Mechanistic study of chemoselectivity in ni-catalyzed coupling reactions between azoles and aryl carboxylates

Article abstract of DOI:10.1021/ja4127455

Itami et al. recently reported the C-O electrophile-controlled chemoselectivity of Ni-catalyzed coupling reactions between azoles and esters: the decarbonylative C-H coupling product was generated with the aryl ester substrates, while C-H/C-O coupling pro

Full text of DOI:10.1021/ja4127455

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Article
pubs.acs.org/JACS
Mechanistic Study of Chemoselectivity in Ni-Catalyzed Coupling
Reactions between Azoles and Aryl Carboxylates
Qianqian Lu,†,§ Haizhu Yu,‡,§ and Yao Fu*,†
†Department of Chemistry, University of Science and Technology of China, Hefei 230026, PR China
‡Department of Polymer Science and Engineering, University of Science and Technology Beijing, Beijing 100083, PR China
S
* Supporting Information
ABSTRACT: Itami et al. recently reported the C−O electro-
phile-controlled chemoselectivity of Ni-catalyzed coupling
reactions between azoles and esters: the decarbonylative C−
H coupling product was generated with the aryl ester
substrates, while C−H/C−O coupling product was generated
with the phenol derivative substrates (such as phenyl pivalate).
With the aid of DFT calculations (M06L/6-311+G(2d,p)-
SDD//B3LYP/6-31G(d)-LANL2DZ), the present study sys-
tematically investigated the mechanism of the aforementioned
chemoselective reactions. The decarbonylative C−H coupling
mechanism involves oxidative addition of C(acyl)−O bond,
base-promoted C−H activation of azole, CO migration, and
reductive elimination steps (C−H/Decar mechanism). This
mechanism is partially different from Itami’s previous proposal
(Decar/C−H mechanism) because the C−H activation step is unlikely to occur after the CO migration step. Meanwhile, C−H/
C−O coupling reaction proceeds through oxidative addition of C(phenyl)−O bond, base-promoted C−H activation, and
reductive elimination steps. It was found that the C−O electrophile significantly influences the overall energy demand of the
decarbonylative C−H coupling mechanism, because the rate-determining step (i.e., CO migration) is sensitive to the steric effect
of the acyl substituent. In contrast, in the C−H/C−O coupling mechanism, the release of the carboxylates occurs before the rate-
determining step (i.e., base-promoted C−H activation), and thus the overall energy demand is almost independent of the acyl
substituent. Accordingly, the decarbonylative C−H coupling product is favored for less-bulky group substituted C−O
electrophiles (such as aryl ester), while C−H/C−O coupling product is predominant for bulky group substituted C−O
electrophiles (such as phenyl pivalate).
bonds. For example, Crabtree et al. described the Pd-catalyzed
decarboxylative C−H coupling between (hetero)arenes and
aryl carboxylic acids.9 Yu et al. reported the Rh(I)-catalyzed
decarbonylative C−H coupling of (hetero)arenes with acid
chlorides.10 Recently Itami et al. developed the Ni(cod)2/dcype
catalyzed coupling reactions of aryl esters and azoles, and the
decarbonylative C−H coupling products were finally obtained
(Scheme 1a).11 Interestingly, in a similar catalytic system
(Scheme 1b), a distinct C−H/C−O coupling reaction occurred
when using the phenol derivatives (such as triflates, carbamates,
tosylates, and sulfamates) as electrophiles.12 Itami’s reactions
provide an efficient way to synthesize bioactive compounds
(e.g., muscoride A) and also reveal an interesting chemo-
selectivity by using different C−O electrophiles.
1. INTRODUCTION
Bi(hetero)aryls widely exist in natural substances,1 pharma-
ceutical molecules2 and materials,3 hence the preparation of
bi(hetero)aryls is of great importance. In the past decades,
transition-metal-catalyzed cross-coupling reactions between
(hetero)aryl halides and (hetero)arylmetallic reagents have
been widely used to synthesize bi(hetero)aryls.4 Nevertheless,
the development of more efficient and green catalytic systems
remains requisite, due to the shortcomings of the cross-
coupling reactions (such as the high cost of organometallic
substrates). In 2006, a pioneering study describing a transition-
metal-catalyzed decarboxylative coupling between aryl carbox-
ylic acids and aryl halides was reported by Goossen and co-
workers.5 Then much effort has been paid to the low-cost and
readily accessible aroyl compounds, from which arylmetallic
reagents could be generated through decarboxylation6 or
decarbonylation process.7 More importantly, with the develop-
ment of C−H activation/functionalization,8 transition-metal-
catalyzed decarboxylative or decarbonylative coupling reactions
between aroyl derivatives and simple (hetero)arenes become an
effective method to construct C(heteroaryl)−C(heteroaryl)
As to the mechanism of the fascinating chemoselectivity
shown in Scheme 1, Itami et al. proposed that the different
mechanisms might be responsible for different coupling
reactions. The decarbonylative C−H coupling reaction was
Received: December 16, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja4127455 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Journal of the American Chemical Society
Article
most feasible mechanism for decarbonylative C−H coupling
reaction involves the oxidative addition of C(acyl)−O bond,
C−H activation of azole, CO migration, and reductive
elimination steps (C−H/Decar mechanism). On the other
hand, the C−H/C−O coupling reaction proceeds through the
oxidative addition of C(phenyl)−O bond, C−H activation of
azole (i.e., the rate-determining step), and reductive elimination
steps (in accord with the Itami’s previous proposal14 in Figure
1). The steric hindrance of the C−O electrophile significantly
suppresses the CO migration step in the C−H/Decar
mechanism, and results in the relative priority of the C−H/
C−O coupling reaction. In other words, decarbonylative C−H
coupling product is favored for less-bulky group substituted C−
O electrophiles (such as aryl ester), while C−H/C−O coupling
product is predominant for bulky group substituted C−O
electrophiles (such as phenyl pivalate).
Scheme 1. Ni-Catalyzed Coupling Reactions between Azoles
and Different Types of Esters
suggested to occur via the oxidative addition of C(acyl)−O
bond, CO migration, azole C−H nickelation, reductive
elimination, and CO extrusion steps (Decar/C−H mechanism,
Figure 1).11 The C−H/C−O coupling reaction was proposed
to begin with the oxidative addition of C(phenyl)−O bond,
following with the C−H nickelation and reductive elimination
steps (C−H/C−O mechanism, Figure 1).12 Meanwhile, the
azole C−H nickelation step was proposed to be unlikely the
rate-determining step of the decarbonylative C−H coupling
reaction, because no significant kinetic isotope effect (KIE) was
observed (KIE = 1.24).
Despite of the aforementioned mechanistic understandings,
many details remain ambiguous. First, the details of decarbon-
ylative C−H coupling mechanism (including the sequence of
CO migration, C−H activation process, and the effects of base)
are unclear. Second, the oxidative addition of different C−O
bonds (C(acyl)−O or C(phenyl)−O bond) causes different
mechanisms, whereas the driving force for oxidative addition of
the stronger C(phenyl)−O bond (relative to C(acyl)−O
bond)13 in the C−H/C−O coupling reaction remains
unknown. More importantly, the reasons for the C−O
electrophile-mediated chemoselectivity need to be clarified.
To settle the above problems, we conducted a mechanistic
study on Ni-catalyzed coupling reaction between azoles and
aryl carboxylates (Scheme 1) using DFT method. The
calculation results indicate that Itami’s previous proposal
(Decar/C−H mechanism in Figure 1) is unfavorable, because
the Ni(II) center generated from the CO migration step is
electron deficient and disfavors the subsequent C−H activation
process. Alternatively, a prior C−H activation (of azole) before
the CO migration step can avoid this problem. As a result, the
2. COMPUTATIONAL METHODS
All the DFT calculations were performed with Gaussian09 package.15
The geometry optimizations of all species in this study were performed
using B3LYP method.16 The 6-31G(d) basis set was used for all the
atoms except Ni and K, for which the LANL2DZ basis set17 was
applied (BSI). Frequency calculations at the B3LYP/BSI level of
theory were carried out to characterize each stationary point
(minimum or transition state) and to obtain the thermodynamic
corrections to Gibbs free energy. Intrinsic reaction coordinate (IRC)18
was calculated to confirm the connection between the transition state
and the right reactant/product. Solvent effect was considered by
employing the single-point calculations on the gas-phase optimized
geometries with SMD model19 and 1,4-dioxane solvent. All the single-
point calculations were performed with M06-L/BSII level20 (BSII
designates the combination of 6-311+G(2d,p) for C, H, O, P, S, N and
SDD21 for Ni, K). The additional polarization functions were Ni(ζ(f))
= 3.130, and K(ζ(d)) = 1.000.22 The reported Gibbs free energy in
this study was calculated by adding the gas-phase Gibbs free energy
correction with the solution-phase single-point energy.23,24
3. RESULT
In this study, the coupling reactions between benzoxazole
(azole) and phenyl thiophene-2-carboxylate (ester-1) or
phenyl pivalate (ester-2) catalyzed by Ni(cod)2/dcype were
chosen as the model reactions (eqs 1 and 2).
Figure 1. Proposed mechanisms of the Ni-catalyzed coupling reaction between aryl ester (1) and azole (2).
B
dx.doi.org/10.1021/ja4127455 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Journal of the American Chemical Society
Article
CO migration processes (such as the involvement of P-TS2a,
Figure 2) have also been examined. Nonetheless, these
possibilities have been excluded due to the high-energy
demands (please see SI for more details).
Azole C−H Nickelation. From P-Int6, benzoxazole (azole)
would participate into the catalytic cycle to carry on the C−H
nickelation step, and three possible pathways were proposed for
this process (Scheme 2).
The first C−H nickelation pathway involves the direct
activation of azole C−H bond with the intramolecular PhO− of
P-Int6 (Scheme 2a). In this pathway, the dissociation of one
arm of the diphosphine ligand and the coordination of azole
occur to form P-Int7. Subsequently, P-Int7 undergoes a four-
centered C−H activation transition state P-TS3 to produce the
Ni(II) complex P-Int8. The high energy barrier of +58.3 kcal/
mol precludes the possibility of this pathway.
As the partial dissociation of dcype causes the instability of
the Ni(II) complexes, the intermolecular PhO− assisted azole
C−H activation pathway was next studied. As shown in Scheme
2b, the coordinated complex P-Int7a is first formed through
the weak interaction between O atom (in P-Int6) and H atom
(in azole). Subsequently, the C−H activation transition state P-
TS3a occurs to generate the deprotonated product P-Int8a.
Although the energy barrier of P-TS3a (ΔG⧧ = +47.8 kcal/
mol) is ∼10 kcal/mol lower than that of P-TS3 (ΔG⧧ = +58.3
kcal/mol), it is still high for the coupling reaction to occur at
150 °C.11 Thus, the PhO− assisted C−H activation pathways
(either via intermolecular or intramolecular pathways) were
excluded.
For clarity reasons, the following discussions are divided into
two parts: the decarbonylative C−H coupling mechanism and
the C−H/C−O coupling mechanism.
3.1. Decarbonylative C−H coupling mechanism. In
this section, we took the transformation of eq 1 as an example
to discuss the detailed elementary steps. The intermediates/
transition states in catalytic cycles of eqs 1 and 2 are named as
P-Int*/P-TS* and T-Int*/T-TS*, respectively.
3.1.1. Decar/C−H Mechanism of Eq 1. Oxidative Addition
of C(acyl)−O Bond. According to the calculation results, the
coordination of the thienyl group of ester-1 to Ni(dcype)
occurs first to form the reactant−catalyst complex P-Int1
(Figure 2). Then P-Int1 goes through isomerization25 to
Considering that the base promoted C−H activation has
been previously proposed,29 we also examined the possibility of
−
the K2PO4 promoted azole C−H activation in the present
study. According to the calculation results (Scheme 2c), the
−
ligand exchange between −OPh and K2PO4 occurs first to
generate P-Int7b.30 Thereafter, the partial dissociation of
−
dcype, rearrangement of K2PO4 , and the ligation of Ar−H
occur subsequently to generate P-Int8b (all these steps are
quite facile, and the details are given in SI). After that, the C−H
bond of azole was deprotonated by K2PO4− via the six-centered
transition state P-TS3b to give P-Int9. During the trans-
formation from P-Int8b to P-TS3b, the Ni−O2 bond distance
is lengthened from 1.918 to 2.340 Å (Figure 3), the Ni−C
distance is shortened from 3.767 to 2.662 Å, and the forming
O2−H bond length is 1.012 Å in P-TS3b. All these structural
parameters are consistent with the formal concerted metal-
ation/deprotonation mechanism (CMD C−H activation
mechanism).31 What’s more important, the relative free energy
of P-TS3b (ΔG = +29.2 kcal/mol) is lower than that of P-TS3
and P-TS3a, therefore it is more plausible in describing the C−
H nickelation of azole.
Reductive Elimination. The product of C−H nickelation
(i.e., P-Int9) then dissociates K2HPO4 to generate the square-
planar intermediate P-Int10 (ΔG = +5.4 kcal/mol, Figure 4).
Then reductive elimination occurs via the four-coordinated
transition state P-TS4 to give P-Int11 (ΔG = −3.3 kcal/
mol).32,33 Finally, the ligand exchange process between P-Int11
and ester-1 occurs to release prod-1 and regenerate the Ni(0)
complex P-Int1.
Figure 2. Oxidative addition and CO migration steps involved in
Decar/C−H mechanism of eq 1.
produce the η2-coordinated complex P-Int2 (ΔG = +4.6 kcal/
mol), from which oxidative addition occurs via the three-
centered transition state P-TS1 to give the square-planar
intermediate P-Int3 (ΔG = +0.4 kcal/mol). As shown in Figure
2, the oxidative addition of C(acyl)−O bond is reversible, and
its energy barrier is +19.8 kcal/mol.
CO Migration and CO Dissociation. From P-Int3, the
dissociation of one arm of the bidentate phosphine ligand
(dcype) occurs first to provide a vacant site and facilitate the
CO migration step.26 This step gives a T-shaped intermediate
P-Int4, and the subsequent CO migration occurs via the
transition state P-TS2 (ΔG = +17.3 kcal/mol). From the
immediate product of CO migration (P-Int5), CO dissociation
occurs easily to form the bidentate ligand coordinated
intermediate P-Int6, and this step is exergonic by 7.6 kcal/
mol.27,28 Note that in the present study, some other possible
The Overall Picture of Decar/C−H Mechanism of Eq 1. For
clarity reasons, the key species in Decar/C−H mechanism of eq
1 are given in Figure 4. First, oxidative addition of C(acyl)−O
bond of ester-1 occurs with an energy barrier of +19.8 kcal/
mol. Then the formed intermediate P-Int3 undergoes CO
C
dx.doi.org/10.1021/ja4127455 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Journal of the American Chemical Society
Article
Scheme 2. Three Possible Azole C−H Nickelation Pathways on P-Int6
→ P-Int10) and reductive elimination steps occur (P-Int10 →
P-TS4 → P-Int11), and their energy barriers are +29.2 and
+19.6 kcal/mol, respectively. The C−H nickelation (with an
energy barrier of +29.2 kcal/mol) represents the rate-
determining step. Interestingly, this conclusion is inconsistent
with Itami’s previous studies (KIE = 1.24).11 In this context, we
wondered if there are other more plausible decarbonylative C−
H coupling mechanisms.
3.1.2. C−H/Decar Mechanism of Eq 1. Different from the
Decar/C−H mechanism in Section 3.1.1, the C−H nickelation
of azole might possibly occur before the CO migration step,
and this pathway is named as C−H/Decar Mechansim.
The oxidative addition step involved in C−H/Decar
mechanism is the same as that of Decar/C−H mechanism
(Figure 4). Thereafter, the C−H nickelation on the immediate
product of oxidative addition step (P-Int3) occurs favorably
Figure 3. Optimized structures of P-Int8b, P-TS3b, and P-Int9. The
bond lengths are given in Å. For clarity, the cyclohexanes of ligand are
omitted in this and the following figures.
migration−dissociation step (P-Int3 → P-TS2 → P-Int6), and
this step needs to overcome an energy barrier of +17.3 kcal/
mol. Subsequently, azole C−H nickelation (P-Int6 → P-TS3b
with the aid of K2PO4 . As shown in Figure 5, PhO− of P-Int3
−
Figure 4. Energy profiles of Decar/C−H mechanism of eq 1.
D
dx.doi.org/10.1021/ja4127455 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX