Copper-mediated direct aryl C-H cyanation with azobisisobutyronitrile via a free-radical pathway

Article abstract of DOI:10.1021/ol401404y

An unprecedented protocol for the copper-mediated direct cyanation of aryl C-H by employing 2,2′-azobisisobutyronitrile (AIBN) as a free radical "CN" source is presented. The protocol not only provides a more efficient pathway for the synthesis of aryl nitriles in terms of the yields and the loading amount of copper salts but also, more importantly, represents a novel strategy for aryl C-H cyanation via a CN free-radical mechanism as compared to the CN anion-participating protocols often reported.

Full text of DOI:10.1021/ol401404y

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Copper-Mediated Direct Aryl CꢀH
Cyanation with Azobisisobutyronitrile
via a Free-Radical Pathway
Hao Xu,† Peng-Tang Liu,‡ Yun-Hui Li,‡ and Fu-She Han*,§
Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, 5625 Renmin
Street, Changchun, Jilin 130022, Changchun University of Science and Technology,
Changchun, Jilin 130022, China, and State Key Laboratory of Fine Chemicals,
Dalian University of Technology, Dalian 116024, China
fshan@ciac.jl.cn
Received May 17, 2013
ABSTRACT
An unprecedented protocol for the copper-mediated direct cyanation of aryl CꢀH by employing 2,20-azobisisobutyronitrile (AIBN) as a free radical
“CN” source is presented. The protocol not only provides a more efficient pathway for the synthesis of aryl nitriles in terms of the yields and the
loading amount of copper salts but also, more importantly, represents a novel strategy for aryl CꢀH cyanation via a CN free-radical mechanism as
compared to the CN anion-participating protocols often reported.
Owing to the great importance of aryl nitriles either as
the key building blocks in natural products and designed
molecules1 or as a versatile latent group for facile conver-
sion into a diverse class of functionalities,2 the develop-
ment of efficient methods for the synthesis of aryl nitriles
constitutes a continuing focus in synthetic organic
chemistry. The transition-metal-catalyzed cyanation of
aryl (pseudo)halides has been comprehensively investi-
gated with various cyanation reagents3 including MCN
(M = Na, K, Cu, Zn, Me3Si, etc.) and a few organic
sources such as acetone cyanohydrin,4 malononitrile,5
benzyl cyanide,6 and formamide.7 In addition, cyanation
of organometallics such as organoboron compounds8 and
Grignard reagents9 has also been intensively investigated.
Recently, with the pioneering work of Yu,10 rapidly
growing interest has been observed in direct cyanation of
(5) Jiang, Z.; Huang, Q.; Chen, S.; Long, L.; Zhou, X. Adv. Synth.
Catal. 2012, 354, 589.
(6) Wen, Q.; Jin, J.; Hu, B.; Lu, P.; Wang, Y. RSC Adv. 2012, 2, 6167.
(7) Sawant, D. N.; Wagh, Y. S.; Tambade, P. J.; Bhatte, K. D.;
Bhanage, B. M. Adv. Synth. Catal. 2011, 353, 781.
(8) For selected recent publications, see: (a) Zhang, Z.; Liebeskind,
L. S. Org. Lett. 2006, 8, 4331. (b) Liskey, C. W.; Liao, X.; Hartwig, J. F.
J. Am. Chem. Soc. 2010, 132, 11389. (c) Anbarasan, P.; Neumann, H.;
Beller, M. Angew. Chem., Int. Ed. 2011, 50, 519. (d) Zhang, G.; Zhang,
L.; Hu, M.; Cheng, J. Adv. Synth. Catal. 2011, 353, 291. (e) Kim, J.; Choi,
J.; Shin, K.; Chang, S. J. Am. Chem. Soc. 2012, 134, 2528. (f) Ye, Y.;
Wang, Y.-H.; Liu, P.-T.; Han, F.-S. Chin. J. Org. Chem. 2013, 31, 27.
(9) For an example, see: Anbarasan, P.; Neumann, H.; Beller, M.
Chem.;Eur. J. 2010, 16, 4725.
† Changchun Institute of Applied Chemistry.
‡ Changchun University of Science and Technology.
§ State Key Laboratory of Fine Chemicals.
(1) (a) Larock, R. C. Comprehensive Organic Transformations; VCH:
Weinheim, 1989; p 819. (b) Kleemann, A.; Engel, J.; Kutscher, B.;
Reichert, D. Pharmaceutical, Substance: Synthesis, Patents, Applica-
tions, 4th ed.; Georg Thieme: Stuttgart, 2001.
(2) Rappoport, Z. The Chemistry of the Cyano Group; Interscience
Publishers: London, 1970.
(3) For selected recent publications, see: (a) Anbarasan, P.; Schareina,
T.; Beller, M. Chem. Soc. Rev. 2011, 40, 4879. (b) Klinkenberg, J. L.;
Hartwig, J. F. J. Am. Chem. Soc. 2012, 134, 5758.
(4) For selected examples, see: (a) Sundermeier, M.; Zapf, A.; Beller,
M. Angew. Chem., Int. Ed. 2003, 42, 1661. (b) Park, E. J.; Lee, S.; Chang,
S. J. Org. Chem. 2010, 75, 2760. (c) Schareina, T.; Zapf, A.; Cott, A.;
Gotta, M.; Beller, M. Adv. Synth. Catal. 2011, 353, 777.
(10) Chen, X.; Hao, X.-S.; Goodhue, C. E.; Yu, J.-Q. J. Am. Chem.
Soc. 2006, 128, 6790.
(11) For selected examples, see: (a) Jia, X.; Yang, D.; Zhang, S.;
Cheng, J. Org. Lett. 2009, 11, 4716. (b) Jia, X.; Yang, D.; Wang, W.;
Luo, F.; Cheng, J. J. Org. Chem. 2009, 74, 9470. (c) Do, H.-Q.; Daugulis,
O. Org. Lett. 2010, 12, 2517. (d) Yan, G.; Kuang, C.; Zhang, Y.; Wang, J.
Org. Lett. 2010, 12, 1052. (e) Liskey, C. W.; Liao, X.; Hartwig, J. F.
J. Am. Chem. Soc. 2010, 132, 11389.
r
10.1021/ol401404y
XXXX American Chemical Society

aryl CꢀH with MCN,11 or organic CN surrogates such
as the combination of ammonium salts and DMF or
DMSO,12 or the independent use of DMF13 and benzyl
nitrile.14 Effective reactions typically required the use of a
noble transition metal catalyst associated with the assis-
tance of stoichiometric amounts of copper salts, although
a few methods have realized the transformation only by
using stoichiometric amounts of copper metals.
Despite these significant advances, a general problem for
the CN anion-participating cyanation is the fast deactiva-
tion of the catalysts owing to the strong bonding affinity
of the CN anion to metals. This resulted in a demand for
a high catalyst loading, and/or a decreased yield of the
products. Deliberations on the shortcomings of the extant
methods led us to initiate a study aimed at achieving the
direct aryl CꢀH cyanation via a radical pathway. To the
bestof ourknowledge, sucha novelstrategyhas never been
investigated and would provide an alternative option for
more efficient synthesis of aryl nitriles. Herein, we present
thedesign, implementation, and mechanistic elucidation of
the new protocol.
reaction. After an exhaustive screening of various combi-
nations of an array of copper salts16 and solvents,17 we
were pleased to find that Cu(OAc)2 or CuOAc coupled
with the use of acetonitrile as solvent was a promising
combination. Thedesired ortho-cyanated product2acould
be obtained in 60% and 51% yields, respectively, under O2
with the presence of 5.0 equiv of AIBN (Table 1, entry 2).
The reaction did not occur in the absence of copper salts or
under N2. Only 12% of 2a was obtained under air. These
results imply that both copper and O2 are of crucial
importance for the cyanation. The control experiment by
replacing AIBN with CuCN did not afford any product,
indicating that CuCN, a possible intermediate formed
in situ from Cu(OAc)2 and AIBN, is not the latent CN
source. In addition, the reaction also did not occur in the
absence of AIBN, implying that MeCN solvent is not the
provider of the CN source.
Table 1. Optimization of the Reaction Conditionsa
Conceptually, we conceived that the reactivity as well as
the affinity property of a CN radical to a metal ion differs
from its anion form. Moreover, to compare with the CN
anion, the CN radical possesses oxidation ability. There-
fore, the formation of the MꢀCN bond would be accom-
panied by an oxidation of the metals to a higher valent
state. Such a high-valent metal is catalytically more
active for CꢀH activation and swiftly undergoes reductive
elimination, thereby, leading to an improved catalytic
efficiency. For the CN radical source, the relatively safe
and readily available 2,20-azobisisobutyronitrile (AIBN) is
assumed to be suitable. Although AIBN has been popu-
larly used as a radical initiator in a myriad of chemical
transformations by extruding a molecule of N2 to form the
2-cyanoprop-2-yl radical A (eq 1), it has never been studied
as a CN source for cyanation. However, with the early
reports15 showing that A could be trapped by oxygen to
form the radical C whose thermolysis might further release
a CN radical, we reasoned that AIBN might be a suitable
CN radical source for cyanation if the reaction conditions
could be properly controlled.
Cu(OAc)2
(equiv)
AIBN
time
(h)
temp
yield
(%)b
entry
(equiv)
(°C)
1
ꢀ
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
3.0
5.0
5.0
5.0
5.0
13
13
13
13
13
13
13
13
13
24
48
48
48
135
135
135
135
135
145
125
115
135
135
135
135
135
n.r.c
60 (51)d
65
2
1.1
1.5
1.8
2.0
1.1
1.1
1.1
1.1
1.1
1.1
0.55
0.30
3
4
75
5
75
6
40
7
39
8
24
9
50
10
11
12
13
80
54
74
11
a Reaction conditions: 1a (0.4 mmol), AIBN (2.0 mmol), additives
(1.1 equiv), Cu(OAc)2 (amount showed in table), MeCN (4.0 mL) under
an O2 atmosphere in a sealed tube. b Isolated yield. c No reaction. d 1.1
equiv of CuOAc was used for the yield in parentheses.
With these preliminary conditions established, we further
optimized the reaction parameters (Table 1). A gradual
increase of Cu(OAc)2 from 1.1 to 2.0 equiv resulted in a
complete conversion of 1a and an increased yield of 2a to
75% (entries 2ꢀ5). However, the formation of a small
amount of dicyanated byproduct made the purification
of the desired monocyanated 2a laborious. Therefore,
1.1 equiv of Cu(OAc)2 was used in our following optimiza-
tion. The suitable reaction temperature is ca. 135 °C. Either
elevating or lowering the temperature led to a decreased
yield (entries 5ꢀ8). The yield was also somewhat decreased
when 3.0 equiv of AIBN was used (entry 9). In addition,
an appropriate reaction time is also crucial. For instance,
by elongating the reaction time from 13 to 24 h, 2a was
To examine the feasibility of the new protocol, we chose
the cyanation of 2-phenylpyridine 1a (Table 1) as a model
(12) (a) Kim, J.; Chang, S. J. Am. Chem. Soc. 2010, 132, 10272. (b)
Ren, X.; Chen, J.; Chen, F.; Chen, J. Chem. Commun. 2011, 47, 6725. (c)
Kim, J.; Choi, J.; Shin, K.; Chang, S. J. Am. Chem. Soc. 2012, 134, 2528.
(13) Ding, S.; Jiao, N. J. Am. Chem. Soc. 2011, 133, 12374.
(14) Jin, J.; Wen, Q.; Lu, P.; Wang, Y. Chem. Commun. 2012, 48,
9933.
(15) (a) Janzen, E. G.; Krygsman, P. H.; Lindsay, D. A.; Haire, D. L.
J. Am. Chem. Soc. 1990, 112, 8279. (b) Sick, H.; Kim, K. B. Korean
Chem. Soc. 1985, 6, 284.
(16) Copper salts being screened: Cu(OAc)2, Cu(OAc), Cu(OTf)2,
CuSO4, CuSO4 5H2O, Cu(NO3)2 3H2O, CuCl2, CuF2, CuO, CuI,
3
3
CuCN, CuBr, CuCl, Cu2O.
(17) Solvents being screened: DMSO, DMF, DME, DCE, dioxane,
toluene, MeCN, PhCN, AcOH, and H2O.
B
Org. Lett., Vol. XX, No. XX, XXXX

obtained in 80% yield (entry 10). In contrast, the yield was
lowered remarkably to 54% when the time was elongated
to 48 h (entry 10 vs 11). Control experiments revealed that
the product 2a may participate in some side reactions such
as hydrodecyanation in the presence of copper. Attempted
suppression of the side reactions by adding various N- or
O-containing additives, which is assumed to weaken the
interaction between copper and 2a through a coordina-
tion effect, proved to be futile. The O-containing additives
such as 1,2-dimethoxyethane, ethylene glycol, HOAc, and
NaOAc exhibited only a slight influence on the yields (data
not shown). Surprisingly, the presence of N-containing
additives such as L-proline, DMEDA, TMEDA, and 2,20-
bipyridine completely suppressed the reaction. Efforts
toward lowering the loading amount of Cu(OAc)2 showed
that the reaction also proceeded smoothly in the presence
of 0.55 equiv of Cu(OAc)2 for 48 h (entry 12). However, the
yield was lowered markedly when the amount of Cu(OAc)2
was further decreased (entry 13).
amount of copper. However, the reaction was sluggish
when a Me group is present at the carbon adjacent to the
N-atom (2l). The poor reactivity results from the steric
effect of the Me group which prevents the coordination of
copper and the pyridyl ring.10,14
We also inspected the efficacy of the protocol for
electron-deficient substrates. According to the reported
methods,12a,14 such substrates are often less reactive in the
direct cynation reaction. We also found that, in compar-
ison with the relatively electron-rich substrates, the reac-
tion of the electron-deficient derivatives was somewhat less
reactive in the presence of 0.55 equiv of Cu(OAc)2. Only
low to moderate yields were observed under the standard
conditions. However, the yield was improved substantially
when Cu(OAc)2 was increased to 1.1 equiv. Under the
slightly modified conditions, various electron-deficient
2-aryl pyridines were demonstrated to be viable substrates.
The cyanation of an array of compounds proceeded un-
eventfully to afford the corresponding products in moder-
ately high to high yields (2mꢀ2s).
Scheme 1. Products for Cu-Mediated CꢀH Cyanation of
Various Substratesa
Scheme 2. Isotopic Labeling Experiments
Having examined the generality of the protocol, we
became interested in exploring the reaction mechanism.
GC-mass analysis revealed that acetone was really formed
in the reaction system (Figure S2).18 Isotopic labeling
experiments under 18O2 (95 atom % 18O) generated the
18O acetone 3a paired with the formation of a minor
amount of unlabeled 3a0 (Scheme 2, conditions a). The
abundance ratio of 3a:3a0 was ca. 2.5:1, indicative of the
fact that ca. 71% of oxygen atom in acetone originates
from O2 (Figure S3). Alternatively, when the reaction was
performed with 18O-labeled Cu(18O2CCH3)2 (95 atom %
18O)19 under unlabeled O2, the 18O-labeled 3a could also be
detected with an abundance ratio of 3a:3a0 = ca. 1:4.8; ca.
17% of 18O acetone was formed (conditions b; Figure S4).
Finally, by carrying out the reaction with Cu(18O2CCH3)2
under 18O2, the component of 18O acetone was increased
remarkably to 7.3:1; ca. 88% of 18O acetone was formed
(conditions c; Figure S5). The overall content of the 18O
acetone equals exactly the sum of that generated from 18O2
(71%) and the 18O2CCH3 anion (17%), respectively.
Thus, the generation of acetone mainly from the combi-
nation of O2 and AIBN is in good agreement with our
proposed working hypothesis in eq 1. However, the
formation of a minoramount of acetone from AcOꢀ raised
a Reaction conditions: 1 (0.4 mmol), AIBN (2.0 mmol), Cu(OAc)2
(0.55 equiv) in MeCN (4.0 mL) under an O2 atmosphere at 135 °C for
48 h; isolated yields. The reaction was carried out with 1.1 equiv
b
of Cu(OAc)2 for 24 h; isolated yields. c The reaction was carried out with
1.1 equiv of Cu(OAc)2 for 48 h; isolated yields.
With the suitable conditions for the direct aryl CꢀH
cyanation using AIBN as the CN source established, we
next examined the scope and limitation of the protocol
under two different conditions as shown in entries 10 and
12 in Table 1. The results are summarized in Scheme 1.
Either in the presence of 0.55 or 1.1 equiv of Cu(OAc)2,
comparably good yields were obtained under both condi-
tions for 2-aryl pyridines whose aryl ring was modified
by weak alkyl (2b and 2c) or strong alkyloxy electron-
donating groups (2dꢀ2g). Moreover, substrates bearing a
substituent at the 5- and 4-position of the pyridyl ring were
also well tolerated (2hꢀ2j). In addition, a fused aromatic
derivative was also a competent substrate (2k). These
extensive results clearly exemplify that AIBN is a powerful
CN source for efficient cyanation of relatively electron-
rich aryl substrates in the presence of a substoichiometric
(18) See Supporting Information for detailed GC-Mass, ESI-MS,
and XPS analyses.
(19) See Supporting Information for the preparation of anhydrous
Cu(18O2CCH3)2.
Org. Lett., Vol. XX, No. XX, XXXX
C

two interesting questions of how is it formed and what
is the relatiohship between this forming process and
the cyanation reaction. Through extensive spectroscopic
studies, we found that acetyl acetone cyanohydrin 4 could
be detected from the reaction mixture by ESI-MS analysis
(eq 2). Correlated with this observation, X-ray photoelec-
tron spectroscopic (XPS) analyses revealed that Cu(OAc)2
was almost completely reduced to the Cu(I) species by
AIBN either in the presence or in the absence of 2-phenyl-
pyridine 1a (Figure S6). On the basis of these results, we
proposed a plausible pathway toexplain the formation of 4
and Cu(I) species. Namely, the redox reaction of radical A
and Cu(II) through single electron transfer (SET) generates
the Cu(I) ion and the cation species D. Then, the highly
reactive species D is readily trapped by AcOꢀ to form 4.
Conversion of 4 into acetone 3 should readily occur in
various ways, e.g., through the nucleophilic attack of
AcOꢀ to 4. However, extensive control experiments clearly
demonstrated that 4 or the deacetylated 5 was not, or at
most just a very minor, supplier of the CN source for the
cyanation (see Scheme S1 in the Supporting Information
for details). Consequently, the real reactive CN must be
generated predominantly from the combination of O2
and AIBN.
On the basis of the comprehensive control experiments
and spectroscopic analyses, we could propose a full
mechanism for the newly developed cyanation protocol
although a detailed pathway awaits further investigation
(Scheme 3b). The real catalytically active copper species
should be Cu(I)L, which generates in situ through the
reduction of Cu(II) by AIBN. Coordination of Cu(I)L with
1a forms the Cu(I) complex E. Then oxidation of E by the
coaction of the CN radical and O2 generates the high-
valent Cu(III) complex G via F through the highly electro-
philic Cu(III) promoted dearomatization and rearomatiza-
tion procedure.21 Subsequently, reductive elimination of G
delivers the product 2a and regenerates Cu(I).
Scheme 3. Isotope Effect (a) and Proposed Mechanism (b) for
the Cu-Mediated Cyanation of 2-Aryl Pyridine with AIBN
Further ESI-MS spectroscopic studies showed that
when the reaction was carried out under N2 in the presence
of Cu(OAc)2, only two signals corresponding to the
Cu(I)[1a]2 and [Cu(I)]2[1a]2[CN] complexes were detected
(Figure S7). This indicates that only simple coordination
between Cu(I) and 1aoccurs. Incontrast, when the reaction
was carried out under O2, several new strong signals cor-
responding to Cu(III)[1a-H]2, Cu(III)[CN][1a][1a-H], and
Cu(III)[CN][1a-H][2a] or Cu(III)[CN][1a][2a-H] complexes
appeared (Figure S8). However, there was no detection of
any possible Cu(II) complexes. The results along with the
XPS data stronglysuggest that the reaction should proceed
via a Cu(I) to Cu(III) cycle, although attempted further
clarification of the Cu(III) species was unsuccessful owing
to its rather reactive nature.20 We also investigated the
isotopic effect through an intramolecular competition re-
In summary, we have presented an unprecedented strat-
egy for the direct aryl CꢀH cyanation via a free-radical
pathway. The new protocol together with the current CN
anion-participating approaches significantly expands the
options for the straightforward synthesis of aryl nitriles.
Mechanistic study indicates that a Cu(I) to Cu(III) cycle
should be rather preferred. The preliminary results would
be valuable for the development of new copper catalysts or
a CN source for more effective cyanation. Finally, with the
extensive experience obtained in this study, a detailed
investigation into the reaction mechanism and the devel-
opment of a more practical cyanation protocol are cur-
rently underway.
Acknowledgment. Financial support from Hundred
Talent Program of CAS and State Key Lab of Fine
Chemicals (KF1201) is acknowledged.
1
action using the deuterated 1a-D (Scheme 3a). H NMR
analysis revealed that the ratio between the undeuterated
2a and the deuterated 2a-D was 1:4. The large isotopic
effect clearly demonstrated that CꢀH cleavage should
be involved in the rate-determining steps. In contrast, no
isotopic effect was observed for an anion participating
Cu-catalyzed CꢀH functionalization.10
Supporting Information Available. Experimental pro-
1
cedures, characterization data, and H and 13C NMR,
GC-Mass, ESI-MS, XPS spectra. This material is avail-
able free of charge via Internet at http://pubs.acs.org.
(21) Phipps, R. J.; Gaunt, M. J. Science 2009, 323, 1593.
(20) For a review on the high-valent copper in catalysis, see: Hickman,
A. J.; Sanford, M. S. Nature 2012, 484, 177.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX