Mechanistic Studies of Ullmann-Type C–N Coupling Reactions: Carbonate-Ligated Copper(III) Intermediates

Article abstract of DOI:10.1002/cctc.201601174

In Ullmann-type C?N coupling reactions, the involvement of Cu^III species has been proposed many times on the basis of the oxidative addition–reductive elimination (OA-RE) path for these reactions, but actual species could not be traced in experimental studies. In the C?N coupling reactions, carbonate and phosphate were considered widely as bases. In the present study, Cu-mediated C?N coupling reactions of aryl halides and NuNH (amide and imide) were investigated extensively, and we provide direct spectroscopic evidence of actual Cu^III species. For the first time, we reveal that carbonate and phosphate ions act as bidentate ligands as well as a base in the catalytic cycle, and thus the actual intermediate species is a carbonate- or phosphate-ligated, distorted octahedral Cu^III complex. Our experimental and computational studies have strengthened the hypothesis that these reactions follow an OA-RE path.

Full text of DOI:10.1002/cctc.201601174

Accepted Article
Title: Mechanistic Studies of Ullmann Type C-N Coupling Reactions:
Carbonate Ligated Cu(III) Intermediate
Authors: Kamlesh K Gurjar and Rajendra Kumar Sharma
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201601174
Link to VoR: http://dx.doi.org/10.1002/cctc.201601174
A Journal of
www.chemcatchem.org

ChemCatChem
10.1002/cctc.201601174
FULL PAPER
Mechanistic Studies of Ullmann Type C-N Coupling Reactions:
Carbonate Ligated Cu(III) Intermediate
a
b
Kamlesh K Gurjar , Rajendra K Sharma*
Abstract: In Ullmann type C-N coupling reactions involvement of
Cu(III) species was proposed many times considering the oxidative
addition–reductive elimination (OA-RE) path for these reactions but
actual species could not be traced out in experimental studies. In the
C-N coupling reactions carbonate and phosphate were widely
considered as base. In the present study, copper-mediated C-N
coupling reactions of aryl halides and NuNH (amide and imide) were
extensively investigated and present studies provide direct
spectroscopic evidences of actual Cu(III) species. First time, these
studies have also revealed that carbonate and phosphate ions act as
bidentate ligand as well as base in the catalytic cycle and thus actual
intermediate species is carbonate or phosphate ligated distorted
octahedral Cu(III) complex. Present experimental and computational
studies have strongly strengthened the hypothesis that these
reactions follow OA-RE path.
invalidated negative radical clock experiment and lack of
[10]
inhibition by radical scavengers.
In 2010 Buchwald and co-
workers suggested that in free radical mechanism, the short
lived free radical remains in caged pair and therefore, invalidate
[
20]
the radical clock experiment. More recently, on the basis of
[
21-23]
experimental and computational findings
some research
[
10,11,24]
groups have suggested the OA-RE pathway (Scheme 1a).
(
a)
(b)
NuNH, CO32-
CO 2-
I-
NuNH
3
L2CuCO3
PhNNu
Base
RE
Ph
L2Cu-HNNu
L2Cu-NNu
PhI
OA
L2CuI
PhI
PhNNu
OA
O
O
C
L2Cu
I
O
HNNu
RE
I
L2Cu
NHNu
Ph
L2CuNNu
Ph
I-
CO32-
L2= H C NH HN CH3
3
Introduction
Copper-catalyzed nucleophilic aromatic substitution reactions
Scheme 1. OA-RE mechanism pathways (a) Previously suggested (b)
proposed catalytic cycle in current study (NuNH= amides and imides).
[
1]
were discovered more than a century ago by Ullmann and
[
2]
[3]
Goldberg. However, due to remarkable work of Buchwald
[
4,5]
and Taillefer
groups, the reaction conditions could be
revitalized by introduction of auxiliary ligands and as a result
coupling reactions have been widely investigated by considering
them as cheaper alternative of the Pd-catalyzed reactions. But
mechanism of Cu-mediated reactions are not well understood
[24]
Taillefer and co-workers
have highlighted a very interesting
aspect of these reactions that electron rich and electron poor
sites of the ligands play individual role in oxidative addition(OA)
and reductive elimination (RE) respectively. In another work
same group reported the discriminative role of carbonate and
hydroxide in diaryl ether synthesis. In reaction of ArI, phenol is
[
6-8]
like Pd-catalyzed coupling reactions.
Identification of actual intermediate as well as mechanism of
these reactions has always been exciting challenge for the
the product in presence of CsOH while additional non-equivalent
[9]
chemists.
The free radical substitution via Cu /Cu and oxidative addition-
[23]
amount of Cs
2 3
CO in same reaction gives diaryl ether.
I
II
[25]
Recently Sambiagio et al.
reported that electron withdrawing
I
III
reductive elimination (OA-RE) via Cu /Cu
mechanisms
group on picolinamide ligands (acetates as solvent) increases
the efficiency of catalytic system.
It has also been reported that organic bases like amines
[
10,11]
(
Scheme 1a) are under debate for Cu-mediated coupling.
is commonly accepted that Cu is the active catalytic species in
On the basis of experimental and
computational studies, several research groups
proposed the involvement of Cu intermediates and Cu has
been demonstrated in electrochemical studies by Taillefer and
It
I
[10]
are
[
12,13]
coupling reactions.
[26a,b]
[10]
not efficient whereas organic ionic
like K CO , Cs CO and K PO are excellent additives and play
vital role in deprotonation of nucleophile.
and inorganic bases
[
13-19]
have
2
3
2
3
3
4
III
III
[26a]
Diversity of the
interesting theoretical and experimental results prompted us to
further strengthen the mechanistic aspects of these reactions.
[
16]
III
Jutand groups.
pincer complex by Stahl and co-workers.
intermediates could not be demonstrated through spectroscopic
Cu has been demonstrated in predefined
[14,15]
III
However, Cu
[
9]
studies in the reaction of typical coupling partners. Free radical
mechanism was invoked many times but discarded due to
Results and Discussion
A number of reactions were performed with typical substrates,
aryl iodides (1a, 1b and 1c), amide/imide (2a, 2b and 2c) and
2
-
3-
[
a] Prof. Kamlesh K Gurjar, Chemical Department, VGEC, Ahmedabad
Gujarat, India-382424
b]* Dr. Rajendra K Sharma, DESM, RIE, NCERT, Ajmer, Rajasthan,
widely used bases (CO
4
and PO ). Reactions were in situ
3
monitored spectroscopically. DFT studies were also performed
to investigate the thermodynamic possibilities of the proposed
intermediate species. Specific emphasis was given to find out
the role of the bases.
[
India-305004
gurjarkamlesh@vgecg.ac.in
rajncert@gmail.com
Initially we screened some ligands for Cu-mediated PhenylC-Nimide
Supporting information for this article is available on http://dx.doi.org/........
1 2
bond formation (Scheme 2) and L and L were found most
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601174
FULL PAPER
efficient ligands from the list of L
1
-L
. Ligands with poor electron
6
) or have greater steric hindrance were not
6
. However, L
2
was found to
increases with increasing amount of base (up to 100%) but
decrease later on in present experimental conditions (entry 5,
table 1). This harsh effect is more in DMSO than toluene.
be slightly more efficient than L
donor sites (L -L
1
3
found suitable in our experimental conditions.
I
O
O
Toluene 8ml
o
1
00
C
2 3
Table 1. Variation in yield with [Ligand]/[K CO ] ratio.
+
HN
N
CuI 5 mol%
.3 eq L1-L6,
.5 eq K CO
3
I
O
O
0
10
Yield%
O
1 eq
O
0
5 mol% CuI
1
1
6
.2 eq
a (PhI) 2b (PhtNH)
mmol 5 mmol
2
+
HN
N
L1 36%*
L2 52%
L3-L6 trace
o
10
Reflux, 17 h
100 C, 17h
1.2 eq
O
1
eq
O
Entry
K
(
2
CO
eq)
3
Ligand
[Ligand]/[ K
2 3
CO ]
Isolated
yield%
Ligands
O
O
1 2
L or L (eq)
NH2
F3C
Ph
H N
O
t
2
L5
1
2
3
4
5
0
0
0
0
H
N
L1
N
N
OH
N
H
N
d
HO
1
0.2 L1
0.2
0.3
0.2
0.1
0.2
0.6
0.3
0.3
0
L2
L3
L4
L6
d
1
0.3 L
0.2 L
0.3 L
0.2 L
1
2
1
2
36
t
1 6
Scheme 2. Model reaction for screening of ligands (L -L ).
1
32
t
3
2
1
. Base free
21
2
1.0 2. Pyridine 0.5 eq
a
t
6
1
30
3
4
5
6
7
. L 0.15 eq
4
. K HPO4 0.5 eq
2
t
7
8
9
a
0.5
1
0.3 L
0.3 L
0.3 L
2
2
2
52
. Potassium ascorbate 0.5 eq
. Potassium tartrate 0.5 eq
t
41
5
. L 1 eq
4
4
5
4
2
0
d
1
5
2
1
[
2 3
a] Cs CO as a base [b] Solvent: d-DMSO and t-toluene
0
1
0
2
3
4
6
7
Base (2-7)
Table 2. Variation in yield with [Ligand/CuI] ratio.
Figure 1. Screening of bases (serial No.2-7). Reaction conditions: 0.3 eq
ligand L
under N
2
, 1.2 eq PhI and 1 eq phthalimide in 8 mL toluene, refluxed for 17 h
at 100 C). For more detail, see Table S2.
I
o
O
O
2
0
.2 eq K PO
3
4
+
HN
N
o
1
00 C, 17h
1
.2 eq
O
10
During screening of the bases (Figure 1 and table S2), bases (2-
) were not found suitable like K CO and K PO4.Therefore, for
next reactions L -L and K CO /K PO were selected as ligand
and base respectively and studies were carried out taking
carbonate and L -L in different ratio in toluene and DMSO (CuI
mol% as constant amount).It has been observed that the yield
significantly improves with increasing ligand to carbonate ratio
Figure 2a and table 1). It may be due to three reasons, the
positive order in respect of ligand concentration or negative
order in respect of the base concentration or both. Careful
comparison of entry 7 and 8 (Table 1), clearly confirms that
reaction shows negative order in respect to the amount of base.
1 eq
(mol%)
O
7
2
3
3
Entry
CuI mol%
L
1
[L
1
]/[CuI]
Isolated yield%
1
2
2
3
3
4
d
1
2
3
4
5
6
17
10
5
20
30
30
15
20
30
1.17
42
1
2
d
3
20
5
t
6
30
(
d
2
7.5
2
43
t
10
0
7
d/t
-
0
3 4
In next set of reactions, base (K PO ) was fixed in sub-
Solvent: d-DMSO and t-toluene
equivalent amount 20 mol% (Table 2), entry 4 indicates that
catalytic cycle significantly depends (positive order) on ligand
concentration (Figure 2b). These results have good agreement
with previous studies reported by Davies and co-workers for
It was assumed that at very high concentration of base,
deprotonation of NuNH (nucleophile 2b or 2c) occurs and inert
multi ligated copper species (NuN) Cu are formed. As a result,
2
catalytic process does not initiate. To overcome this
consequence, more concentration of ligand is required to
[
22]
organic bases. With respect to product yield, K
a better additive in DMSO than in toluene and vice versa for
CO (Figure 2). We found that the yield of product initially
3
PO
4
is slightly
[27]
K
2
3
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601174
FULL PAPER
generate the active LCuNNu species (L=L
2
1
or L ). At lower
(15%), even at very low amount of CuI (2%). However, 21%
concentration of base, catalytic process can initiate via
LCuNHNu (without deprotonation).
yield was noted with L
(as base, instead of K
substitute of K CO but not a substitute of L
2 4
(ligand) and stoichiometric amount of L
2
CO
3
).This indicates that L
4
can act as
2
3
2
4
ligand. L is poor
60
50
40
30
20
10
0
(
a)
base than pyridine and it seems that it has some other role
besides deprotonation of nucleophile. L is relatively electron
poor and has two harder donor sites like carbonate. Thus, above
observations indicate that L and K CO have similar role (may
t, L2
4
t,L2
t,L1
t,L2
t,L2
4
2
3
be a second bidentate ligand) in the catalytic cycle. In this case,
pyridine (unidentate and borderline hard base) is not expected to
be the substitute of carbonate (bidentate and hard base).
From these observations, we suspected some other role of
carbonate species instead of base alone and thus further
investigation were initiated to find out the actual role of
carbonate in term of base and second ligand.
[
28]
[
[
L1 or L2]
K2CO3]
CuI 5 mol%
Cs2CO3
B=
d,L1
d,L2
t
t,L1
0.1
0.0
0.2
0.3
0.4
0.5
0.6
B
50
40
30
20
10
0
In view of the above, our current hypothesis is that, 1.
(
b)
I
d,2*
Nucleophile can coordinate with Cu even without deprotonation
d,17*
Masked
2-
3-
and 2. CO
3
/ PO
4
can act as chelating ligand. While working on
the hypothesis, we found that in the electronic spectrum of 2b
t,5*
(
Figure S1) hypso and hyper-chromic shifts are observed on
[
29]
d,10*
addition of CuI (see the supporting information).
observation reflects the ligation of 2b on Cu even in the absence
of base. Similar result was also observed for 2c and Cu
This
I
I
[
L1]
B=
[
CuI]
t,10*
interaction. Shifting in λmax on addition of different additives
*
CuI mol%
K3PO4 20%
I
clearly indicates that interaction between Cu and nucleophile
2b/2c can take place in absence of base also and deprotonation
0
1
2
3
4
5
6
7 8
B
of nucleophile is essentially not required to coordinate with
Cu.
I [16]
Figure 2. Linear fitting of yield% with (a) [Ligand]/[K
2 3
CO ] ratio. (b)
[
Ligand]/[K PO ] ratio. Solvent: d-DMSO (red dots) and t-toluene (blue dots).
3
4
NH
CuI
NH
Toluene, 100 o
NH
Cu NHNu
NH
C
NuNH
+
3
0 min reflux, N2
(
2b or 2c)
1
eq
(CuI 5 mol%, 0.3 eq L2)
(3b or 3c)
I
3
0% L2
3
0% L
nucleophiles
O
4
aryl halides (PhI)
I
5% CuI
5
% CuI
21%
0
%
O
O
I
1
eq K CO3
1eq L4
2
I
COOH
N
HN
HN
2b(NHPht) 2c(NHPy) 1a
+
O2N
5
0
% CuI
.5eq Pyridine
O
2% CuI
O
O
2a(NPht)
1b
1c
1
1
1
eq Pyridine
5% L2
5% L4
0%
13%
3
0%L2
HN
Scheme 4. Synthesis of 3b/3c in toluene and list of typical nucleophiles and
aryl halides.
O
o
Reflux, 17 h, Toluene, 100 C
[
27]
Hartwig and co-workers have demonstrated that LCuNPht and
LCuNPy are the actual species that react with aryl halides. DFT
calculations suggest that thermodynamic factors are slightly in
favor of the formation of LCuNPy than LCuNHPy. However, on
the basis of above mentioned observations and not much
Scheme 3. Synthesis of N-phenylphthalimide (10). (5 mmol phthalimide and 6
mmol PhI in 8 mL toluene). Pyridine and carbonate/L play different roles while
carbonate and L play similar roles).
4
4
[
29]
difference in the thermodynamic data
(8.29 kcal/mol), we
carefully assumed that LCuNHPht and LCuNHPy are also
formed in the solution.
When L
noticeable yield is obtained. However, comparatively
substantial yield is obtained when it is used in stoichiometric
amount as a substitute of K CO in the reaction (L as ligand and
as substitute of base) (Scheme 3). Aforesaid reactions
completely rest in presence of organic base pyridine while 13%
yield was recorded when second ligand L was also added
4 2 3
is used as ligand in presence of base (K CO ), no
a
In order to investigate the formation of 4c, solution of 3c was
o
refluxed with 1a (PhI) in absence of base at 100 C and in situ
2
3
2
electronic spectrum of the resulting solution was recorded at
room temperature in toluene. Electronic spectrum indicates the
L
4
III
formation of pentacoordinated (square pyramidal) Cu species
4
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601174
FULL PAPER
[
30,31]
4
c (552, 632 and 806 nm) (Figure S3).
The observed
role up to the process of OA and thus as per second point of
hypothesis their role starts after this step. Later role of these
species may be as a second ligand.
spectral pattern is corresponding to OA of PhI to 3c (Scheme 5).
Similar results were also recorded for 4a species by OA of PhI to
2
L CuNPht (3a) (Figure S4 and table 3) and spectrum has very
0
0
0
0
.10
.08
.06
.04
2+
(a)
(890)
74
good agreement with reported pentacoordinated complex of Ni
8
8
[30a,b]
(
3d ).
It was believed without electronic spectroscopic
studies that 4b also form in OA. Apparently, the actual species
that can react with PhI are 3a, 3b and 3c.
(1020)
(
748)
910
1
014 (1065)
058
714, 730
1
0
Ph
500
600
700
800
WL /nm
900
1000
1100
NH
III O
1
+
Cu
O
1
+
2-
CO3
C
O
NHNu
NH
NH
CuNHNu
NH
NH IPh
0.1
0.0
I
PhI
OA
III
NHNu=
(b)
Cu
5b or 5c
P
O
NH NHNu
O
3
4
-
1-
Ph
or
NH
3
b or 3c
4b or 4c
NH
NH
NH
-0.1
-0.2
III O
O-
1012
(1001)
Cu
2
b
2c
(PyNH)
O
757
(796)
P
O
O
O
(PhtNH)
NHPy
-
0.3
0.4
1225
11
1303 (1247)
-
2
-
Bidentate CO
Free Ionic CO
III
1590
1580)
1
395
3 2-
Scheme 5. Formation of Cu intermediates through oxidative addition and
ligation of carbonate/phosphate (nucleophile NHNu is PhtNH in 3b/4b and
PyNH in 3c/4c).
-0.5
0.6
(
3
-
1600
1400
1200
WN/cm
1000
800
-1
(
.00
c)
0
There is no product corresponding to OA if the aforesaid
6
44*
reactions are carried out in absence of ligand and K
C temperature. However, more reducible 1b undergoes OA with
CuI in toluene, even in absence of ligand and base
2
CO
3
at 100
-0.05
-0.10
-0.15
-0.20
864*
(843)
(
666)
0
1280*
(
1280)
1
422*
(
tetracoordinated species 8) (Scheme 6). When ligand L
2
is
added in solution of 8, expected pentacoordinated Cu complex
is formed. Experimental and theoretical spectra validate the
formation of a square planar 8 (electronic absorptions at 686,
(
1436)
III
3-
PO
4
9
1008, 977
*
PyN
H
O-P
(
1036, 941)
[
32,33]
1600
1400
1200
WN/cm
1000
800
600
7
88, 854 and 978 nm)
and square pyramidal 9 (electronic
-
1
[
30,31]
absorptions at 612, 844 and 1048 nm)
(Figure S12). These
observations indicate more reactivity and thus effective
participation of relatively electron deficient 1b in OA. This is
parallel to the reactivity trends of Ar-X species in OA-RE
pathway. Again it is important to reiterate that PhI does not
undergo OA in similar chemical environment even after 4 hrs.
Figure 3. a. Electronic spectrum of species 5c in toluene. b. FTIR spectrum of
bidentate carbonate in species 5c (Spectrum of solution of 4c subtracted from
the spectrum of solution of 5c) and c. FTIR spectrum of bidentate phosphate in
species 11 (Spectrum of solution of 4c subtracted from the spectrum of
solution of 11). Calculated values are in bracket. (See supporting information
for more details).
I
NHPht
NH I
Cu
Reaction progress was monitored spectroscopically for proper
validation of the hypothesis that carbonate may act as a second
ligand. On addition of base (K CO ) in solution of 4b/4c,
2 3
electronic spectrum is significantly changed and we assumed
that this is due to ligation of carbonate in the resulting species.
Therefore, all the possible species were considered and spectra
were computed on optimized geometries by DFT. We finally
concluded on the basis of thermodynamic data (Figure 6) and
similarity in the calculated and experimental spectra that these
species are 5b/5c (octahedral and carbonate ligated) (Scheme
Ph
0.3eq L2
toluene
NHPht
Cu
I
+
NHPht
I
o
100 C
O N
2
NH Ph
CuI 20 mol%
1b
8
9
Ph= p-nitrophenyl moiety
Scheme 6. OA of 1b with CuI in absence of base and ligand. On addition of
ligand, species 8 turns into 9. Experimental conditions: 1.2 eq (6 mmol) 1b and
eq (5 mmol) phthalimide. Refluxed under N for 0.5 h.
2
1
5).
III
From the above discussion it can be concluded that Cu
In situ electronic absorption studies of 5b and 5c (in DMSO and
toluene) shows transitions corresponding to the octahedral Cu
III
intermediate species are formed in these reactions even in
absence of base. This also indicates that widely considered
bases, carbonate/phosphate are not having any unavoidable
complex (Figure 3a and S6). The 5b species in toluene shows
3
3
3
absorptions at ν
1
=1058, 1020 ( A2g
2
T2g), ν = 910, 874 ( A2g
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601174
FULL PAPER
3
3
3
3-
T
1g, F) and ν
3
=474 nm ( A2g
T
1g, P). A doublet (714, 730 nm)
higher frequencies than ionic PO
4
. Broadening of N-H bands
3
1
[34,35]
is attributed to forbidden transition A2g
the position of absorption bands was observed on exchange of
b with 2c. However, in case of 5a (Scheme 8), spectrum greatly
differs in position of bands (478, 578, 662, 726 and 880 nm)
E
1g
.
No shifting in
(PyNH) due to hydrogen bonding with phosphate is an additional
evidence of phosphate ligation. Optimized geometry of 11 also
confirms the above observations (Figure 4).
2
[
31]
(
Figure S5).
-
1
Table 4. C=O and N-H IR absorbance bands (cm ) in 5c, 6 species and
reported values for carbonate complexes in literature.
Table 3. Experimental and DFT calculated electronic absorption bands
WL/nm).
(
a
[
2 3
Co(en) CO ]I
Intermediate 5c
DFT calculated bands for 5c
Species
Experimental
DFT Calculated
1
1
1
7
565
278,
034-1055
56
1590
1225
1012
757
1580
1247
1001
796
4a
4c
5a
5c
492, 660, 784, 928
477, 661, 793, 948
551, 651, 823, 924
478, 578, 662, 726, 880
748, 890, 1020, 1065
552, 632, 806
b
3-
b
-
b
+
[
Co(CO
3 3
) ]
[Co(NH
3
) (CO
2
3
2
) ]
[Co(NH
3 5 3
) CO ]
446, 564, 630, 726, 884
(714, 730), (874, 910), 1014,
C=O 1590
C=O 1623
Complex 6 [CuL
N-H 3300, N-H--O 3190-3115
1058
2
3
CO ]
8
9
416, 686, 788, 854, 978
612, 724, 844, 958, 1050
664, (876, 910), 984, 1008
III
412, 687, 785, 888, 970
Bond
C=O
N-H
Experimental
1627
DFT Calculated bands for species 6
634, 734, 849,1048
-
1638
3377
3061
3329
3166
a*
N-H--O
Reported by [a] Gatehouse et al. (Ref. 35) [b] V S Sastri (Ref. 42).
a* Spectrum of Cu species in reaction of CuI, NHPht, 2-Iodobenzoic acid
1c), carbonate and L in toluene. See the supporting information for spectra.
(
2
This is important to note that splitting of spectral bands is
III
observed. It is strong evidence of distortion in octahedral Cu
complexes. Similar results have also been noticed for coupling
reaction of PhI and 2b in presence of K
Absorbance is observed at ν =1008, 984, ν
=664 nm when 1c is an aryl halide in reaction (Figure S10).
3
4
PO in DMSO.
1
2
=910, 876 and
ν
3
Chlorobenzene does not respond even after 4 hr in similar set of
condition. (Figure S15).
In order to strengthen the second point of hypothesis, in situ
FTIR spectra were also recorded. These studies were
conducted with nucleophile 2c (PyNH) due to its liquid nature. In
the in situ FTIR spectrum of the complex 5c in toluene,
vibrational bands are observed at 1590 (C=O sym str.), 1303 (C-
Figure 4. Optimized geometries of 5c and 11 (O
●
N
●
Cu
●
C
●
H● and
o
P
●
). Cu-N/C bond lengths (A ) and elongated coordination of PyNH with Cu.
-1
Cu
O asym str.) 1012 (C-O sym str.) and 757 (C=O bending) cm
(
0.031)
(
-0.006)
[
36,37]
C
C
(
Table 4, figure 3b and figure S7).
Above spectral pattern
C
8 (0.027)
9 (0.133)
CO32-
.088 Ao
3
-1
-1
Cu 3.245 Ao
NHPy
has a difference of more than 250 cm (Δν=287 cm ) between
the higher position bands appearing at 1590 and 1303 cm .
Difference of 287 cm in the positions of these bands strongly
reinforces the hypothesis that carbonate is acting as bidentate
Thus, these studies strongly
indicate the formation of carbonate ligated octahedral Cu
intermediates 5c (Figure 4).
(CO3)Cu
(-0.191)
-1
NHPy
(
-0.185)
NO2
-1
5c
4
c
[
38]
ligand in these complexes.
III
Scheme 7. DFT calculated Mulliken charges and spatial distance between
coupling atoms (C and N) in species 4c, 5c, 8 and 9. Carbonate ligation makes
carbon atom (phenyl ring) positive and N atom (amide) more negative in 5c.
Coupling C atoms of species 8 and 9 have positive charge even without
coordination of carbonate and ligand. Mulliken charges are in bracket.
2
-
This seems appropriate to mention that bands of CO
3
in 5c
-
1
species (in DMSO) are observed at 1598 and 794 cm and
unfortunately, other bands overlap with solvent bands. FTIR
studies for corresponding phosphate ligated species 11 were
also conducted in toluene and similar results were obtained
There is no sufficient literature to support the electronic spectra
-1 [39]
(
1008 and 977 cm )
(Figure 3c and S8).These band are at
III
of octahedral Cu species and therefore, it is complimented with
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601174
FULL PAPER
DFT studies. The fully optimized geometry of 5c validates the
we can conclude that base also helps in preventing the
protonation of these species by HI which is generated in reaction.
formation of
a
distorted octahedral complex, elongated
o
coordination with 2c (bond length 2.414 A ). Similar pattern of
elongated coordination is reported for Cu complexes.
III
[40]
1
-
O
0
Ph
I
III
Cu
NH
Moreover, we also found a justification of carbonate ligation from
DFT. Carbonate plays two roles, it reduces spatial distance and
increases favorable charge on coupling atoms (C of phenyl ring
get positive charge and N of nucleophile get more negative
charge) (Scheme 7). It is interesting that coupling C atom of p-
nitrophenyl ring in species 8 and 9 carry positive charges even
without coordination of ligand and carbonate. On the basis of
Mulliken charges it appears that RE is much facile in the present
case. This strongly indicates that RE can easily take place
without deprotonation of nucleophile. This is also supported by
the theoretical thermodynamic data of 5c.The free energy barrier
NH
CuI
NH
NH
CO32-
III O
I
PhI
L2
Toluene
Ph
Cu
NK +
C O
NPht
NH NPht
NH
O
2
a
O
(
PhtNK)
4a
5a
III
Scheme 8. Formation of Cu intermediates through oxidative addition and
ligation of carbonate.
Uv-vis spectroscopic studies were also carried out with
deprotonated nucleophile 2a (PhtNK) and observations confirm
the formation of 4a and 5a in the similar set of experimental
(
+1.07 kcal/mol) of transition state (TS) of RE for 5c (without
[
29,31]
deprotonation) is significantly lower than earlier reported by Guo
condition.
However, 4a and 5a poorly turn into the final
[
41]
[23]
et al. (+2.8 kcal/mol) and Taillefer et al. (+4.37 kcal/mol) for
deprotonated nucleophile ligated species (Figure 6).
product (0-5% yield) and yield is regardless of the amount of
base (0-100 mol% K CO ) used (Figure S13). It may be due to
2
3
2
-
[27]
In view of the above it can be concluded that CO
3
facilitates the
formation of inactive multi-ligated species.
RE step by lowering activation energy of the TS. It is worth to
mention that deprotonation is not essential up to RE. As
discussed above approach of carbonate moiety to 4c/4b
facilitates reductive elimination and therefore, there should be
important role of steric hindrance in these reactions. It is
reported in the literature that o-substituted (e.g. o-methyl) aryl
halides give poor yield comparative to the corresponding p-
substituted aryl halides while presence of coordinating group
To ensure the formation of 6 (Scheme 9), the crude product of
the reaction of 1a and 2b was washed with acetone (to remove
un-reacted 2b and product 10) and FTIR spectrum was recorded.
Appearance of strong absorption bands at 3329, 3166 (doublet,
-1
NH), 1627, 1390, and 781cm (C=O) indicates presence of N, N
[42]
I
donor L and bidentate carbonato in the Cu complex 6 (Table 4
2
and figure S14). Absorption of C=O at higher energy, denotes
str
[
43]
-1
the covalent interaction
and Δν=237 cm indicates the
(
e.g. o-COOH) facilitates the reaction.
bidentate or bridged nature.
Ph-I
0.8
0.7
0.6
0.5
0.4
0.3
0.2
Base (pyridine)
Acid (n-butanoic acid)
Natural decay
OA
I
3b or 3c
NH
NH
Ph
4
b or 4c
NuHN CuI
CuIII
NH
NH
NHNu
CO32-
CO 2
-
3
I-
NHNu
NH O
CuI
O
Ph
O
C
NH
CuIII
6
C
O
O
5b or 5c
NH
NH O NHNu
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
RE
Time/min
Base/-H+
NuNH=amides/imidess
Figure 5. Relative stability of 5c species in basic and acidic medium (under N
2
and at room temperature in toluene).
Ph-NNu
Scheme 9. Proposed catalytic cycle for C-N coupling of aryl iodide and
amides/imides.
For further strengthening the role of carbonate, effect of acid and
base were studied on species 5c. In this order 0.1 mL pyridine
and n-butanoic acid were slowly injected in 3 mL solutions
-1
In addition, one NH band occurs at lower energy 3166 cm due
to intramolecular hydrogen bonding (N-H--O) between N-H of L
(
toluene) of 5c under nitrogen. Absorbance peak 910 nm was
2
selected as a reference. On addition of acid, 5c species
disappear immediately while base holds the species for longer
period (Figure 5). We assumed that in presence of acid 5c has
disappeared due to protonation of ligand/nucleophile
neutralization of ligated base. On the basis of above discussion
[
43]
and oxygen of carbonate. These spectral bands also have
[
29]
good agreement with calculated frequencies.
[
15]
The displacement of carbonate moiety by nucleophile is not
much energetically favored (+59.53 kcal/mol) (Figure 6).
or
I
Secondly, tetracoordinated Cu complex 6 has poor solubility
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601174
FULL PAPER
and thus probably, species 6 is responsible for reported off cycle
be significantly reduced by introducing of second ligand (a
substitute of carbonato ligand). We believe that present studies
will serve more insight into the mechanism. Therefore, scope is
open to investigate the C-O/C-S coupling and better alternatives
of carbonate and phosphate in term of their behavior as second
ligand. We have successfully isolated one of the intermediate
and analysis is underway.
[
22]
parking
of copper catalyst after a few hours of run of the
catalytic cycle. Therefore, higher loading of CuI is required.
Above discussion also gives an insight that presence of higher
2
-
3-
amount of the CO
3
/PO
4
plays negative role in the catalytic
I
cycle by remaining coordinated with Cu . Most recently, similar
negative effect of HPO
Nickel-mediated reactions. In this situation greater amount of
ligand or nucleophile is required to generate active species
3b/3c) and for mass transfer
2
-
4
is reported by T B Lu and co-workers in
[
44]
[
45]
(
(species 6 is insoluble in Experimental and computational Section
toluene).
In all cases until not specified 5 mmol phthalimide or 2-pyrrolidone, 1.2
eq (6 mmol) arylhalide, 5 mol% CuI, base and 0.3 eq ligand refluxed
o
under N
2
in 8 mL toluene at 100 C. For in situ studies, 40 mL solvent
was used and chemical equivalents were remained same. See the
supporting information for detail.
+1.07
TS
[
47]
5c
RI-J auxiliary basis sets in combination with Aldrich
basis sets def2-
Becke-Johnson damping
and COSMO solvation model (toluene) were implemented using the
[
48]
[49]
XVP def2-XVP/J,
dispersion correction
+
7.74
TS
4c
[
50]
+
47.98
ORCA quantum chemistry program 3.03.
Accordingly, geometries
were fully optimized without any constrain. Free energies, numerical
frequencies and electronic spectra were calculated for species.
-
116.32
3c
[
29]
+66.56
+5
9
.
53
NH
H
Cu I 0
NH
N
O
6
Cu
C O
O
Relative free energy
kcal/mol)
PyNH Nucleophile
PhI Iodobenzene
NH
(
Acknowledgements
Figure 6. DFT calculated relative free energy of various intermediates in
catalytic cycle in toluene at 298 K (free energy of L CuI is taken 0). Geometry
and bond lengths of transition state (TS) for reductive elimination (O
Cu and H●).
Authors are thankful to the UGC, New Delhi for financial support
and grateful to the VGEC, Ahmedabad and RIE, NCERT, Ajmer
for providing necessary facility.
2
● N●
● C●
Keywords: cross-coupling • copper • reaction mechanisms •
homogeneous catalysis• density functional calculation
On the basis of this conclusion, one can easily explain the
[
46]
results of kinetic studies reported by Buchwald et al.
Possibly,
[
1]
F. Ullmann, J. Bielecki, Berichte Dtsch. Chem. Ges. 1901, 34, 2174–
185.
soluble organic bidentate bases are efficient alternatives due to
2
[
22]
mass transfer (better solubility).
(
We believe that carbonate
[
[
2]
3]
I. Goldberg, Berichte Dtsch. Chem. Ges. 1906, 39, 1691–1692.
Jean-François Marcoux, Sven Doye, S. L. Buchwald, J. Am. Chem. Soc.
1997, 119, 10539–10540.
[
23]
Taillefer et al.) is second ligand in C-O bond formation.
Sometimes initially, on addition of ligand Cu species (630 nm)
II
[
[
4]
5]
a) H.-J. Cristau, P. P. Cellier, J.-F. Spindler, M. Taillefer, Chem. – Eur. J.
appear and these species disappears in presence of nucleophile
2004, 10, 5607–5622; b) M. Taillefer, H.-J. Cristau, P. P. Cellier, J.-F.
II
2
b/3b (Figure S11).These Cu species are reduced by
Spindler, patents Fr2833947-WO0353225 (Pr. Nb. Fr 2001 16547; c) M.
Taillefer, H.-J. Cristau, P. P. Cellier, J.-F. Spindler, A. Ouali,
Fr2840303-WO03101966 (Pr.Nb.Fr200206717); d) A. Ouali, M.
[
16]
nucleophile.
Taillefer, J.-F. Spindler, A. Jutand, Organometallics 2007, 26, 65–74.
H. Kaddouri, V. Vicente, A. Ouali, F. Ouazzani, M. Taillefer, Angew.
Chem. Int. Ed. 2009, 48, 333–336.
Conclusions
[
[
6]
7]
D. S. Surry, S. L. Buchwald, Chem Sci 2010, 1, 13–31.
a) F. Monnier, M. Taillefer, Angew. Chem. Int. Ed. 2008, 47, 3096–3099.
b) F. Monnier, M. Taillefer, Angew. Chem. Int. Ed. 2009, 48, 6954–6971.
J. Bariwal, E. Van der Eycken, Chem Soc Rev 2013, 42, 9283–9303.
A. Casitas, X. Ribas, Chem Sci 2013, 4, 2301–2318.
Mechanism is proposed for aryl iodide and amide/imide C-N
bond formation on the basis of experimental and DFT studies.
The different steps of coupling reaction involve ligation of
nucleophile (without deprotonation), well defined roles of base
[8]
[
[
9]
10] C. Sambiagio, S. P. Marsden, A. J. Blacker, P. C. McGowan, Chem Soc
Rev 2014, 43, 3525–3550.
2
-
3-
III
(
CO
3
/PO
4
) as a bidentate second ligand, octahedral Cu
[11] E. Sperotto, G. P. M. van Klink, G. van Koten, J. G. de Vries, Dalton
Trans 2010, 39, 10338–10351.
I
intermediates and tetracoordinated Cu complexes are reported.
Most importantly, the triple role of CO
[
12] A. J. Paine, J. Am. Chem. Soc. 1987, 109, 1496–1502.
2
-
3-
3
/PO
4
as proton
[13] M. Mansour, R. Giacovazzi, A. Ouali, M. Taillefer, A. Jutand, Chem
Commun 2008, 6051–6053.
acceptor, ligand and inhibitor in reaction is defined. The present
study provides the direct spectroscopic (Uv-vis and IR)
[
[
14] L. M. Huffman, S. S. Stahl, J. Am. Chem. Soc. 2008, 130, 9196–9197.
15] A. Casitas, A. E. King, T. Parella, M. Costas, S. S. Stahl, X. Ribas,
Chem Sci 2010, 1, 326–330.
III
evidences of Cu species and OA-RE pathway (Scheme 9).
[
16] G. Lefèvre, G. Franc, A. Tlili, C. Adamo, M. Taillefer, I. Ciofini, A. Jutand,
Organometallics 2012, 31, 7694–7707.
Herein, first time this work has defined broader role of bases.
Additionally, we argue that amounts of ligand, CuI and base can
[17] T. Cohen, I. Cristea, J. Am. Chem. Soc. 1976, 98, 748–753.
18] R. Giri, J. F. Hartwig, J. Am. Chem. Soc. 2010, 132, 15860–15863.
[
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601174
FULL PAPER
[
[
[
[
[
[
[
[
19] G. Lefèvre, G. Franc, C. Adamo, A. Jutand, I. Ciofini, Organometallics
012, 31, 914–920.
20] G. O. Jones, P. Liu, K. N. Houk, S. L. Buchwald, J. Am. Chem. Soc.
010, 132, 6205–6213.
21] M. Rovira, M. Soler, I. Güell, M.-Z. Wang, L. Gómez, X. Ribas, J. Org.
Chem. 2016, 81, 7315–7325.
22] S. Sung, D. Sale, D. C. Braddock, A. Armstrong, C. Brennan, R. P.
Davies, ACS Catal. 2016, 6, 3965–3974.
23] G. Lefevre, A. Tlili, M. Taillefer, C. Adamo, I. Ciofini, A. Jutand, Dalton
Trans 2013, 42, 5348–5354.
2
2
24] A. Ouali, J.-F. Spindler, A. Jutand, M. Taillefer, Adv. Synth. Catal. 2007,
349, 1906–1916.
25] C. Sambiagio, R. H. Munday, A. John Blacker, S. P. Marsden, P. C.
McGowan, RSC Adv 2016, 6, 70025–70032.
26] a) C.-T. Yang, Y. Fu, Y.-B. Huang, J. Yi, Q.-X. Guo, L. Liu, Angew.
Chem. Int. Ed. 2009, 48, 7398–7401. b) D. Maiti, Chem Commun 2011,
47, 8340–8342.
[
27] J. W. Tye, Z. Weng, A. M. Johns, C. D. Incarvito, J. F. Hartwig, J. Am.
Chem. Soc. 2008, 130, 9971–9983.
[
[
[
28] R. G. Pearson, J. Am. Chem. Soc. 1963, 85, 3533–3539.
29] See the supporting information for more detail.
30] a) S. Yamada, K. Iwasaki, A. Takeuchi, Inorganica Chim. Acta 1968, 2,
3
1
91–394. b) M. J. D’Aniello, E. K. Barefield, J. Am. Chem. Soc. 1976, 98,
610–1611.
[
[
31] See table 3.
32] Hiromitsu Maeda, Yuichi Ishikawa, Tomoyuki Matsuda, Atsuhiro Osuka,
and Hiroyuki Furuta, J. Am. Chem. Soc. 2003, 125, 11822–11823.
33] J. Hanss, A. Beckmann, H.-J. Krüger, Eur. J. Inorg. Chem. 1999, 1999,
[
[
[
[
1
63–172.
34] S. Kumar, M. Munjal, J. Singh, R. Gupta, Eur. J. Inorg. Chem. 2014,
014, 4957–4965.
2
35] C. Krebs, T. Glaser, E. Bill, T. Weyhermüller, W. Meyer-Klaucke, K.
Wieghardt, Angew. Chem. Int. Ed. 1999, 38, 359–361.
36] B. M. Gatehouse, S. E. Livingstone, R. S. Nyholm, J Chem Soc 1958,
3137–3142.
[
[
37] A. K. Sengupta, A. K. Nandi, J. Inorg. Nucl. Chem. 1974, 36, 2479–2484.
38] A. M. Greenaway, T. P. Dasgupta, K. C. Koshy, G. G. Sadler,
Spectrochim. Acta Part Mol. Spectrosc. 1986, 42, 949–954.
39] X. Zhu, S. D. Alexandratos, Chem. Eng. Sci. 2015, 127, 126–132.
40] H. Zhang, B. Yao, L. Zhao, D.-X. Wang, B.-Q. Xu, M.-X. Wang, J. Am.
Chem. Soc. 2014, 136, 6326–6332.
[
[
[
[
41] Song-Lin Zhang, Lei Liu , Yao Fu, and Qing-Xiang Guo,
Organometallics 2007, 26, 4546–4554.
42] R. A. Altman, A. M. Hyde, X. Huang, S. L. Buchwald, J. Am. Chem. Soc.
2008, 130, 9613–9620.
[
[
43] V. S. Sastri, Inorganica Chim. Acta 1972, 6, 264–266.
44] J.-W. Wang, X.-Q. Zhang, H.-H. Huang, T.-B. Lu, ChemCatChem 2016,
8
, 3287–3293.
45] P.-F. Larsson, C.-J. Wallentin, P.-O. Norrby, ChemCatChem 2014, 6,
277–1282.
46] E. R. Strieter, B. Bhayana, S. L. Buchwald, J. Am. Chem. Soc. 2009,
31, 78–88.
[
[
1
1
[
[
[
47] A. Schäfer, H. Horn, R. Ahlrichs, J. Chem. Phys. 1992, 97, 2571–2577.
48] F. Weigend, Phys Chem Chem Phys 2006, 8, 1057–1065.
49] S. Grimme, S. Ehrlich, L. Goerigk, J. Comput. Chem. 2011, 32, 1456–
1465.
[
50] F. Neese, Wiley Interdiscip. Rev. Comput. Mol. Sci. 2012, 2, 73–78.
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601174
FULL PAPER
Layout 2:
FULL PAPER
DFT studies
O
Kamlesh K Gurjar, Rajendra K Sharma*
NH
O
O
N
In situ
O
PhI
CO32-
C
O
Page No. – Page No.
Uv-vis
Cu
FTIR
CuI
NH
NH
Cu
NH
O
dmeda
NH
C
O
Mechanistic Studies of Ullmann Type
C-N Coupling Reactions: Carbonate
Ligated Cu(III) Intermediate
PyNH
O
Inactive
Direct observation of
Cu(III) and ligation of CO32-
(
2-Pyrrolidone)
Triple role of carbonate: Mechanism has been proposed for Ullmann C-N
coupling on the basis of direct observation of Cu(III) intermediates in Uv-vis and
2
3
-
FTIR studies. CO
is a second ligand and facilitates the reductive elimination.
Actual intermediate species are distorted octahedral Cu(III) complexes.
Experimental results are complimented by DFT.
This article is protected by copyright. All rights reserved.