Mechanistic studies on the palladium-catalyzed cross-dehydrogenative coupling of 4-phenoxy-2-coumarins: Experimental and computational insights

Article abstract of DOI:10.1039/c8dt00842f

Delineating the mechanistic features of C-H activation in aryl-heteroaryl coupling is an important step in the design of selective, catalytic processes. Herein we use the intramolecular dehydrogenative coupling of 4-phenoxy-2-coumarins as a model study. Computational results and experimental studies suggest that CMD is operative in the cleavage of both C-H bonds, and that the heteroaryl C-H is cleaved initially.

Full text of DOI:10.1039/c8dt00842f

Dalton
Transactions
View Article Online
View Journal
PAPER
Mechanistic studies on the palladium-catalyzed
cross-dehydrogenative coupling of 4-phenoxy-2-
coumarins: experimental and computational
insights†
Cite this: DOI: 10.1039/c8dt00842f
a
b
b
Aisling M. Prendergast,
Gerard P. McGlacken
Zhihan Zhang,
Zhenyang Lin * and
a
*
Received 5th March 2018,
Accepted 6th April 2018
Delineating the mechanistic features of C–H activation in aryl-heteroaryl coupling is an important step in
the design of selective, catalytic processes. Herein we use the intramolecular dehydrogenative coupling
of 4-phenoxy-2-coumarins as a model study. Computational results and experimental studies suggest
that CMD is operative in the cleavage of both C–H bonds, and that the heteroaryl C–H is cleaved initially.
DOI: 10.1039/c8dt00842f
rsc.li/dalton
emerged on the mechanism of cross-dehydrogenative coupling
Introduction
1
9–22
(
CDC).
We chose to focus on the 2-coumarin substrate to study the
Effective routes to aryl-heteroaryl (Ar-HetAr) bonds are among
1
the most important in organic synthesis due to the abun- mechanistic features of an intramolecular dehydrogenative
1
8
dance of the Ar-HetAr moiety in natural products and pharma- coupling, for two reasons: firstly, Choi et al. elegantly unra-
2
ceuticals. The development of direct arylation protocols invol- velled a number of divergent regioselectivity outcomes using
ving aryls/heteroaryls and at least one C–H activation event is chromones and enaminones, and applied their insight to
3
4
highly sought after by academic and industrial groups.
explain the reactivity and regioselectivity in published alkeny-
2
3
24
Davies and Macgregor provided a key insight into C–H acti- lation and arylation of 4-H coumarins. In the arylation for
vation with Pd(OAc)
assistance (AMLA),
2
–7
in terms of ambiphilic metal ligand example, arene activation occurs first. This is followed by car-
describing a Pd-coordinated acetate bopalladation (rather than CMD) of the coumarin at C-3,
5
acting as proton-abstracting base. Echavarren reported that which then delivers the aryl group to C-4. In the absence of an
deprotonation mechanisms by either an internal or external available site at C-4 (our case), we wished to reveal the alterna-
base were consistent with the reactivity enhancements seen tive mechanistic pathway. Secondly, along with the related
8
,9
25–27
28
experimentally. However, analysis of a wide range of (hetero) 2-pyrones,
2-coumarins have been shown to be both a
aromatic substrates by Fagnou and co-workers led to the gener- chemically diverse (acts as an electrophile, nucleophile aryl,
alization of a key mechanistic feature, termed concerted metal- alkene, diene, ring-opens, etc.) motif and a privileged biologi-
9
–16
lation–deprotonation (CMD).
Computational studies con- cal scaffold. Thus the moiety represents a highly challenging,
tinue to provide critical insight into the modes of action invol- yet rewarding, synthon and an excellent choice to challenge a
ving CMD, and an excellent review on carboxylate-assisted C–H new methodology.
1
7
activation and functionalisation has recently emerged.
We recently reported the dehydrogenative coupling of
1
8
29
However, as stated by Choi et al., mechanistic aspects of C–H 2-coumarins (Scheme 1) and 2-pyrones, which gave interest-
bond activation involving non-aromatic substrates remain ing polycyclic ring structures. The method was applied to the
somewhat elusive and even fewer in-depth studies have
a
Analytical & Biological Chemistry Research Facility and School of Chemistry,
University College Cork, Cork, Ireland. E-mail: g.mcglacken@ucc.ie
b
Department of Chemistry, The Hong Kong University of Science and Technology,
Clear Water Bay, Kowloon, Hong Kong, People’s Republic of China.
E-mail: chzlin@ust.hk
†
Electronic supplementary information (ESI) available: Experimental details,
characterisation data, NMR spectra and additional DFT calculations. See DOI:
0.1039/c8dt00842f
1
Scheme 1 Cross-dehydrogenative coupling of 4-phenoxy-2-coumarin.
This journal is © The Royal Society of Chemistry 2018
Dalton Trans.

View Article Online
Paper
Dalton Transactions
synthesis of natural product, flemichapparin C, demonstrating
a rare application of CDC on a less-robust natural product.
Chen and Xu later arrived at similar conditions for the CDC
3
0,31
of 4-phenoxy- and 4-anilino-2-coumarins.
Gogoi and co-
workers reported the direct synthesis of coumestans via a
palladium-catalysed C–H activation/C–C and C–O bond for-
3
2
33
mation cascade invoking aryne chemistry, while Hajra and
Litinas
Scheme 3 Experimentally determined kinetic isotope effect for aryl
C–H cleavage.
3
4
have also reported conditions for the CDC of
4
-anilino-2-coumarins.
The determination of a positive KIE (approx. 3) by both of
the initial rate experiments confirms that C–H bond cleavage
on the aryl ring is involved in the rate-determining step of this
Results and discussion
We first investigated the lability of the C-3–H bond. The ease
of H/D exchange can be easily determined in intermolecular
coupling reactions by eliminating one reaction partner, and
applying the usual reaction conditions. However, in intra-
molecular cases, this is not possible. Herein, re-isolation of 1
after 5 min showed incorporation of deuterium (14%) at the
3
5
reaction mechanism.
29
Our initial tentative mechanistic proposal involved CMD
of the C-3–H of the coumarin, which indeed has been shown
to be the likely scenario, consistent with the experiments
described herein. However, we previously relied on subtle elec-
tronic factors to suggest that the second C–H cleavage was an
29
3
-position (Scheme 2), as we previously reported. In the
E E
S Ar-type activation. In an S Ar-type mechanism, the pallada-
opposite vein, when using 3-deuteriocoumarin and PivOH, the
level of H-incorporation rose from 3% (initial level) to 11%
after 5 min (ESI†). No exchange was observed at any other posi-
tion of the 2-coumarin framework. In the absence of Pd, no
H/D exchange occurred. These experiments suggest that the
C-3–H cleavage is reversible. We also reported a KIE of 1.08 for
this position, suggesting that C-3–H cleavage was not involved
in the rate-determining step of the reaction.
We then turned our attention to the determination of the
KIE for the aryl C–H bond, using a pentadeuterated analogue
5
of 1, 4-(phenoxy-d )-2-coumarin. This was complicated by the
observation of H/D exchange in the reaction, which appeared
to occur after the product was formed (ESI†).
tion is usually the rate-determining step, but the C–H bond
cleavage (to restore aromaticity) is fast. Since a positive KIE
was observed herein (Scheme 3) for the cleavage of the aryl
C–H bond, a Wheland intermediate cannot be evoked, and
therefore the mechanism of aryl C–H bond activation could
E
not involve an S Ar-type step. At this stage, we sought to delin-
eate between the potential mechanistic pathways. In terms of
which C–H bond is cleaved first, it is tempting to suggest C-3–
H, given that this C–H bond has been shown to be most labile.
However, it was not possible to rule out a reversible, non-pro-
ductive C-3–H bond cleavage. To provide some further clarity,
we turned to DFT calculations at the level of ωB97X-D/
LanL2DZ+f(Pd)/6-31G(d,p)/SMD//ωB97X-D/LanL2DZ+f(Pd)/6-
Thus, the scrambling of deuterium on the aryl ring post-
reaction prevented us from confidently using 4-(phenoxy-d )-2-
311++G(d,p)/SMD; see ESI† for more details. We chose to use
5
M. Head-Gordon’s long-range corrected (LC) hybrid density
coumarin to determine the KIE at the aryl C–H bond, as the
3
6
1
functional ωB97X-D as it has been reported to take noncova-
integrals arising from H NMR spectroscopy were unreliable.
lent interactions into consideration and it has worked well for
To circumvent this issue, 5-methoxy-4-phenoxy-2-coumarin 3
and a deuterated analogue 4 were synthesized. Since the
1
7,18,37
similar systems.
We first considered an initial CMD of the phenyl C–H, fol-
5
-methoxy group is unlikely to participate in a C–H activation
1
lowed by carbopalladation of the coumarin, in line with the
process, the H NMR signals corresponding to this group
1
8
1
observations of Choi et al. In that case, the electrophilic Pd
would approach the C-3 position (greater π-electron density).
However, the absence of a suitable destination for the phenyl
group appeared to preclude this route. Indeed, a conformation
to fulfill a carbonpalladation-type addition could not be found
allowed the progress of the reaction to be monitored using H
NMR spectroscopy. A KIE of 2.76 was determined, using the
initial rates method (Scheme 3). Finally, to confirm a positive
KIE, 5,7-dimethyl-4-phenoxy-2-coumarin, and the corres-
ponding deuterated analogue were also tested and a KIE of
(
ESI†). We next calculated for a double CMD pathway, in which
3
.25 was observed (ESI†).
aryl C–H bond activation occurred first (ESI†). However, the
palladation and C–H cleavage at C-3 was calculated to be irre-
versible with this pathway. Other mechanisms involving C–H
oxidative insertion and σ-bond metathesis have been shown to
1
8
give very high activation barriers and were not considered
here.
Finally, the most likely mechanistic route was calculated. In
the DFT calculations, we employed acetate to model pivalate
3
8
Scheme 2 H/D exchange of the 4-phenoxy-2-coumarin substrate.
for simplicity. The reaction coordinate diagram for the CDC
Dalton Trans.
This journal is © The Royal Society of Chemistry 2018

View Article Online
Dalton Transactions
Paper
−
1
of 4-phenoxy-2-coumarin (1) via sequential CMD-type C–H product 2 (ΔG = −15.6 kcal mol ) is lower in energy than IM5
(ΔG = −8.2 kcal mol− ), consistent with an irreversible aryl
1
bond activations is shown in Scheme 4.
The first step of the reaction pathway was calculated to be a C–H bond cleavage. Interestingly, the most favourable mecha-
κ –κ displacement of an acetate ligand (TS0, ΔG = 8.7 nistic route presented in Scheme 4 is the one in which the
2
1
‡
kcal mol− ) to give a π-complex IM1, which progresses through more acidic C-3–H proton (calculated pK = 34.5) is abstracted
1
a
‡
−1
the CMD transition state (TS1, ΔG = 10.2 kcal mol ) to give first, followed by the less acidic phenoxy C–H proton (calcu-
IM2. The abstraction of the proton at C-3 could be reversible lated pK = 38.5). The calculation of pK values follows the lit-
due to the similarity of the free energies of the starting erature method.
material (ΔG = 0.0 kcal mol ), IM1 (ΔG = −0.2 kcal mol
and IM2 (ΔG = −1.4 kcal mol ). This correlates well with the transfer to a Brønsted base such as tert-butoxide (tBuO ) or
a
a
3
9,40
−
1
−1
)
2
The protons on the acetic acid ligands of Pd(HOAc) could
−
1
−
−
experimentally observed ease of H/D exchange (Scheme 2). pivalate (PivO ). One or more silver species could act as an in-
Additionally, the initial activation of the C-3–H bond of the soluble proton sink, similar to the proposed role of the carbon-
4
1
2
-coumarin is not turnover-limiting, which is consistent with ate bases proposed by Fagnou and co-workers. Finally, Pd(0)
the observed KIE of 1.08 for the isotopic substitution of this is likely oxidised to Pd(II) by Ag O.
2
bond. From IM2, a π-complex (IM3) is formed via TSIM2–3. IM3
The true nature of the active catalytic and deprotonating
‡
−1
could then interconvert to IM3–2 (ΔG = −8.5 kcal mol ). species are not fully understood. It is difficult to study the role
−1
IM3–2 (ΔG = −7.4 kcal mol ) contains a stronger hydrogen of Ag O and related species in a meaningful way using DFT
2
bond and is calculated to be more stable than IM3 (ΔG = calculations as there are no reasonable structural models for
1
1
.1 kcal mol− ). The summit of the potential energy surface, Ag
2
O at this time. However, it would be reasonable to postulate
4
2
i.e. the turnover-limiting step, for the pathway is the activation that a number of equilibria between basic silver species
‡
−1
of the aryl C–H bond via TS2 (ΔG = 19.6 kcal mol ) to IM4. could exist in the reaction mixture. The presence of numerous
A KIE value of 3.6 was calculated for this turnover-limiting basic species could offer an explanation for why the starting
step. This is consistent with the experimentally observed KIE material 1 was fully consumed and a 63% yield of product 2
of approx. 3 for the cleavage of the aryl C–H bond (Scheme 3).
was measured in the absence of NaOtBu (ESI†). Finally,
IM4 is calculated to isomerize to the lower energy IM5 an interesting lag period was observed in most of the
through the formation of a hydrogen bond between the two kinetic studies described herein and further studies on a pre-
acetate ligands. Reductive elimination of Pd(HOAc)
2
from IM5 catalyst → catalyst step, are underway.
‡
to give product 2 proceeds through TS3 (ΔG = 14.3 kcal At this stage it would be pertinent to draw a comparison
−
1
18
mol ). The reductive elimination is relatively facile, and with the work of Choi et al. They found that in the substrates
Scheme 4 Reaction coordinate diagram for cross-dehydrogenative coupling of 4-phenoxy-2-coumarins.
This journal is © The Royal Society of Chemistry 2018
Dalton Trans.

View Article Online
Paper
Dalton Transactions
which proceed via a CMD, the more electron-rich position is
favoured (by Pd and ultimately by the coupling partner). They
determined that a CMD route, which would give arylation at
the C-3 position, due to the greater π-electron density at that
position, was in contrast to the experimentally observed C-2
arylation. However, when carbopalladation was evoked by Choi
et al., palladation did occur at the most electron-rich site, but
the coupling partner (i.e. aryl group) was delivered to the
2 D. A. Horton, G. T. Bourne and M. L. Smythe, Chem. Rev.,
2003, 103, 893–930.
3 G. P. McGlacken and L. M. Bateman, Chem. Soc. Rev., 2009,
38, 2447–2464.
4 D. J. Constable, P. J. Dunn, J. D. Hayler, G. R. Humphrey,
J. L. Leazer Jr., R. J. Linderman, K. Lorenz, J. Manley,
B. A. Pearlman and A. Wells, Green Chem., 2007, 9, 411–
420.
neighbouring
regioselectivity.
carbon,
resulting
in
the
observed
5 D. L. Davies, S. M. Donald and S. A. Macgregor, J. Am.
Chem. Soc., 2005, 127, 13754–13755.
In contrast to that used by Choi et al., our system does not
possess a suitable destination for the phenyl group, after C-3
palladation. Moreover, the electron donating properties of the
O-tether may play a role. Thus in our case, CMD becomes the
mode of choice in both C–H activation steps, with the cou-
marin C-3–H cleaved first. Finally, our system does not benefit
from the statistical effect of a large excess of the aryl coupling
partner (40-fold vs. 1 equivalent), although it is difficult to
assess this contribution.
6 D. L. Davies, S. M. Donald, O. Al-Duaij, S. A. Macgregor and
M. Pölleth, J. Am. Chem. Soc., 2006, 128, 4210–4211.
7 Y. Boutadla, D. L. Davies, S. A. Macgregor and
A. I. Poblador-Bahamonde, Dalton Trans., 2009, 5820–
5831.
8 D. García-Cuadrado, A. A. C. Braga, F. Maseras and
A. M. Echavarren, J. Am. Chem. Soc., 2006, 128, 1066–1067.
9 D. García-Cuadrado, P. de Mendoza, A. A. Braga, F. Maseras
and A. M. Echavarren, J. Am. Chem. Soc., 2007, 129, 6880–
6886.
1
1
1
1
1
0 M. Lafrance and K. Fagnou, J. Am. Chem. Soc., 2006, 128,
16496–16497.
1 M. Lafrance, C. N. Rowley, T. K. Woo and K. Fagnou, J. Am.
Chem. Soc., 2006, 128, 8754–8756.
2 S. I. Gorelsky, D. Lapointe and K. Fagnou, J. Am. Chem.
Soc., 2008, 130, 10848–10849.
Conclusions
In conclusion, the CDC of 4-phenoxy-2-coumarins is most
likely to proceed via the mechanism shown in Scheme 4. It
comprises of an initial, reversible C–H activation of the C-3–H
bond of the 2-coumarin via CMD, followed by irreversible,
turnover-limiting CMD of the aryl C–H bond. Reductive elimin-
ation of Pd(0) gives the product, and deprotonation and oxi-
dation of the Pd(0) back to the active catalyst Pd(II) completes
the catalytic cycle. The experimental evidence is consistent
with the DFT calculations. This work demonstrates that a
double CMD mechanism is accessible in the absence of suit-
able substrates for carbopalladation and provides key insights
into the C–H activation of coumarins, and the order of C–H
cleavage.
3 D. Lapointe and K. Fagnou, Chem. Lett., 2010, 39, 1118–
1
126.
4 S. I. Gorelsky, D. Lapointe and K. Fagnou, J. Org. Chem.,
012, 77, 658–668.
2
1
1
1
5 S. I. Gorelsky, Organometallics, 2012, 31, 794–797.
6 S. I. Gorelsky, Coord. Chem. Rev., 2013, 257, 153–164.
7 D. L. Davies, S. A. Macgregor and C. L. McMullin, Chem.
Rev., 2017, 117, 8649–8709.
8 H. Choi, M. Min, Q. Peng, D. Kang, R. S. Paton and
S. Hong, Chem. Sci., 2016, 7, 3900–3909.
1
1
2
2
9 C.-J. Li and Z. Li, Pure Appl. Chem., 2006, 78, 935–945.
0 C.-J. Li, Acc. Chem. Res., 2009, 42, 335–344.
1 S. A. Girard, T. Knauber and C.-J. Li, Angew. Chem., Int. Ed.,
Conflicts of interest
2014, 53, 74–100.
There are no conflicts of interest to declare.
2
2 S. A. Girard, T. Knauber and C.-J. Li, in From C-H to C-C
Bonds: Cross-Dehydrogenative-Coupling, The Royal Society of
Chemistry, Cambridge, UK, 2015, pp. 1–32.
Acknowledgements
23 M. Min, Y. Kim and S. Hong, Chem. Commun., 2013, 49,
196–198.
This research was supported by Science Foundation Ireland
2
2
2
4 M. Min and S. Hong, Chem. Commun., 2012, 48, 9613–9615.
5 J. Dickinson, Nat. Prod. Rep., 1993, 10, 71–98.
6 G. P. McGlacken and I. J. S. Fairlamb, Nat. Prod. Rep., 2005,
(
SFI/12/IP/1315 & SFI/12/RC/2275) and the Synthesis and Solid
State Pharmaceutical Centre (SSPC) (G. P. M. and
A. M. P.). A. M. P. and G. P. M wish to thank Dr Davide Tiana
for helpful discussion during the preparation of the
manuscript.
22, 369–385.
27 A. Goel, G. Taneja, A. Raghuvanshi, R. Kant and
P. R. Maulik, Org. Biomol. Chem., 2013, 11, 5239–5253.
28 M. Costa, T. A. Dias, A. Brito and F. Proença, Eur. J. Med.
Chem., 2016, 123, 487–507.
2
Notes and references
9 K. Mackey, L. M. Pardo, A. M. Prendergast, M.-T. Nolan,
L. M. Bateman and G. P. McGlacken, Org. Lett., 2016, 18,
2540–2543.
1
J. Hassan, M. Sévignon, C. Gozzi, E. Schulz and
M. Lemaire, Chem. Rev., 2002, 102, 1359–1470.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2018

View Article Online
Dalton Transactions
Paper
3
3
3
3
3
3
3
0 C. Cheng, W.-W. Chen, B. Xu and M.-H. Xu, Org. Chem. 37 T. Sperger, I. A. Sanhueza, I. Kalvet and F. Schoenebeck,
Front., 2016, 3, 1111–1115. Chem. Rev., 2015, 115, 9532–9586.
1 C. Cheng, W.-W. Chen, B. Xu and M.-H. Xu, J. Org. Chem., 38 This simplification should not affect the conclusions of
2
016, 81, 11501–11507.
2 K. Neog, A. Borah and P. Gogoi, J. Org. Chem., 2016, 81,
1971–11977.
our investigation, as we mainly focus on the comparison of
different reaction pathways.
39 K. Shen, Y. Fu, J.-N. Li, L. Liu and Q.-X. Guo, Tetrahedron,
2007, 63, 1568–1576.
40 H. Tajuddin, P. Harrisson, B. Bitterlich, J. C. Collings,
N. Sim, A. S. Batsanov, M. S. Cheung, S. Kawamorita,
A. C. Maxwell and L. Shukla, Chem. Sci., 2012, 3, 3505–3515.
1
3 A. Dey, M. A. Ali, S. Jana, S. Samanta and A. Hajra,
Tetrahedron Lett., 2017, 58, 313–316.
4 T. Balalas, A. Abdul-Sada, D. J. Hadjipavlou-Litina and
K. E. Litinas, Synthesis, 2017, 2575–2583.
5 E. M. Simmons and J. F. Hartwig, Angew. Chem., Int. Ed., 41 S. Rousseaux, S. I. Gorelsky, B. K. Chung and K. Fagnou,
012, 51, 3066–3072. J. Am. Chem. Soc., 2010, 132, 10692–10705.
6 J.-D. Chai and M. Head-Gordon, Phys. Chem. Chem. Phys., 42 B. Ren, M. Wang, J. Liu, J. Ge and H. Dong, ChemCatChem,
008, 10, 6615–6620. 2015, 7, 761–765.
2
2
This journal is © The Royal Society of Chemistry 2018
Dalton Trans.