
1-methyl-1H-pyrazole
1-methyl-1H-pyrrole
benzo[d]thiazole
160
2
160
10
5
2
160
0
a Reaction conditions: 1-bromo-3,5-bis(trifluoromethyl)benzene (2 mmol), K4[Fe(CN)6] (20 mol%), CuI (10 mol%), ligand (200 mol%), tolu-
ene (2 mL), tetradecane (200 mL, internal GC standard), 16 h.
b Determined by GC.
c NMP as solvent.
d 100 mol% ligand.
e 1-Methyl-1H-imidazole as solvent.
Synthesis 2008, No. 20, 3351–3355 © Thieme Stuttgart · New York

PRACTICAL SYNTHETIC PROCEDURES
Convenient Cyanation of Aryl Halides
3353
place (Table 1, entry 5). Testing a series of commercially yields of the corresponding nitriles. A particularly inter-
available imidazoles and similar ligands revealed that 1- esting feature is apparent when the yields of ortho- and
butyl-1H-imidazole performed best, with an almost quan- para-substituted isomers are compared. In general, the
titative yield (98%; Table 1, entry 6). Histidine and imida- more sterically hindered starting material gave the better
zole itself, as well as other similar nitrogen ligands, gave yield! Examples constitute 1-bromo-2-nitrobenzene and
only low yields of 3,5-bis(trifluoromethyl)benzonitrile 1-bromo-4-nitrobenzene (92% vs 80%, Table 2, entries 1
(Table 1, entries 8–14).
and 32) as well as 4-bromotoluene and 2-bromotoluene
(70 vs 85%, Table 2, entries 8 and 9). Noteworthy, 1-bro-
mo-2,6-dimethylbenzene gave 89% of the corresponding
product. This behavior is in sharp contrast to most known
palladium-catalyzed coupling reactions and might be an
Scope and Limitations
The new protocol has been proven successfully for the cy- indication of a fundamental mechanistic difference be-
anation of diverse aryl and heteroaryl bromides. Selected tween copper- and palladium-catalyzed cyanations. Fol-
results of the scope and limitations are shown in Table 2. lowing this trend 1-bromo-2-isopropylbenzene and the
It should be noted that for certain substrates a proper adap- very
hindered
2-bromo-1,3,5-triisopropylbenzene
tion of the reaction conditions might be crucial for the suc- (Table 2, entries 11 and 12) gave excellent yields (both
cess of the reaction (Table 2, entries 2 and 3). In general, 85%)! The challenging 2-bromo-1,3,5-tri-tert-butylben-
1-butyl-1H-imidazole is the best ligand. Most of the sub- zene gave the desired benzonitrile in a respectable 63%
strates, both electron-poor, for example, 1-bromo-2-ni- yield (Table 2, entry 16), with almost perfect selectivity.
trobenzene (Table 2, entry 1), and electron-rich ones, such At this point, a few comments about the reaction temper-
as 1-bromo-2-methoxybenzene or 4-bromoaniline ature are appropriate: Although the commonly used tem-
(Table 2, entries 19 and 20), gave high yields and very perature (140–160 °C) may seem high, the observed
good selectivity. Notably, substrates with primary amino chemoselectivity is excellent. Hence, typical side reac-
groups, which did not react to the corresponding benzoni- tions of coupling processes like homocoupling and reduc-
triles in the presence of palladium catalysts, readily gave tion, as well as nitrile side reactions like hydrolysis and
the desired products (Table 2, entries 20, 24, and 25). In trimerization occur only to a very low extent.
addition, heterocyclic substrates also perform well
(Table 2, entries 22–28). Hence, pyridines, thiophenes,
thiazoles, indoles, and furans deliver fair to excellent
Table 2 Scope and Limitations of the Copper-Catalyzed Cyanationa
Entry Substrate
Product
Methodb
Conversionc Yieldc
Selectivity
(%)
(%)
100
94
36
100
84
87
97
72
86
90
86
90
85
78
30
64
(%)
92
75
31
92
83
86
78
70
85
89
85
85
77
73
29
63
1
2
1-bromo-2-nitrobenzene
2-nitrobenzonitrile
Ae
Bf
92
80
86
92
99
99
80
98
99
99
98
95
90
94
97
99
4-bromobenzonitrile
terephthalonitrile
3
4-bromobenzonitrile
terephthalonitrile
Af
Ae
Af
Af
Ad,f
Af
Af
Af
Af
Af
Af
Af
Af
Cg
4
ethyl 2-bromobenzoate
ethyl 2-cyanobenzoate
2-(trifluoromethyl)benzonitrile
4-(trifluoromethyl)benzonitrile
phthalonitrile
5
1-bromo-2-(trifluoromethyl)benzene
1-bromo-4-(trifluoromethyl)benzene
1,2-dibromobenzene
6
7
8
4-bromotoluene
4-methylbenzonitrile
9
2-bromotoluene
2-methylbenzonitrile
10
11
12
13
14
15
16
2-bromo-1,3-dimethylbenzene
1-bromo-2-isopropylbenzene
2-bromo-1,3,5-triisopropylbenzene
2-bromobiphenyl
2,6-dimethylbenzonitrile
2-isopropylbenzonitrile
2,4,6-triisopropylbenzonitrile
biphenyl-2-carbonitrile
biphenyl-4-carbonitrile
2,4,6-tri-tert-butylbenzonitrile
2,4,6-tri-tert-butylbenzonitrile
4-bromobiphenyl
2-bromo-1,3,5-tri-tert-butylbenzene
2-bromo-1,3,5-tri-tert-butylbenzene
Synthesis 2008, No. 20, 3351–3355 © Thieme Stuttgart · New York

3354
T. Schareina et al.
PRACTICAL SYNTHETIC PROCEDURES
Table 2 Scope and Limitations of the Copper-Catalyzed Cyanationa (continued)
Entry Substrate
Product
Methodb
Conversionc Yieldc
(%)
Selectivity
(%)
(%)
85
75
75
69
61
82
49
53
93
67
68
58
25
>99
95
80
>99
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
1-bromonaphthalene
1-naphthonitrile
Af
Af
Af
Af
Af
Af
Af
Bf
Be
Be
Af
Be
Af
Be
Af
Af
Be
87
98
100
90
88
97
94
84
96
93
96
96
92
100
99
95
82
99
2-bromo-6-methoxynaphthalene
1-bromo-2-methoxybenzene
4-bromoaniline
6-methoxy-2-naphthonitrile
2-methoxybenzonitrile
4-aminobenzonitrile
75
83
78
N-(4-bromophenyl)acetamide
3-bromopyridine
N-(4-cyanophenyl)acetamide
nicotinonitrile
63
87
3-bromofuran
furan-3-carbonitrile
58
5-bromopyridin-2-amine
3-bromo-5-fluoropyridin-2-amine
2-bromothiophene
6-aminonicotinonitrile
2-amino-5-fluoronicotinonitrile
thiophene-2-carbonitrile
1H-indole-4-carbonitrile
isoquinoline-4-carbonitrile
pyridine-2-carbonitrile
thiazole-2-carbonitrile
pyrimidine-5-carbonitrile
4-nitrobenzonitrile
55
100
70
4-bromo-1H-indole
71
4-bromoisoquinoline
2-chloropyridine
63
25
2-bromothiazole
100
100
98
5-bromopyrimidine
1-bromo-4-nitrobenzene
2-bromopyridine
pyridine-2-carbonitrile
100
a Reaction conditions: substrate (2 mmol), anhydrous K4[Fe(CN)6] (20 mol%), CuI (10 mol%), solvent (2 mL), tetradecane (200 mL, internal
standard for GC), 16 h.
b Method A: 1-butyl-1H-imidazole (200 mol%), toluene; Method B: 1-methyl-1H-imidazole (solvent); Method C: 1-butyl-1H-imidazole
(200 mol%), o-xylene.
c Determined by GC; average of 2 parallel experiments.
d 40 mol% K4[Fe(CN)6].
e 140 °C.
f 160 °C.
g 180 °C.
K4[Fe(CN)6]·3 H2O was ground to a fine powder and dried in vacu-
um (ca. 2 mbar) at 80 °C overnight. This material was stored in a
stoppered bulb and can be used for several months without a notice-
able decrease in activity. NMP was distilled from CaH2 under re-
duced pressure and stored under argon. 1-Methyl-1H-imidazole
was fractionally distilled under reduced pressure and stored under
argon. Toluene and o-xylene (>99.5% quality) were used without
drying after a triple cycle of vacuum and argon. All other materials
were purchased (Aldrich, Fluka, Alfa-Aesar etc.) and used as re-
ceived.
by GC. Conversion and yield were calculated as average of 2 paral-
lel runs. For isolation of the products the mixture was transferred to
a separation funnel with the help of an appropriate nonwater misci-
ble solvent and washed with H2O. In experiments with >10 mmol of
substance often a dark bluish precipitate and emulsions were
formed, which conceal the interface and lengthen the workup. In
these cases, filtration through a sintered glass frit removes both. The
organic phase was dried (Na2SO4). After evaporation of the sol-
vents, the residue was subjected to column chromatography (silica
gel, hexane–EtOAc). All known benzonitriles were identified by
comparison with commercially available materials and literature
data.
Aryl Cyanides; General Procedure
Anhydrous K4[Fe(CN)6] (148 mg, 0.4 mmol), CuI (38 mg,
0.2 mmol), and aryl halide (2 mmol) were placed in an Ace pressure
tube under argon. Depending on the method used, either tetradecane
(200 mL) (internal standard for GC), 1-butyl-1H-imidazole (520 mL,
4 mmol), and solvent (2 mL) (Method A, C) or tetradecane (200 mL)
and 1-methyl-1H-imidazole (2 mL) (Method B) were added and the
mixture was stirred for 1 min. The pressure tube was sealed and
heated for 16 h at the temperature specified in Table 2. After cool-
ing to r.t., CH2Cl2 (3 mL) was added and the mixture was analyzed
Acknowledgment
Financial support for this work from Lanxess (Saltigo GmbH), the
state of Mecklenburg-Vorpommern and the BMBF as well as the
DFG (Leibniz Price) are gratefully acknowledged. The authors
thank Dr. W. Baumann, Dr. C. Fischer, Mrs. S. Buchholz and Mrs.
S. Schareina for analytical support.
Synthesis 2008, No. 20, 3351–3355 © Thieme Stuttgart · New York

PRACTICAL SYNTHETIC PROCEDURES
References
Convenient Cyanation of Aryl Halides
3355
(13) (a) Jensen, R. S.; Gajare, A. S.; Toyota, K.; Yoshifuji, M.;
Ozawa, F. Tetrahedron Lett. 2005, 46, 8645. (b) Veauthier,
J. M.; Carlson, C. N.; Collis, G. E.; Kiplinger, J. L.; John,
K. D. Synthesis 2005, 2683. (c) Chidambaram, R.
Tetrahedron Lett. 2004, 45, 1441.
(1) Lindley, J. Tetrahedron 1984, 40, 1433.
(2) For a recent catalytic variant of the Sandmeyer reaction see:
Beletskaya, I. P.; Sigeev, A. S.; Peregudov, A. S.; Petrovskii,
P. V. J. Organomet. Chem. 2004, 689, 3810.
(3) For recent examples see: (a) Martin, A.; Kalevaru, N. V.;
Lücke, B.; Sans, J. Green Chem. 2002, 4, 481. (b) Liu, Y.;
Zhong, M.; Yu, W.; Ma, Y. L. Synth. Commun. 2005, 35,
2951. (c) Rombi, E.; Ferino, I.; Monaci, R.; Picciau, C.;
Solinas, V.; Buzzoni, R. Appl. Catal., A 2004, 266, 73.
(d) Lücke, B.; Narayana, K. V.; Martin, A.; Jähnisch, K.
Adv. Synth. Catal. 2004, 346, 1407.
(14) The price per mol CN– is 1.6 for K4[Fe(CN)6]·3 H2O (5 kg,
99%, technical grade) and 2.0 for KCN (6 kg, 96%, technical
grade); source Sigma-Aldrich Chemicals Catalogue 2008.
(15) (a) Schareina, T.; Zapf, A.; Beller, M. J. Organomet. Chem.
2004, 689, 4576. (b) Schareina, T.; Zapf, A.; Beller, M.
Chem. Commun. 2004, 1388. (c) Schareina, T.; Zapf, A.;
Mägerlein, W.; Müller, N.; Beller, M. Tetrahedron Lett.
2007, 48, 1087.
(4) (a) Hagedorn, F.; Gelbke, H.-P. In Ullmanns Encyklopädie
der technischen Chemie, 4th ed., Vol. 17; Bartholomé, E.;
Biekert, E.; Hellmann, H.; Ley, H.; Weigert, W. M.; Weise,
E., Eds.; Verlag Chemie: Weinheim, 1979, 333. (b) Ellis,
G. P.; Romney-Alexander, T. M. Chem. Rev. 1987, 87, 779.
(5) Takagi, K.; Okamoto, T.; Sakakibara, Y.; Oka, S. Chem.
Lett. 1973, 471.
(6) Tschaen, D. M.; Desmond, R.; King, A. O.; Fortin, M. C.;
Pipik, B.; King, S.; Verhoeven, T. R. Synth. Commun. 1994,
24, 887.
(7) Sundermeier, M.; Zapf, A.; Beller, M. Eur. J. Inorg. Chem.
2003, 3513.
(8) Sundermeier, M.; Zapf, A.; Beller, M.; Sans, J. Tetrahedron
Lett. 2001, 42, 6707.
(16) (a) Schareina, T.; Zapf, A.; Beller, M. Tetrahedron Lett.
2005, 46, 2585. (b) Schareina, T.; Zapf, A.; Mägerlein, W.;
Müller, N.; Beller, M. Chem.–Eur. J. 2007, 6249.
(17) (a) Weissman, S. A.; Zewge, D.; Chen, C. J. Org. Chem.
2005, 70, 1508. (b) Grossman, O.; Gelman, D. Org. Lett.
2006, 8, 1189.
(18) (a) Pinto, A.; Jia, Y. X.; Neuville, L.; Zhu, J. P. Chem.–Eur.
J. 2007, 13, 961. (b) Mariampillai, B.; Alliot, J.; Li, M.;
Lautens, M. J. Am. Chem. Soc. 2007, 129, 15372.
(19) Schareina, T.; Zapf, A.; Mägerlein, W.; Müller, N.; Beller,
M. Synlett 2007, 555.
(20) (a) Sendra, V.; Cannella, D.; Bersch, B.; Fieschi, F.;
Menage, S.; Lascoux, D.; Coves, J. Biochemistry 2006, 45,
5557. (b) Hernandez-Romero, D.; Sanchez-Amat, A.;
Solano, F. FEBS J. 2006, 273, 257. (c) Klabunde, T.;
Eicken, C.; Sacchettini, J. C.; Krebs, B. Nat. Struct. Biol.
1998, 5, 1084. (d) Brown, K.; Tegoni, M.; Prudencio, M.;
Pereira, A. S.; Besson, S.; Moura, J. J.; Moura, I.; Cambillau,
C. Nat. Struct. Biol. 2000, 7, 191.
(21) For the use of Cu/imidazole in other coupling reactions see:
(a) Schareina, T.; Zapf, A.; Cotté, A.; Müller, N.; Beller, M.
Org. Process. Res. Dev. 2008, 12, 537. (b) Schareina, T.;
Zapf, A.; Cotté, A.; Müller, N.; Beller, M. Tetrahedron Lett.
2008, 49, 1851.
(9) Ramnauth, J.; Bhardwaj, N.; Renton, P.; Rakhit, S.;
Maddaford, S. P. Synlett 2003, 2237.
(10) Sundermeier, M.; Zapf, A.; Beller, M. Angew. Chem. Int. Ed.
2003, 42, 1661; Angew. Chem. 2003, 115, 1700.
(11) Sundermeier, M.; Mutyala, S.; Zapf, A.; Spannenberg, A.;
Beller, M. J. Organomet. Chem. 2003, 684, 50.
(12) Chobanian, H. R.; Fors, B. P.; Lin, L. S. Tetrahedron Lett.
2006, 47, 3303.
Synthesis 2008, No. 20, 3351–3355 © Thieme Stuttgart · New York