3-Homoacyl coumarin: An all carbon 1,3-dipole for enantioselective concerted (3+2) cycloaddition

Article abstract of DOI:10.1039/c8cc07271j

A new type of all-carbon 1,3-dipole precursor, 3-homoacylcoumarin, was employed for the stereoselective (3+2) cycloaddition with indandione alkylidenes to generate a series of coumarin/indandione-fused spirocyclopentanes bearing four contiguous stereogenic centers. While two reaction pathways progressed simultaneously, detailed mechanistic investigation revealed that the highly efficient stereoselective concerted route dominated the extremely slow stepwise pathway.

Full text of DOI:10.1039/c8cc07271j

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
3-Homoacyl coumarin: an all carbon 1,3-dipole
for enantioselective concerted (3+2)
cycloaddition†
Cite this: Chem. Commun., 2018,
54, 12702
Received 7th September 2018,
Accepted 17th October 2018
Yi-Ru Chen, ‡ Madhusudhan Reddy Ganapuram, ‡ Kai-Hong Hsieh,
Kai-Han Chen, Praneeth Karanam, Sandip Sambhaji Vagh, Yan-Cheng Liou
Wenwei Lin
and
DOI: 10.1039/c8cc07271j
*
rsc.li/chemcomm
A new type of all-carbon 1,3-dipole precursor, 3-homoacylcoumarin,
was employed for the stereoselective (3+2) cycloaddition with indan-
dione alkylidenes to generate a series of coumarin/indandione-
fused spirocyclopentanes bearing four contiguous stereogenic
centers. While two reaction pathways progressed simultaneously,
detailed mechanistic investigation revealed that the highly efficient
stereoselective concerted route dominated the extremely slow
stepwise pathway.
With the advent of cycloaddition reactions, efficient synthesis
of complex natural products has become more practical.1 In
particular, (3+2) cycloadditions involving 1,3-dipoles are con-
sidered as one of the powerful strategies for the construction of
5-membered ring systems, which are an integral part of a large
number of biologically active systems.2 However, most of
the reported 1,3-dipoles bear a heteroatom, while all-carbon
1,3-dipoles are seldom reported to undergo cycloadditions
unless assisted by neighbouring functional groups which
induce bond polarization (Scheme 1).3 For instance, the use
of substrates such as (a) allenoates4a (b) MBH-carbonates4b,c (c)
metal-induced dipolar systems4d and (d) oxyallyl cations4e, f
involves the assistance of a bond polarizing unit (BPU) to
Scheme 1 Earlier reported formal all-carbon 1,3-dipoles, their equiva-
lents and our design.
Similarly, other approaches such as trimethylenemethane
generate formal all-carbon 1,3-dipoles (Scheme 1). Introduction of cycloadditions,5a,b (Scheme 1e) the cyclopropane donor–acceptor
a BPU in these substrates resulted in the creation of dipoles as a strategy5a,c,d (Scheme 1f) and the organocatalytic donor–acceptor
result of delocalization of the allylic p-system. However, all these strategy5a,e–g (Scheme 1g) have been developed utilizing non-
substrates require very specific methods of activation such as conjugated (no allylic p-system between dipoles) synthetic
nucleophilic (in the case of allenoates and MBH-carbonates, equivalents of all-carbon 1,3-dipoles for (3+2) cycloaddition
Schemes 1a and b) or metal catalysis (Scheme 1c) or rely on prior employing specific activation modes (palladium-activation;
generation of cations (e.g. oxyallyl cations, Scheme 1d) to induce Brønsted acid/Lewis acid or base).
1,3-dipolar properties, which limits their application to a great extent.
With regard to the concept of designing an all-carbon
1,3-dipolar system, it would be preferred that the substrate
displays properties such as (i) presence of a BPU (ii) inherent
1,3-dipolar properties without requiring a specific method
for its generation, and (iii) effective delocalization of the allylic
p-system. Although Du’s group was successful in designing a
substrate fulfilling these criteria,6 coumarin derivatives have
not yet been reported that display such functional group-
induced 1,3-dipolar properties.
Department of Chemistry, National Taiwan Normal University, 88, Sec. 4, Tingchow
Road, Taipei 11677, Taiwan, Republic of China. E-mail: wenweilin@ntnu.edu.tw;
Fax: +886 2 2932 4249
† Electronic supplementary information (ESI) available: Detailed time vs. conver-
sion plots, synthetic procedures, characterisation data and copies of spectra of all
compounds. CCDC 1484301 (3ad). For ESI and crystallographic data in CIF or
other electronic format see DOI: 10.1039/c8cc07271j
‡ These authors contributed equally.
12702 | Chem. Commun., 2018, 54, 12702--12705
This journal is ©The Royal Society of Chemistry 2018

View Article Online
Communication
ChemComm
Table
1
Optimization of (3+2) cycloaddition for the generation of
As one of the privileged scaffolds in medicinal chemistry and
materials science, coumarin derivatives are being passionately
investigated for their further applications.7 In the synthetic aspect,
the utility of coumarin derivatives as substrates has drawn much
attention because of their reactivity.8 Although the inherent
aromatic-like properties of coumarin shield its electrophilic nature
to a great extent, recent reports show that an activating group
installed on the 3-position could indeed enhance its reactivity.9
On the other hand, incorporating a hydroxyl group on either
the 3 or 4-position of coumarin would render it nucleophilic,
thus opening up the possibility of utilizing it in cycloaddition
reactions.10 However, such amphiphilic coumarin derivatives
have never been utilized as all-carbon 3-atom partners for cyclo-
additions. Herein, we report a new type of all-carbon 1,3-dipole
precursor, 3-homoacylcoumarin, that is employed for the stereo-
selective (3+2) cycloaddition with indandione alkylidenes
to generate a series of coumarin/indandione-fused spirocyclo-
pentanes bearing four contiguous stereogenic centers.
To begin with, substrates 1a and 2a were treated in the
presence of different amine catalysts (Fig. 1) in dichloromethane
at 30 1C to understand the activation mode of the (3+2) cyclo-
addition for the generation of 3aa (Table 1, entries 1–11). This
showed that the use of tertiary amine catalysts resulted in better
reactivity of substrates when compared to the use of primary
and secondary amines. In these cases, minor amounts of 4aa
(Michael adduct) could also be isolated, which made us assume
that the reaction may proceed via a step-wise pathway rather
than a concerted one. (However, upon detailed mechanistic
study, the reaction was found to be concerted, which is dis-
cussed in later sections.) A similar reactivity trend was observed
in the case of both chiral and achiral catalysts, which indicated
that the reaction may follow a Brønsted-base catalysis mode,
rather than enamine or iminium activation. It was observed
adduct 3aaa
Entry
Catalyst
Solvent
Yieldc [%]
eed [%]
b
1
2
3
4
5
6
7
8
Cy-NH2
Pyrrolidine
DABCO
QN-NH2
Hayashi’s cat.
QN
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CHCl3
Toluene
THF
52
74
99
42
67
89
85
92 (89)
67
69
86
63
84
79
84
91
78
35
88
68
43
Racemic
Racemic
Racemic
21
33e
54
85
QN-SQA
QN-T
QD-T
CN-T
CD-T
QN-T
QN-T
QN-T
QN-T
QN-T
QN-T
QN-T
QN-T
QN-T
QN-T
96 (96)
9
84e
10
11
12
13
14
15
16
17
18
19 f
20g
21h
81e
91
93
93
71
88
33
4
73
EtOAc
CH3CN
EtOH
NMP
CH2Cl2
CH2Cl2
CH2Cl2
96
96
96
a
Unless otherwise specified, all the reactions were carried out using 1a
(0.1 mmol), 2a (1.1 equiv.) and catalyst (10 mol%) in the given solvent
b
c
(0.5 mL) and stirred at 30 1C. Cy-NH2 = cyclohexylamine. Deter-
mined by H NMR analysis of the crude reaction mixture using Ph3CH
as an internal standard. Determined by HPLC analysis on a chiral
1
d
e
f
stationary phase. ent-3aa was obtained. 15 mol% of QN-T was used.
g
h
Reaction carried out at 10 1C. 10 mol% of benzoic acid was used.
that the hydrogen-bonding thiourea and squaramide catalysts probably be due to the perturbation of hydrogen bonding
resulted in products with better stereoselectivity and among interactions in the reacting model between substrates and the
them QN-T provided the best result.
catalyst by the protic solvent. Additionally, when the reaction
A brief screening of different parameters such as the solvent, was carried out in the presence of an acidic additive, the yield of
catalyst loading and temperature indicated that carrying out the product dropped to a large extent (entry 21). This clearly
the reaction with 1a and 2a (1.1 equiv.) in the presence of indicated that an acidic additive would decrease the activity
10 mol% of QN-T in CH2Cl2 at 30 1C gave the best result in of the catalyst, thereby resulting in reduced reactivity.
terms of reactivity and selectivity. It is noteworthy that when the
Under the optimized conditions, a series of substrates 1 and
reaction was carried out in EtOH, 3aa was obtained with very 2 bearing different substitutions were employed in the (3+2)
poor enantioselectivity (Table 1, entry 17). This could most cycloaddition to furnish a wide range of products 3 (Scheme 2).
It was observed that the electronic or steric effects of the R1 and
R2 substituents on 1 were not prominent and all the substitu-
ents resulted in good yields and stereoselectivities of the
corresponding products. However, the R3 substituent on 2
displayed variable effects on the outcome of the reaction. The
presence of an electron-withdrawing group on R3 enhanced
the reactivity of 2, while an electron-donating group reduced
the reactivity and required longer reaction times. The position
of the R3 substituent also had a considerable impact on the
reactivity of 2. A substitution at the ortho position of the
aryl ring resulted in reduced reactivity (3aj–3ak, Scheme 2).
Moreover, the presence of an o-hydroxy substituent resulted
Fig. 1 Catalysts used for the screening of (3+2) cycloaddition.
in the product in poor yield and selectivity (3aj, Scheme 2).
This journal is ©The Royal Society of Chemistry 2018
Chem. Commun., 2018, 54, 12702--12705 | 12703

View Article Online
ChemComm
Communication
Scheme 3 Control experiments to elucidate the reaction mechanism.
step being higher than the first step or (ii) the reaction could
be following a major concerted pathway, along with a minor
step-wise pathway that is responsible for the formation of
minor amounts of 4aa. Unfortunately, the enantioselectivity
of 4aa, which would have given some crucial information, was
not disclosed since its enantiomers could not be separated
despite extensive efforts employing various HPLC methods.
To rule out one of the two possibilities, another control
experiment was conducted with rac-4aa under optimized con-
ditions (chiral catalyst) with real-time monitoring of different
Scheme 2 Evaluation of substrate scope for (3+2) cycloaddition.a aUnless
otherwise specified, all the reactions were carried out using 1 (0.1 mmol),
2 (1.1 equiv.) and QN-T (10 mol%) in CH2Cl2 (0.5 mL) at 30 1C. bEnantiomeric
excess was determined by HPLC analysis on a chiral stationary phase.
cWhen the reaction was performed on a gram scale (4.0 mmol), 3aa was
obtained in 78% yield and 93% ee. dThe absolute configuration of 3ad was
determined by X-ray analysis.
1
species in the reaction via H NMR spectroscopy (Scheme 3c).
The formation of 3aa from 4aa required a longer time (35 h, 72%
yield) than that from 1a and 2a (23 h, 95%), which contradicted
the first possibility. This clearly revealed that the step-wise path-
way was less efficient. Moreover, it was surprising to see that the
This indicated that the o-hydroxy group may reduce the basicity resulting adduct 3aa in our control experiment (Scheme 3c) was
of the catalyst and may also interfere with the hydrogen enantiopure (92% ee from rac-4aa). This hinted at a possible
bonding between the substrate and the catalyst. Interestingly, DKR (dynamic kinetic resolution) of rac-4aa. A closer investigation
substrates bearing electron-rich heteroaryl groups such as furyl
and thienyl substituents could also result in the corresponding
products 3al and 3am with good stereoselectivities, although
the reactivity was slightly lower. However, the electron-deficient
pyridinyl substitution displayed relatively better reactivity to
afford 3an. It was also demonstrated that the reaction could
also be performed on a gram-scale with similar efficiency
(Table 1, 3aa).
Mechanistically, this (3+2) cycloaddition could proceed
either via a two-step double Michael addition pathway or a
concerted pathway. The formation of intermediate 4aa during
optimization studies (Table 1) initially hinted at a step-wise
pathway. Real-time monitoring of 4aa at various stages of the
reaction under optimized conditions (Scheme 3a) revealed that
only minor amounts of 4aa were present in the reaction
medium at any given time, with its maximum observed amount
being 11% at t = 3 h (based on 1H NMR data of the crude
reaction mixture; see the ESI† for the reaction progress moni-
toring). Even when the reaction was slowed down (DABCO,
À5 1C), the maximum observed amount of 4aa was 30% at
t = 30 h (Scheme 3b). This indicated two possibilities: (i) the
Scheme 4 Mechanism of (3+2) cycloaddition as revealed by control
experiments.
overall reaction could be step-wise, with the rate of the second
12704 | Chem. Commun., 2018, 54, 12702--12705
This journal is ©The Royal Society of Chemistry 2018

View Article Online
Communication
ChemComm
other bond-polarizing units to design all-carbon 1,3-dipolar pre-
cursors is currently in progress.
The authors thank the Ministry of Science and Technology
of the Republic of China (MOST 104-2113-M-003-002-MY3) for
financial support.
Conflicts of interest
There are no conflicts to declare.
Notes and references
1 (a) N. Fu and T. T. Tidwell, Tetrahedron, 2008, 64, 10465; (b) G. Pandey,
D. Dey and S. K. Tiwari, Tetrahedron Lett., 2017, 58, 699; (c) R. P. Hsung,
A. V. Kurdyumov and N. Sydorenko, Eur. J. Org. Chem., 2005, 23;
(d) M. Harmata, Chem. Commun., 2010, 46, 8886; (e) M. M. Heravi,
T. Ahmadi, M. Ghavidel, B. Heidari and H. Hamidi, RSC Adv., 2015,
5, 101999; ( f) K. E. O. Ylijoki and J. M. Stryker, Chem. Rev., 2013, 113, 2244.
2 Y. Xing and N.-X. Wang, Coord. Chem. Rev., 2012, 256, 938.
3 (a) T. Hashimoto and K. Maruoka, Chem. Rev., 2015, 115, 5366;
(b) M. S. Singh, S. Chowdhury and S. Koley, Tetrahedron, 2016,
72, 1603; (c) R. Huisgen, J. Org. Chem., 1976, 41, 403.
4 (a) T. M. V. D. Pinho e Melo, Curr. Org. Chem., 2009, 13, 1406;
(b) Y.-Y. Zhao, S. Zhao, J.-K. Xie, X.-Q. Hu and P.-F. Xu, J. Org. Chem.,
2016, 81, 10532; (c) B. Tan, N. R. Candeias and C. F. Barbas, J. Am.
Chem. Soc., 2011, 133, 4672; (d) G. Zhang and L. Zhang, J. Am. Chem.
Soc., 2008, 130, 12598; (e) H. Li and J. Wu, Synthesis, 2015, 22; ( f ) C. Liu,
E. Z. Oblak, M. N. Vander Wal, A. K. Dilger, D. K. Almstead and
D. W. C. MacMillan, J. Am. Chem. Soc., 2016, 138, 2134.
5 (a) I. Kumar, RSC Adv., 2014, 4, 16397; (b) B. M. Trost, D. A. Bringley,
T. Zhang and N. Cramer, J. Am. Chem. Soc., 2013, 135, 16720;
(c) N. R. O’Connor, J. L. Wood and B. M. Stoltz, Isr. J. Chem., 2016,
56, 431; (d) T. F. Schneider, J. Kaschel and D. B. Werz, Angew. Chem.,
Int. Ed., 2014, 53, 5504; (e) D. Cheng, Y. Ishihara, B. Tan and
C. F. Barbas, ACS Catal., 2014, 4, 743; ( f ) J. Zhang, D. Cao,
H. Wang, C. Zheng, G. Zhao and Y. Shang, J. Org. Chem., 2016,
81, 10558; (g) W. Sun, L. Hong, G. Zhu, Z. Wang, X. Wei, J. Ni and
R. Wang, Org. Lett., 2014, 16, 544.
6 B.-L. Zhao and D.-M. Du, Chem. Commun., 2016, 52, 6162.
7 For biological activity of coumarins, please see: (a) F. G. Medina,
J. G. Marrero, M. Macias-Alonso, M. C. Gonzalez, I. Cordova-
Guerrero, A. G. Teissier Garcia and S. Osegueda-Robles, Nat. Prod.
Rep., 2015, 32, 1472; (b) J. Grover and S. M. Jachak, RSC Adv., 2015,
5, 38892; (c) S. Sandhu, Y. Bansal, O. Silakari and G. Bansal, Bioorg.
Med. Chem., 2014, 22, 3806. For the importance of coumarins in
materials sciences, please see: (d) A. M. Breul, M. D. Hager and
U. S. Schubert, Chem. Soc. Rev., 2013, 42, 5366; (e) W. Shi and H. Ma,
Chem. Commun., 2012, 48, 8732.
8 (a) A. Y. Fedorov, A. V. Nyuchev and I. P. Beletskaya, Chem. Heterocycl.
Compd., 2012, 48, 166; (b) W. Zhao, L. Xu, Y. Ding, B. Niu, P. Xie,
Z. Bian, D. Zhang and A. Zhou, Eur. J. Org. Chem., 2016, 325; (c) H. Guo,
E. Herdtweck and T. Bach, Angew. Chem., Int. Ed., 2010, 49, 7782;
(d) J. F. Teichert and B. L. Feringa, Chem. Commun., 2011, 47, 2679.
9 (a) A. Song, X. Zhang, X. Song, X. Chen, C. Yu, H. Huang, H. Li and
W. Wang, Angew. Chem., Int. Ed., 2014, 53, 4940; (b) Y.-T. Lee, U. Das,
Y.-R. Chen, C.-J. Lee, C.-H. Chen, M.-C. Yang and W. Lin, Adv. Synth.
Catal., 2013, 355, 3154; (c) Y. Wang, Z.-H. Yu, H.-F. Zheng and
D.-Q. Shi, Org. Biomol. Chem., 2012, 10, 7739.
10 (a) L.-P. Fan, P. Li, X.-S. Li, D.-C. Xu, M.-M. Ge, W.-D. Zhu and
J.-W. Xie, J. Org. Chem., 2010, 75, 8716; (b) Y. Gao, Q. Ren, L. Wang
and J. Wang, Chem. – Eur. J., 2010, 16, 13068.
11 (a) J. Wang, H. Xie, H. Li, L. Zu and W. Wang, Angew. Chem., Int. Ed.,
2008, 47, 4177; (b) S. Meninno, G. Croce and A. Lattanzi, Org. Lett.,
2013, 15, 3436; (c) C. Yu, Y. Zhang, A. Song, Y. Ji and W. Wang,
Chem. – Eur. J., 2011, 17, 770.
Fig. 2 Real-time monitoring of the reaction in Scheme 4c. (a) t = 0 h to
35 h; (b) t = 0 h to 3 h.
revealed that in the initial stages of this reaction (from t = 0 to
t = 2 h), no cycloaddition product 3aa could be seen even when
4aa was being consumed (Fig. 2; for detailed time vs. conver-
sion plots, see the ESI†), which was contradictory to the
characteristics of a DKR.11 Instead, the generation of 1a and
2a was observed via a retro Michael reaction (control experi-
ments 3d and 3e revealed that 1a and 2a are indeed obtained
from 4aa and not from two-step decomposition of 3aa). As the
reaction progressed, the formation of 3aa was noticed, the rate
of which was proportional to the rate of consumption of
generated 1a and 2a (t = 10 to t = 35 h), indicating the formation
of enantiopure 3aa from 1a and 2a and not from 4aa. All these
results clearly revealed that the minor amounts of intermediate
4aa generated via the less efficient step-wise pathway reverted
back to substrates 1a and 2a, presumably due to the inhibition
from the inherent aromatic like nature of the coumarin motif
(in the intermediate 4aa) in promoting further cyclization. The
thus generated 1a and 2a further resulted in 3aa via the highly
efficient concerted pathway.
In conclusion, we have successfully demonstrated amphiphilic
3-homoacyl coumarins as all-carbon 1,3-dipolar precursors for
the (3+2) cycloaddition reaction with arylidene indandione
derivatives resulting in interesting adducts in excellent yields
and stereoselectivities. This result revealed that the concept of
designing an all-carbon 1,3-dipolar precursor by incorporating
a bond-polarizing unit can be realized. Detailed mechanistic
studies indicated that the reaction followed a highly efficient
concerted pathway to result in products with high stereo-
selectivities. Moreover, the step-wise pathway was found to be
ineffective under our reaction conditions as the intermediate
could only revert back to the starting materials that sub-
sequently delivered the final product, making the concerted
pathway even more efficient. Further study towards incorporating
This journal is ©The Royal Society of Chemistry 2018
Chem. Commun., 2018, 54, 12702--12705 | 12705