N-Heterocyclic Carbene-Pd Polymers as Reusable Precatalysts for Cyanation and Ullmann Homocoupling of Aryl Halides: The Role of Solvent in Product Distribution

Article abstract of DOI:10.1002/cctc.201500383

A main-chain N-heterocyclic carbene (NHC) polymer with N-dodecyl-substituted groups was found to be a highly efficient precatalyst for cyanation of various aryl halides by using nontoxic cyanic source [K₄Fe(CN)₆] under relatively mild reaction conditions. Several aryl iodides, bromides, and activated chlorides were successfully converted to their corresponding benzonitrile derivatives without using additives (such as reducing agents or exogenous organic ligands) in good to excellent yields. Our investigations showed that changing the solvent from DMF to polyethylene glycol (PEG-200) resulted in dramatic decline of cyanation product formation to the benefit of the Ullmann homocoupling reaction. The catalyst could be reused four and five times with only slight loss of activity in cyanation and the Ullmann coupling, respectively.

Full text of DOI:10.1002/cctc.201500383

DOI: 10.1002/cctc.201500383
Full Papers
N-Heterocyclic Carbene–Pd Polymers as Reusable
Precatalysts for Cyanation and Ullmann Homocoupling of
Aryl Halides: The Role of Solvent in Product Distribution
Babak Karimi,* Majid Vafaeezadeh, and Pari Fadavi Akhavan[a]
A main-chain N-heterocyclic carbene (NHC) polymer with N-do-
decyl-substituted groups was found to be a highly efficient
precatalyst for cyanation of various aryl halides by using non-
toxic cyanic source [K4Fe(CN)6] under relatively mild reaction
conditions. Several aryl iodides, bromides, and activated chlor-
ides were successfully converted to their corresponding benzo-
nitrile derivatives without using additives (such as reducing
agents or exogenous organic ligands) in good to excellent
yields. Our investigations showed that changing the solvent
from DMF to polyethylene glycol (PEG-200) resulted in dramat-
ic decline of cyanation product formation to the benefit of the
Ullmann homocoupling reaction. The catalyst could be reused
four and five times with only slight loss of activity in cyanation
and the Ullmann coupling, respectively.
Introduction
Benzonitrile derivatives are very important building blocks be-
cause they can convert to various functional groups such as al-
dehyde, amides, carboxylic acids, and tetrazoles.[1] Moreover,
organonitriles are directly present in various pharmaceutical
products such as Fadrozole, Azoxystrobin, Letrozole and Citalo-
pram or they participate in the synthesis of vital drugs such as
HIV-1 protease inhibitor.[2]
lyzed cyanation was reported in 1973 by Takagi et al. for the
cyanation reaction of aryl bromides and iodides.[7] Since then,
various methods have been reported that employ Pd for the
development of the cyanation reaction.[8]
However, major concerns still exist for the cyanation reaction
and these concerns need to be resolved. One concern is the
use of a highly toxic source of cyanide such as NaCN[9] and
KCN.[10] Although the toxicity issue of the cyanating agent
could be circumvented to some extent by using Zn(CN)2,[11]
which has a lower toxicity than the traditional sources of cya-
nide, there are associated economic issues. To address these
issues, in 2004, Beller and co-workers introduced potassium
hexacyanoferrate(II) [K4Fe(CN)6] as a nontoxic and inexpensive
source of CNÀ.[12,13] This reagent is commercially available in
ton scale and does not need special precautions for its han-
dling and storage.
On an industrial scale and on a laboratory scale these deriva-
tives are usually synthesized by the Rosenmund–van Braun re-
action from aryl halides[3] or by the Sandmeyer reaction.[4] Am-
moxidation of toluene derivatives is another method for the
ton-scale preparation of benzonitrile derivatives.[5] The major
drawbacks of the mentioned methods include the release of
stoichiometric amounts of heavy metal salts, employing elevat-
ed temperatures, low reactivity of aryl bromides and aryl chlor-
ides, and restriction to nonfunctionalized heterocyclic starting
materials.
Perhaps, the major drawback of Pd-catalyzed cyanation is
the deactivation of the metal catalyst by dissolved cyanide
ions in the reaction mixture. Therefore, compared to known
coupling reactions (such as Suzuki, Heck, etc.), the catalyst ac-
tivity and turnover number (TON) for cyanation reactions are
often very low[14] and the reaction proceeds well in the pres-
ence of expensive ligands. In this context, despite the signifi-
cant achievements in cyanation reactions of aryl, as well as
heteroaryl halides, in most cases the employed catalysts are
homogeneous and thus suffer from difficult separation from
the reaction mixture and nonrecyclability. The lack of success-
ful examples of reusable catalysts and also extensive catalyst
deactivation in this reaction is mainly attributed to catalyst poi-
soning by cyanide ions and formation of inactive [Pd(CN)4]2À
and [(CN)3PdAr]2À species during the reaction.[15] Heterogene-
ous supported palladium catalysts may be an alternative for
this reaction. Heterogeneous catalyst systems, which are con-
veniently recycled from the reaction media and have the po-
tential to be integrated into flow reactors, are more attractive
Transition-metal-catalyzed cyanation is an alternative ap-
proach for the synthesis of benzonitrile derivatives. Cu, Ni, and
Pd complexes are among the most popular catalysts for pro-
moting this reaction.[6] Among them, Pd is definitely more re-
active than Cu and tolerates various functional groups. More-
over, Pd is relatively less sensitive in contact with air and hu-
midity and is significantly less toxic than Ni. The first Pd-cata-
[a] Prof. Dr. B. Karimi, Dr. M. Vafaeezadeh, Dr. P. F. Akhavan
Department of Chemistry
Institute for Advanced Studies in Basic Sciences (IASBS)
Zanjan 45137-6731 (Iran)
Fax: (+98)241-415-3232
E-mail: karimi@iasbs.ac.ir
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201500383.
This publication is part of a Special Issue on Palladium Catalysis. To view
the complete issue, visit:
http://onlinelibrary.wiley.com/doi/10.1002/cctc.v7.14/issuetoc
ChemCatChem 2015, 7, 2248 – 2254
2248
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
and hold promise for use in large-scale productions. Therefore,
a substantial volume of recent investigations have been direct-
ed towards the development of new and efficient cyanation
protocols based on heterogeneous Pd catalysts.[16]
case of highly challenging substrates such as aryl chlorides and
heteroaryl bromides.
Results and Discussion
Although significant improvements have been achieved by
using heterogeneous catalysts, the “lazy” kinetics of the cyana-
tion reactions, extensive metal leaching, and limited substrate
scope have remained to be major concerns of these methods
and should be resolved. We respect the pioneering advances
in this area, however, we believe that the development of
a novel heterogeneous catalyst system that exhibits improved
recyclability and durability and displays appropriated per-
formance towards less-reactive aryl halides are still welcome.
Following the impressive discovery of Bielawski and co-
workers on the synthesis of highly stable main-chain N-hetero-
cyclic carbene (NHC)-based organometallic polymers[17] (NHC-
MCOPs), we have recently prepared a set of self-supported
NHC–Pd polymers based on various N-substituted bis(imidazo-
lium) compounds (Scheme 1).[18] These materials demonstrated
The route for the preparation of NHC polymer 3 is illustrated
in Scheme 1 according to our previously reported procedure
with slight modifications.[18b] The NHC precursors with an N-
substituted dodecyl group were synthesized from the reaction
between 1,2,4,5-tetraaminobenzene with freshly distilled
formic acid, followed by the alkylation of the resulting benzo-
bis(imidazole) 1 with 1-bromododecane. The obtained bis(imi-
dazolium) bromides
2 were reacted with stoichiometric
amounts of Pd(OAc)2 in DMSO at 110 8C for 10 h.
Finally, the reaction mixture was cooled to room tempera-
ture then added to distilled water. The precipitated polymer
was then collected by means of vacuum filtration and dried
under vacuum at room temperature. Compounds 1 and 2
1
were characterized by H NMR and 13C NMR spectroscopy (see
the Experimental Section). Gel permeation chromatography
(GPC) showed that the average molecular weight (MW) of or-
ganometallic polymer 3 is approximately 85000 Da (Scheme 1).
The efficiency of 3 was initially investigated for the cyana-
tion of bromobenzene as a model substrate with [K4Fe(CN)6].
The data from the optimization reactions is shown in Table 1.
Initially, the effect of the solvent was investigated in the reac-
tion of bromobenzene with 0.6 mol% of catalyst 3, 2 mmol of
Table 1. Screening of the NHC–Pd polymer (3)-catalyzed cyanation of
bromobenzene using [K4Fe(CN)6].
Scheme 1. Schematic procedure for the preparation of the NHC–Pd polymer
catalyst: a) HCOOH, 110 8C, 36 h. b) (i) NaH (2.0 equiv.), PhCH3, 1108C;
(ii) C12H25Br (6.0 equiv.), 25–1108C, 4 h; (iii) DMF, 1108C, 6 h; (iv) dioxane,
110 8C, then 758C. c) Pd(OAc)2 in DMSO at 110 8C, 10 h.
Entry
Solvent
Base
t [h]/T [8C]
Yield [%][a,b]
1
2
3
4
5
6
7
8
H2O
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
Et3N
12/95
12/95
12/95
12/95
24/95
12/95
12/95
12/95
12/95
12/110
12/RT
24/60
25/80
12/95
12/95
12/95
12/95
12/95
12/95
–
–
15
29
35
<5
–
91
70[c]
86 (14)[d]
–
–
15
26[e]
–
–
–
15[f]
48[g]
H2O+1 equiv. TBAB
H2O/tbutanol (1:1)
H2O/DMF (1:1)
H2O/DMF (1:1)
H2O/ethanol (1:1)
DMSO
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
outstanding catalytic activity in the Suzuki–Miyaura coupling
reactions of aryl chlorides in water. In a separate in-depth
mechanistic investigation it was also evident that the high cat-
alytic performance of these NHC-Pd-MCOPs could likely be at-
tributed to the in situ generation of an extremely low concen-
tration of Pd nanoparticles within the framework of the poly-
mers, which not only resulted in the generation of a highly
active region containing trace amounts of Pd nanoparticles,
but also prevented the agglomeration of these Pd nanoclus-
ters.[18b]
9
10
11
12
13
14
15
16
17
18
19
Na2CO3
CS2CO3
K2CO3
K2CO3
Herein, we wish to disclose that a main-chain NHC–Pd poly-
mer bearing an N-dodecyl substituent is also an efficient and
highly recyclable catalyst system for the cyanation of various
type of aryl halides with [K4Fe(CN)6] in DMF. Most interestingly,
we detected that by changing the reaction solvent from DMF
to polyethylene glycol (MWAV ꢀ200), the selectivity patterns of
this reaction were remarkably shifted towards highly efficient
formation of Ullmann homocoupling products, even in the
[a] Reaction conditions: Catalyst (0.6 mol%), bromobenzene (1 mmol),
[K4Fe(CN)6] (0.7 mmol), base (2 mmol), and solvent (2 mL) under Ar atmos-
phere unless otherwise stated. [b] Yields refer to GC (gas chromatogra-
phy) results. [c] 0.5 mmol of [K4Fe(CN)6] was used. [d] 14% of homocou-
pling byproduct was observed. [e] 1 mmol of K2CO3 was used. [f] The
amount of catalyst was 0.2 mol%. [g] The amount of catalyst was
0.4 mol%.
ChemCatChem 2015, 7, 2248 – 2254
www.chemcatchem.org
2249
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
K2CO3, and 0.7 equivalents of [K4Fe(CN)6] as the cyanide source
at 958C. The results indicate that the best yields of benzonitrile
were obtained in DMF rather than other organic or aqueous/
organic solvent mixtures (Table 1, entries 1–8).
Table 2. Cyanation of aryl halides with [K4Fe(CN)6] using NHC–Pd polymer
3 under the optimized reaction conditions.
It has been reported that [K4Fe(CN)6] is able to transfer all of
the six CNÀ ions during the cyanation reaction.[12] However, our
investigations showed that the optimum amount of
[K4Fe(CN)6] is 0.7 mmol per one mmol of bromobenzene and
reducing this value to 0.5 mmol significantly reduces the yield
of benzonitrile to 70% (Table 1, entry 9). Increasing the reac-
tion temperature to 110 8C had a negative effect on the reac-
tion selectivity. Under these conditions, a considerable amount
of the competitive Ullmann homocoupling byproduct (ꢀ14%)
was observed (Table 1, entry 10). Notably, the reaction did not
proceed well at lower temperature even after 24 h (Table 1, en-
tries 11–13). To get better insight into the optimal reaction
conditions, the impact of various bases, as well as the amount
of base, was also studied because these two parameters are
known to be crucial in coupling reactions. Our investigations
revealed that the highest yield was obtained when the cyana-
tion reaction was carried out using K2CO3 as a base (Table 1,
entry 8 vs. 15–17). In the presence of K2CO3, a further decrease
in the amount of base to 1 mmol significantly decreased the
conversion of bromobenzene, giving a disappointing yield
(ꢀ26%) of the cyanation product after 12 h (Table 1, entry 14).
In addition, decreasing the catalyst loading to 0.4 and
0.2 mol% afforded the desired product in much lower yields of
48 and 15%, respectively, under otherwise the same reaction
conditions (Table 1, entries 18, 19). After the screening of differ-
ent parameters, the optimized conditions were found to be
those in which the [K4Fe(CN)6] and catalyst loading are
0.7 equivalents and 0.6 mol%, respectively, in DMF (2 mL) at
958C (Table 1, entry 8).
Entry
R
X
3 [mol%]/t [h]
Yield [%][a,b,c]
1
2
3
4
5
6
7
8
4-NO2
H
4-CH3
4-OCH3
4-Cl
I
I
I
I
Br
Br
Br
Br
Br
Br
Br
Cl
Cl
0.5/3
0.5/5
0.5/8
94 (90)
99 (92)
99
99
88
91 (88)
91
95 (89)
(85)
87
N.R.
61 (58)[d]
N.R.[e]
0.5/10
0.5/10
0.6/12
0.6/16
0.5/10
0.5/10
0.8/15
0.6/12
1/24
H
2-CH3
4-NO2
4-Ph
4-OCH3
H
9
10
11
12
13
4-NO2
H
1/24
[a] Reaction conditions: NHC–Pd polymer 3, K2CO3 (2 equiv.), DMF (2 mL)
at 958C [b] Yields refer to GC results. [c] The data in parenthesis refer to
isolated yields. [d] 1 equiv. TBAB was added. [e] Reaction conditions:
1 equiv. TBAB at 1058C.
For the purpose of comparison, a commercially available Pd/
C catalyst (1wt% Pd) was also tested under the same reaction
conditions (Table 2, entry 11) and no cyanation product was
obtained using this catalyst in the reaction of bromobenzene
with [K4Fe(CN)6] within 24 h. The much higher activity of 3, as
compared to Pd/C and our previously developed IL@SBA-15-
Pd[16] may be attributed to the presence of the NHC ligand and
container-like structure of the NHC–Pd polymer,[18b] which en-
sures appropriate dispersion of highly active (possibly ultra-
small) Pd nanoparticles and facilitates the cyanation reaction.
Encouraged by these results, we then set up experiments to
study the performance of 3 in the cyanation of even more
challenging aryl chlorides under the described reaction condi-
tions. It was found that 4-nitrobenzonitrile was produced with
a good yield of 61% when 4-chloronitrobenzene was heated
for 24 h at 958C in the presence of 3 (1 mol%) and nBu4N+BrÀ
(TBAB, 1 equiv.) under otherwise optimized reaction conditions.
This yield is superior to that obtained from other existing het-
erogeneous Pd catalysts (Table 2, entry 12).[19] In this regard,
the addition of TBAB enhanced the yield significantly, presuma-
bly owing to the stabilization effect of the in situ generated Pd
nanoparticles.[18b,20] Unfortunately, our attempts to use this
method in its present form in the cyanation of chlorobenzene
itself were unsuccessful even when higher temperatures or
a higher loading of catalyst were applied (Table 2, entry 13).
Despite the high catalytic activity of NHC–Pd polymer 3 for
the synthesis of benzonitrile derivatives under relatively mild
conditions, our preliminary investigation showed that the cy-
anation reaction of heteroaryl halides such as bromo- or iodo-
thiophene and pyridine was unsuccessful under the described
conditions. Therefore, to find an alternative and more-appro-
priate reaction media for the cyanation of heteroaryl halides
we focused our attempts on performing this reaction in other
reactions solvents.
With the optimal reaction conditions in hand, the per-
formance and substrate scope of the presented NHC–Pd poly-
mer 3 were investigated with various aryl halides (Table 2). As
shown in Table 2, various aryl iodides with electron-withdraw-
ing and electron-donating groups converted to their benzoni-
trile derivatives with good to excellent yields (Table 2, en-
tries 1–4). For example, the less-reactive 4-iodoanisole convert-
ed into 4-methoxybenzonitrile in almost quantitative yield
after 10 h (Table 2, entry 4).
The same trends were also observed for substituted aryl bro-
mides with both electron-withdrawing and electron-donating
functionalities. Notably, the highly efficient cyanation of 2-bro-
motoluene, as a model for challenging 2-substituted aryl bro-
mides, afforded 2-cyanotoluene in excellent yield under similar
reactions conditions (Table 2, entry 7). Similarly, this catalyst
system was amenable to the efficient cyanation of highly deac-
tivated aryl bromides, but a high loading of 0.8 mol% and
longer reaction time were required to give excellent yields of
the corresponding product (Table 2, entry 10). By comparison
with cyanation coupling reactions catalyzed by most of the ex-
isting Pd catalysts, which employ highly expensive exogenous
phosphine ligands (such as tBu3P), the presented results
strongly confirm the high catalytic activity of 3 in the cyana-
tion reaction.
ChemCatChem 2015, 7, 2248 – 2254
www.chemcatchem.org
2250
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
In this regard, polyethylene glycols (PEGs) with a wide range
of molecular weight are known to be cheap, biodegradable,
safe, and nonvolatile (high boiling) nonionic solvents. PEGs are
capable of creating suitable reaction media in several metal-
catalyzed CÀC as well as CÀheteroatom bond-forming reac-
tions.[21] These solvents are also able to stabilize metal nano-
particles, demonstrating the strong potential of such solvent
systems in establishing highly durable metal nanoparticles-
based catalyst systems. For these reasons, we next decided to
study the impact of PEG-200 as a solvent in the cyanation of
aryl halides in the presence of 3. To do this, we chose bromo-
benzene as a model substrate for the cyanation reaction in
PEG-200 (2 mL) under otherwise optimal reaction conditions as
demonstrated in entry 8 of Table 1, to deduce whether the ex-
cellent conversion obtained in DMF could be repeated in PEG-
200. Surprisingly, we found that under the new reaction condi-
tions, the yield of benzonitrile dramatically decreased to the
benefit of the Ullmann homocoupling reaction of bromoben-
zene, affording an excellent biphenyl yield of ꢀ89%
(Scheme 2).
Table 3. Ullmann homocoupling of aryl halides using NHC-Pd polymer 3
as catalyst.
Entry
R
X
Yield [%][a,b]
1
2
3
4
5
6
7
8
9
H
I
I
98
90
88
92
85
89
54
4-OCH3
H
Br
Br
Br
Br
Br
Cl
Cl
4-CH2CH3
4-OCH3
2-CH3
2,6-CH3
H
75[c]
78[c]
4-CH2CH3
[a] Reaction conditions: NHC–Pd polymer 3 (0.5 mol%), K2CO3 (2 equiv.),
PEG (2 mL), 12 h at 758C. [b] Yields refer to isolated pure products.
[c] 20 mol% KI was added at 908C for 24 h.
Table 4. Ullmann homocoupling of heteroaryl halides using NHC-Pd
polymer 3 as catalyst.
Entry
1
Substrate
Product
Yield [%][a,b]
95
2
90
3
4
93
82
Scheme 2. Results of the cyanation reaction of bromobenzene with PEG-200
as solvent.
[a] Yields refer to isolated products. [b] Reaction conditions: NHC–Pd
polymer 3 (0.5 mol%), PEG-200 (2 mL), K2CO3 (2 equiv.), 758C, and 12 h.
This surprisingly different behavior between DMF and PEG-
200 prompted us to further investigate whether PEG-200 could
be employed as solvent system in the Ullmann coupling of var-
ious aryl halides by using 3 as catalyst.
With this initial interesting data in hand, we next proceeded
to set up appropriate experiments for the Ullmann homocou-
pling reactions because symmetrical biaryls are also very prom-
ising synthetic units in organic transformations.[22] This reaction
is generally performed at very high temperature (over 2008C)
by using copper as a catalyst.[23] Unfortunately, at elevated
temperatures, the scope of reaction is strongly limited to ther-
mostable starting materials. However, our preliminary investi-
gations showed that the present reaction can proceed even at
758C with high efficiency in the presence of 0.5 mol% of the
catalyst. To illustrate the general applicability of NHC–Pd poly-
mer 3, the present protocol was then subsequently extended
to the Ullmann coupling of a series of substituted aryl halides.
Table 3 summarizes the results of this investigation. The proto-
col was amenable to the homocoupling of various types of
substituted aryl halides (I, Br, Cl) as well as sterically hindered
substrates.
action conditions with PEG as a green reaction media. Based
on our findings, various heteroaryl halides, comprising pyridine
and thiophene rings, converted to their corresponding Ull-
mann product with excellent yields under very mild thermal
conditions (Table 4).
The exact role of PEG in the process is not clear at this
stage. However, a plausible explanation for this phenomenon
is the trapping (or scavenging) of CNÀ ions by formation of hy-
drogen bonds with PEG. In other words, it is most likely that
the cyanide ions are entrapped among the solvent molecules.
The most dominant reaction under these conditions is the Ull-
mann homocoupling. A schematic illustration for the scaveng-
ing of CNÀ ions in the PEG media and the formation of the re-
action core is shown in Scheme 3.
Recycling experiments were performed for both cyanation
and homocoupling reactions by using bromobenzene as the
substrate under the optimized reaction conditions. For cyana-
tion, after reaction termination the products were extracted
with organic solvent and the catalyst/solvent (DMF or PEG)
mixture was recharged with starting materials. As shown in
To extend the application of NHC-Pd polymer 3, the Ullmann
homocoupling reactions of extremely challenging heteroaryl
substrates were also investigated under the presented mild re-
ChemCatChem 2015, 7, 2248 – 2254
www.chemcatchem.org
2251
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
obtained in 86% yields (determined by GC analysis), which is
to some extent lower than the same reaction in the absence
of mercury (Table 2, entry 6). Although the mercury poisoning
experiment could not solely prove the formation of nanoparti-
cles during the reaction, the weak mercury positive test ac-
companied with the observed changes in the molecular-
weight profile of the recovered NHC–Pd polymer 3 after the
recycling experiment, as evidenced by GPC of the recovered
catalyst, provides reasonable evidence that this polymer could
be the source of the active Pd species in the form of either Pd
nanoparticles or fragmented NHC–Pd complexes. Although, we
were not able to detect the formation of Pd nanoparticles by
direct TEM investigation, the above-mentioned observations
clearly demonstrate that our NHC–Pd polymer might be
indeed a source of a trace amount of (non-observable) Pd
nanoparticles, which in turn could actually catalyze the titled
reactions. These results are also in agreement with our recent
studies of using the same organometallic polymer catalyst sys-
tems in the Suzuki–Miyaura coupling reaction.[18b] These fea-
tures are schematically demonstrated in Scheme 4.
Scheme 3. Proposed model describing the deactivation of CNÀ ions in the
PEG media by formation of a hydrogen bond to the solvent in favor of the
homocoupling reaction inside the reaction core.
Table 5. Recycling experiments for Ullmann homocoupling and cyanation
of bromobenzene using NHC-Pd polymer 3 as catalyst.
Cyanation
Reaction runs
GC yield [%]
1st 2nd 3rd 4th
90 89 72 72
Ullmann homocoupling Reaction runs
1st 2nd 3rd 4th 5th
87 84 77
Isolated yield [%] 92 91
Scheme 4. The proposed pathway for the in situ generation of Pd nanoparti-
cles by partial decomposition of the NHC–Pd polymer, which could actually
catalyze the described coupling reactions.
Table 5, the catalyst could be reused at least in four successive
reaction runs with only a slight decrease in the yields of ben-
zonitrile. Table 5 summarizes the results of the recycling experi-
ments for cyanation and the Ullmann homocoupling reaction
of bromobenzene.
Conclusions
We have demonstrated that a NHC–Pd polymer 3 bearing an
N-dodecyl group is a highly efficient and recoverable precata-
lyst for the cyanation of aryl halides (I, Br, and Cl) with
[K4Fe(CN)6] in DMF under relatively mild conditions. We also
found that by changing the reaction solvent from DMF to
poly(ethylene glycol) (MWAV ꢀ200), the selectivity patterns of
this reaction were remarkably shifted toward highly efficient
formation of Ullmann homocoupling products even in the case
of highly challenging substrates such as aryl chlorides and het-
eroaryl bromides. A model highlighting the possible role of
PEG in attaining the high degree of selectivity towards the Ull-
mann coupling reaction was also described. The catalyst exhib-
ited high tolerance to CNÀ ions during the cyanation reaction.
The catalyst was successfully recycled and reused for four and
five successive runs in cyanation and the Ullmann homocou-
pling reaction of bromobenzene, respectively, with only slight
decrease in activity. Our studies on the recovered catalyst from
the cyanation reaction demonstrated that this polymer could
be the source of the production of active Pd species in the
form of either Pd nanoparticles or fragmented NHC–Pd com-
In the case of the Pd-catalyzed cyanation reaction the cata-
lyst could be deactivated by CNÀ ions and forms inactive
[Pd(CN)4]2À and [(CN)3PdAr]2À species during the reaction.[15] In
our recycling experiment for the cyanation reaction the first
and second reaction runs gave constant yields of benzonitrile.
However, for the third and fourth recycling experiment the
yield was reduced to 72%. The fact that the yields of both cou-
pling reactions for the reused polymer go down linearly in
each step suggests that organometallic polymer 3 is gradually
decomposing under the described reaction conditions, thus it
is very unlikely that it functions as a truly heterogeneous cata-
lyst.
To investigate whether the real catalytic species are naked
Pd nanoparticles or robustly-bound NHC–Pd, an independent
mercury poisoning test (Hg/Pd=400:1) was set up for the cou-
pling reaction of bromobenzene with [K4Fe(CN)6] to give ben-
zonitrile under the otherwise standard reaction conditions. To
do this, an excess amount of Hg0 (Hg/Pd=400:1) was added
to the reaction mixture. It was found that benzonitrile was still
ChemCatChem 2015, 7, 2248 – 2254
www.chemcatchem.org
2252
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
to cool to RT and then small portion of distilled water was added
to the mixture followed by extraction with n-hexane or diisopropyl
ether. The upper organic phase was separated and washed several
times with water dried over MgSO4 and then the solvent was
evaporated under reduced pressure. Pure products were obtained
after recrystallization or purification by column chromatography on
silica with n-hexane/ethyl acetate as eluent.
plexes. Work is underway to expand the application of this cat-
alyst system in other important chemical transformations.
Experimental Section
Synthesis of benzo-bis(imidazole) (1)
1,2,4,5-Benzenetetraamine tetrahydrochloride (284 mg, 1.00 mmol)
was poured into a round-bottomed flask charged with a magnetic
stir bar. Formic acid (88–99%, 15 mL) was added and the flask was
fitted with an air-jacketed condenser. The reaction was performed
in an oil bath at 1008C for 36 h. The reaction mixture was then al-
lowed to cool, decanted into ice-cold water (equal volume to
formic acid) and neutralized with K2CO3. Neutralization caused pre-
cipitation of the product, which was collected by vacuum filtration,
washed with cold water, and dried under vacuum in the presence
of P2O5. The desired product was obtained as a light brown solid.
1H NMR (250 MHz, [D6]DMSO): d=7.65 (s, 2H), 8.15 ppm (s, 2H);
13C NMR (250 MHz, [D6]DMSO): d=99.7, 135.4, 143.0 ppm.
General procedure for the Ullmann homocoupling reaction
using NHC-polymer 3
Aryl or heteroaryl halides (1 mmol), catalyst 3 (0.5 mol%), K2CO3
(2 mmol), and PEG-200 (2 mL) were placed in a glass flask. The re-
action was heated at 758C with vigorous stirring for 12 h. The reac-
tion progress was monitored by TLC, and after completion of the
reaction, the mixture was allowed to cool to RT and then small por-
tion of distilled water was added to the mixture followed by ex-
traction with n-hexane or diisopropyl ether. The upper organic
phase was separated and dried over MgSO4 and the solvent was
evaporated under reduced pressure. Pure products were obtained
after recrystallization or purification by column chromatography on
silica with n-hexane/ethyl acetate as eluent.
Synthesis of tetradodecyl benzo-bis(imidazolium) bromide
(2)
In a well-dried two-necked 100 mL Schlenk flask, bis(imidazole)
(475 mg, 3.00 mmol) was added to a solution of NaH (245 mg,
60 wt%, 6.10 mmol) in toluene (15 mL) under an argon atmos-
phere. The resulting solution was heated to 110 8C for 1 h. After
cooling the solution to RT, 1-bromododecane (4.8 mL, 20.00 mmol)
was added through a syringe. The suspension was placed in an oil
bath at 110 8C for 1 h, then dry DMF (10 mL) was added through
a syringe and the reaction was maintained at 110 8C for 12 h, then
608C for 8 h. Upon completion, the suspension was allowed to
cool. The concentrated suspension in vacuum was added to 2:1
CH2Cl2/MeOH. After removal of insoluble NaBr, ethyl acetate was
added to the resultant oil until precipitation of the product oc-
curred. The solids were collected by vacuum filtration and dried
under vacuum over P2O5 to give 2.31 g (77%) of the product as
a light pink powder. 1H NMR (500 MHz, CDCl3): d=10.08 (s, 2H),
9.53 (s, 2H), 4.93 (t, J=6.35 Hz, 8H), 1.98 (p, 8H), 1.25–1.35 (m,
80H), 0.90 ppm (t, J=6.9 Hz, 12H), 13C NMR (250 MHz, CDCl3): d=
14.1, 22.6, 26.5, 29.1, 29.3, 29.4, 29.5, 29.6, 29.6, 31,9, 49.0, 100.7,
129.7, 144.9 ppm.
Recycling experiments
All recycling experiments were performed on a 5 mmol scale of
bromobenzene to facilitate easy recovery of the catalysts. As an ex-
ample, after the first use of the catalyst 3 in the Ullmann homo-
coupling of bromobenzene to yield 92% of biphenyl (isolated yield
as described above), the catalyst was separated by centrifugation
and thoroughly washed with deionized water and EtOH and dried
at 408C under vacuum. The recovered materials were then success-
fully used in the same way for five reaction cycles.
Acknowledgements
The authors greatly acknowledge Iran National Science Founda-
tion (INSF) through a grant no. G014 and IASBS research council
for the financial support of this work.
Keywords: coupling reactions
· cyanation · polyethylene
glycol · polymers · Ullmann reaction
Typical procedure for preparation of NHC-polymer 3
[1] a) T. Schareina, A. Zapf, M. Beller, J. Organomet. Chem. 2004, 689, 4576–
4583; b) G. Liang, Y. Xu, I. B. Seiple, D. Trauner, J. Am. Chem. Soc. 2006,
128, 11022–11023; c) S. Jaegli, J.-P. Vors, L. Neuville, J. Zhu, Tetrahedron
2010, 66, 8911–8921; d) Z. Rappoport, Chemistry of the Cyano Group,
Wiley, London, 1970; e) R. C. Larock, Comprehensive Organic Transforma-
tions: A Guide to Functional Group Preparations, VCH, New York, 1989.
[2] M. Alterman, A. Hallberg, J. Org. Chem. 2000, 65, 7984–7989.
[3] a) K. W. Rosenmund, E. Struck, Ber. Dtsch. Chem. Ges. 1919, 52, 1749–
1756; b) J. von Braun, G. Manz, Justus Liebigs Ann. Chem. 1931, 488,
111–126; c) R. G. R. Bacon, H. A. O. Hill, J. Chem. Soc. 1964, 1097–1107;
d) J. Lindley, Tetrahedron 1984, 40, 1433–1456.
Bis(imidazolium) bromide 2 (1.0 mmol) was dissolved in DMSO
(5 mL, ca. 0.2m) and Pd(OAc)2 (1.0 equiv., 112 mg), was added in
one portion. The solution was placed in a preheated oil bath at
110 8C and stirred in air for 5–10 h. The reaction mixtures typically
darkened in color (to orange-brown) as the reaction progressed.
The cooled reaction mixture was added to H2O. The resulting pre-
cipitated polymer was collected by vacuum filtration and dried
under vacuum.
[4] a) T. Sandmeyer, Ber. Dtsch. Chem. Ges. 1885, 18, 1946–1948; b) T. Sand-
meyer, Ber. Dtsch. Chem. Ges. 1885, 18, 1492–1496; c) T. Sandmeyer,
Ber. Dtsch. Chem. Ges. 1884, 17, 2650–2653.
General procedure for the cyanation reaction using NHC-
polymer 3
[5] a) A. C. Stevenson, Ind. Eng. Chem. 1949, 41, 1846–1851; b) W. I.
Denton, R. B. Bishop, H. P. Caldwell, H. D. Chapman, Ind. Eng. Chem.
1950, 42, 796–800; c) A. Martin, N. V. Kalevaru, B. Lücke, J. Sans, Green
Chem. 2002, 4, 481–485.
[6] P. Anbarasan, T. Schareina, M. Beller, Chem. Soc. Rev. 2011, 40, 5049–
5067.
A
mixture of aryl halide (1.0 mmol), finely powdered
[K4Fe(CN)6]·3H2O (0.7 mmol), K2CO3 (2 mmol), and NHC–Pd polymer
3 (0.5–1 mol%) in DMF was stirred vigorously under an argon at-
mosphere at 958C. The reaction progress was monitored by GC
analysis. After completion of the reaction the mixture was allowed
ChemCatChem 2015, 7, 2248 – 2254
www.chemcatchem.org
2253
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[7] K. Takagi, T. Okamoto, Y. Sakakibara, S. Oka, Chem. Lett. 1973, 471–474.
[8] a) S. A. Weissman, D. Zewge, C. Chen, J. Org. Chem. 2005, 70, 1508–
1510; b) G. Y. Zhang, J. T. Yu, M. L. Hu, J. Cheng, J. Org. Chem. 2013, 78,
2710–2714; c) A. V. Ushkov, V. V. Grushin, J. Am. Chem. Soc. 2011, 133,
10999–11005; d) H. Yu, R. N. Richey, W. D. Miller, J. Xu, S. A. May, J. Org.
Chem. 2011, 76, 665–668; e) Z. Zhang, L. S. Liebeskind, Org. Lett. 2006,
8, 4331–4333; f) S. Xu, X. Huang, X. Hong, B. Xu, Org. Lett. 2012, 14,
4614–4617; g) J. Peng, J. Zhao, Z. Hu, D. Liang, J. Huang, Q. Zhu, Org.
Lett. 2012, 14, 4966–4969; h) N. Chatani, T. Hanafusa, J. Org. Chem.
1986, 51, 4714–4716.
Commun. 2006, 1727–1729; i) A. G. Tennyson, D. M. Khramov, C. D. Var-
nado, Jr., P. T. Creswell, J. W. Kamplain, V. M. Lynch, C. W. Bielawski, Orga-
nometallics 2009, 28, 5142–5147; j) C. D. Varnado Jr., V. M. Lynch, C. W.
Bielawski, Dalton Trans. 2009, 7253–7261; k) A. G. Tennyson, J. W. Kam-
plain, C. W. Bielawski, Chem. Commun. 2009, 2124–2126; l) D. J. Coady,
D. M. Khramov, B. C. Norris, A. G. Tennyson, C. W. Bielawski, Angew.
Chem. Int. Ed. 2009, 48, 5187–5190; Angew. Chem. 2009, 121, 5289–
5292; m) K. A. Williams, A. J. Boydston, C. W. Bielawski, J. R. Soc. Interface
2007, 4, 359–362; n) J. A. V. Er, A. G. Tennyson, J. W. Kamplain, V. M.
Lynch, C. W. Bielawski, Eur. J. Inorg. Chem. 2009, 1729–1738.
[9] J. R. Dalton, S. L. Regen, J. Org. Chem. 1979, 44, 4443–4444.
[10] a) A. Sekiya, N. Ishikawa, Chem. Lett. 1975, 277–278; b) Y. Akita, M. Shi-
mazaki, A. Ohta, Synthesis 1981, 974–975; c) K. Tanji, T. Higashino, Het-
erocycles 1990, 30, 435–440; d) M. Sundermeier, A. Zapf, M. Beller, J.
Sans, Tetrahedron Lett. 2001, 42, 6707–6710.
[18] a) B. Karimi, P. Fadavi Akhavan, Chem. Commun. 2009, 3750–3752; b) B.
Karimi, P. Fadavi Akhavan, Inorg. Chem. 2011, 50, 6063–6072; c) B.
Karimi, P. Fadavi Akhavan, Chem. Commun. 2011, 47, 7686–7688.
[19] D. Ganapathy, S. S. Kotha, G. Sekar, Tetrahedron Lett. 2015, 56, 175–178.
[20] a) M. LysØn, K. Kçhler, Synthesis 2006, 692–698; b) M. T. Reetz, J. G. de V-
ries, Chem. Commun. 2004, 1559–1563.
[21] a) J. Chen, S. K. Spear, J. G. Huddleston, R. D. Rogers, Green Chem. 2005,
7, 64–82; b) A. Haimov, R. Neumann, Chem. Commun. 2002, 876–877;
c) J. Mao, J. Guo, F. Fang, S. Ji, Tetrahedron 2008, 64, 3905–3911; d) C.
Luo, Y. Zhang, Y. Wang, J. Mol. Catal. A 2005, 229, 7–12; e) H. Firouzaba-
di, N. Iranpoor, M. Gholinejad, Adv. Synth. Catal. 2010, 352, 119–124;
f) L. Cheng, Y. Zhong, Z. Ni, H. Du, F. Jin, Q. Rong, W. Han, RSC Adv.
2014, 4, 44312–44316; g) M. Gholinejad, B. Karimi, F. Mansouri, J. Mol.
Catal. A 2014, 386, 20–27; h) A. R. Hajipour, G. Azizi, Green Chem. 2013,
15, 1030–1034; i) Y. Zhong, X. Gong, X. Zhu, Z. Ni, H. Wang, J. Fu, W.
Han, RSC Adv. 2014, 4, 63216–63220; j) M. Vafaeezadeh, M. Mahmoodi
Hashemi, J. Mol. Liq. 2015, 207, 73–79.
[11] D. M. Tschaen, R. Desmond, A. O. King, M. C. Fortin, B. Pipik, S. King, T. R.
Verhoeven, Synth. Commun. 1994, 24, 887–890.
[12] a) T. Schareina, A. Zapf, M. Beller, Chem. Commun. 2004, 1388–1389;
b) T. Schareina, A. Zapf, W. Maegerlein, N. Mueller, M. Beller, Tetrahedron
Lett. 2007, 48, 1087–1090; c) T. Schareina, R. Jackstell, T. Schulz, A. Zapf,
A. Cotte, M. Gotta, M. Beller, Adv. Synth. Catal. 2009, 351, 643–648.
[13] [K4Fe(CN)6] is known to be a soluble salt in water with low dissociation
and thus can slowly deliver cyanide ions in the reaction. LD50 of
[K4Fe(CN)6] is even lower than that for NaCl. However, it should be con-
sidered that in contact with acids, potassium ferrocyanide can poten-
tially release HCN. Moreover, the use of [K4Fe(CN)6] in the food industry
is limited to low concentrations.
[14] M. Sundermeier, A. Zapf, M. Beller, Angew. Chem. Int. Ed. 2003, 42,
1661–1664; Angew. Chem. 2003, 115, 1700–1703.
[15] a) K. D. Dobbs, W. J. Marshall, V. V. Grushin, J. Am. Chem. Soc. 2007, 129,
30–31; b) S. Erhardt, V. V. Grushin, A. H. Kilpatrick, S. A. Macgregor, W. J.
Marshall, D. C. Roe, J. Am. Chem. Soc. 2008, 130, 4828–4845.
[16] B. Karimi, A. Zamani, F. Mansouri, RSC Adv. 2014, 4, 57639–57645.
[17] For examples, see: a) A. J. Boydston, K. A. Williams, C. W. Bielawski, J.
Am. Chem. Soc. 2005, 127, 12496–12497; b) D. M. Khramov, A. J. Boy-
dston, C. W. Bielawski, Angew. Chem. Int. Ed. 2006, 45, 6186–6189;
Angew. Chem. 2006, 118, 6332–6335; c) A. J. Boydston, D. M. Khoramov,
C. W. Bielawski, Tetrahedron Lett. 2006, 47, 5123–5125; d) D. M. Khra-
mov, A. J. Boydston, C. W. Bielawski, Org. Lett. 2006, 8, 1831–1834;
e) K. A. Williams, A. J. Boydston, C. W. Bielawski, Chem. Soc. Rev. 2007,
36, 729–744; f) K. A. Williams, C. W. Bielawski, Chem. Commun. 2010, 46,
5166–5168; g) A. B. Powell, C. W. Bielawski, A. H. Cowley, J. Am. Chem.
Soc. 2009, 131, 18232–18233; h) J. W. Kamplain, C. W. Bielawski, Chem.
[22] a) A. Molnµr, Palladium-Catalyzed Coupling Reaction: Practical Aspects
and Future Developments, Wiley-VCH, Weinheim, 2013; b) J. M. Lehn,
Supramolecular Chemistry, Wiley-VCH, Weinheim, 1995; c) T. Yamamoto,
T. Maruyama, Z. Zhou, T. Ito, T. Fukada, Y. Yomeda, F. Begum, T. Ikeda, S.
Sasaki, H. Takezoe, A. Fukada, K. Kubota, J. Am. Chem. Soc. 1994, 116,
4832–4845; d) S. Zhu, T. M. Swager, Adv. Mater. 1996, 8, 497–500; e) J.
Hassan, M. SØvignon, C. Gozzi, E. Schulz, M. Lemaire, Chem. Rev. 2002,
102, 1359–1470; f) Y. Liao, W. Hung, T. Hou, C. Lin, K. Wong, Chem.
Mater. 2007, 19, 6350–6357; g) P. E. Fanta, Chem. Rev. 1964, 64, 613–
632.
[23] F. Ullmann, J. Bielecki, Ber. Dtsch. Chem. Ges. 1901, 34, 2174–2185.
Received: April 6, 2015
Published online on July 1, 2015
ChemCatChem 2015, 7, 2248 – 2254
www.chemcatchem.org
2254
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim