Synthesis of functionalized chromenes and benzofurans from aryloxy propargyl malonates

Article abstract of DOI:10.1002/ijch.201300038

A series of aryloxy propargyl malonates was efficiently converted into various chromenes by a gold-catalyzed cyclization/Br?nsted acid catalyzed isomerization sequence, and the chromenes were subsequently converted into benzofurans by implementing an additional photochemical rearrangement. The photochemically induced ring contraction was proven to involve the intermediate formation of dienones.

Full text of DOI:10.1002/ijch.201300038

Communication
DOI: 10.1002/ijch.201300038
Synthesis of Functionalized Chromenes and Benzofurans
from Aryloxy Propargyl Malonates
I. D. Jurberg,[a] K. Ikeda,[a] D. Antwi-Omane,[a] and F. Gagosz*[a]
Abstract: A series of aryloxy propargyl malonates was effi-
ciently converted into various chromenes by a gold-cata-
lyzed cyclization/Brønsted acid catalyzed isomerization se-
quence, and the chromenes were subsequently converted
into benzofurans by implementing an additional photo-
chemical rearrangement. The photochemically induced ring
contraction was proven to involve the intermediate forma-
tion of dienones.
Keywords: benzofurans · catalysis · chromenes · gold · photochemistry
1
Chromene and benzofuran moieties are found in the
structures of a large number of biologically active natural
and designed products.[1] They are also key scaffolds in
drugs and, therefore, of paramount importance to the
pharmaceutical industry. A selection of natural products
and pharmaceutical agents containing one of these two
structural units is shown in Scheme 1. While numerous
routes to access chromenes and benzofurans have already
been established,[2] the development of new synthetic pro-
cedures is still of high interest, especially if these allow
practical, rapid, and efficient access to these structural
motifs.
CDCl3, at 608C, and the reaction was monitored by H
NMR spectroscopy (Scheme 3). We were pleased to see
that 4a was efficiently converted into chromane 8a and
chromene 5a, which were co-isolated as an inseparable
mixture, in 97% yield, after
a 5h reaction time
(Scheme 3, eq. 3). The formation of the isomeric chro-
mene 5a could be explained by a gold-catalyzed isomeri-
zation of the exo double bond in chromane 8a. This par-
ticular cyclization could be performed with the simple
[(Ph3P)Au]NTf2 catalyst[3c] (1 mol%), and a 1:45 mixture
of 8a and 5a was obtained in this case after an 8h reaction
time (yield: 98%). This catalyst, however, was shown to
be completely inactive in several other cases.[9] The use of
7 as the catalyst appears to be more synthetically valua-
ble, as it allowed selective and sequential access to either
chromane 8a or chromene 5a after an additional Brønsted
acid catalyzed isomerization step (pTsOH [5 mol%],
97% overall yield; see Scheme 3, eq. 4).[10,11]
The scope of the cyclization was next investigated using
a series of aryloxy propargyl malonates 4a–k as the sub-
strates and 7 as the catalyst (1–3 mol%), in chloroform,
at 608C (Table 1). The gold-catalyzed cyclization was fol-
lowed by a Brønsted acid catalyzed isomerization step,
performed in situ, in order to convert the resulting chro-
mane 8 into the corresponding chromene 5. The transfor-
mation proved to be general, and substrates 4a–k, which
possess various substituents on the aromatic scaffold,
could be converted into chromenes 5a–k in good to excel-
As part of our work on homogeneous gold catalysis,[3]
we recently reported a convenient and widely applicable
procedure for the synthesis of dihydroquinolines 2 and in-
doles
3
from N-aminophenyl propargyl malonates
1 (Scheme 2, eq. 1).[4] The synthetic route that was devel-
oped involved a gold-catalyzed 6-exo cyclization[5]/Brønst-
ed acid-catalyzed isomerization sequence that produced
the intermediates 2. A subsequent photochemical rear-
rangement allowed the conversion of 2 into 3. Following
these studies, we imagined that an analogous procedure
could be employed for the synthesis of chromenes 5 and
benzofurans 6 (Scheme 2, eq. 2).[6] This approach ap-
peared to be particularly interesting as the required phen-
oxy propargyl malonate substrates 4 could be rapidly and
easily synthesized from simple, commercially available
compounds.[7] We report herein the results of our investi-
gations.
The simple phenoxy propargyl malonate 4a, which was
rapidly synthesized in a two-step procedure from diethyl
chloromalonate, phenol, and propargyl bromide,[7] was
chosen as a model substrate, in order to validate our ap-
proach to the synthesis of chromenes and benzofurans.
We first examined the possibility of performing gold-cata-
lyzed 6-exo cyclization. On the basis of our previous
work,[4] 4a was treated with the gold(I) complex 7,[8] in
[a] I. D. Jurberg, K. Ikeda, D. Antwi-Omane, F. Gagosz
Dꢀpartement de Chimie
UMR 7652
CNRS/Ecole Polytechnique
91128 Palaiseau (France)
e-mail: gagosz@dcso.polytechnique.fr
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/ijch.201300038.
Isr. J. Chem. 2013, 53, 915 – 922
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
915

Communication
F. Gagosz et al.
Scheme 1. Natural products and pharmaceutical agents possessing a chromene or a benzofuran structural unit.
Scheme 2. Synthetic approach to chromenes and benzofurans.
Scheme 3. Initial attempt of 6-exo cycloisomerization with substrate 4a.
lent overall yields (72–97%). The gold-catalyzed step
(4!8) proved to be generally selective, leading to the for-
mation of chromanes 8 as the major products (8:5=2:1 to
1:0). The time required for completing the cyclization was
highly dependent on the nature of the aromatic substitu-
ent: electron withdrawing groups slowed down the cycli-
zation (Table 1, entries 3, 4, and 7), while electron donat-
ing groups sped it up (Table 1, entries 2 and 5). Notably,
the gold-catalyzed cyclization proved to be completely se-
lective in the cases of the meta-bromo substituted sub-
916
www.ijc.wiley-vch.de
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Isr. J. Chem. 2013, 53, 915 – 922

Synthesis of Chromenes and Benzofurans
Table 1. Scope of the gold-catalyzed cyclization/Brønsted acid catalyzed isomeri-
zation sequence.[a]
1
[a] Substrate concentration: 0.2 M. [b] Ratio determined by H NMR spectroscopy.
[c] Isolated yield.
Isr. J. Chem. 2013, 53, 915 – 922
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.ijc.wiley-vch.de
917

Communication
Table 2. Photochemical rearrangement of chromenes 5a–k.[a]
F. Gagosz et al.
a thermodynamic mixture of chromanes and chromenes
was obtained (Table 1, entries 5 and 9).[12]
We next focused our attention on the possibility of
photochemical rearrangement of chromenes 5a–k into the
corresponding benzofuran derivatives, as such a rearrange-
ment was previously reported for the analogous conver-
sion of dihydroquinolines into indoles (see Scheme 2,
eq. 1).[4] Gratifyingly, the exposure of chloroform solu-
tions of 5a–k to visible light irradiation, using a sun-
lamp,[7,13] led to their clean and efficient conversion into
the corresponding functionalized benzofurans 6a–k,
which were isolated in yields ranging from 76% to 99%
(Table 2).[14,15] This photochemical rearrangement ap-
peared to be a slow process under the experimental con-
ditions employed, as the reaction time was generally
greater than 10 h.
A mechanistic proposal for the formation of benzofur-
an 6 from chromene 5 is presented in Scheme 4. The irra-
diation of 5 with visible light could induce a photochemi-
cal ring opening of the pyran unit that would produce the
intermediate dienone 9. A 5-exo addition of the keto
functionality onto the alkylidene malonate moiety in 9,
followed by a rearomatization/proton transfer sequence,
could then deliver the corresponding benzofuran 6. Such
a mechanism is supported by previous studies reported in
the literature, which show that the analogous 2,2-diaryl
chromenes 10 can undergo ring opening under irradia-
tion, furnishing the corresponding dearomatized dienones
11.[16] While such a photochemical process is generally re-
versible, the presence of the two electrophilic ester moiet-
ies in intermediates 9 should, in our case, drive the reac-
tion towards the irreversible formation of the ring con-
traction benzofuran products 6.
The involvement of dienone 9 in the formation of ben-
zofuran 6 was confirmed by a separate experiment, in
which chromene 5a was irradiated in the presence of an
excess of triethylamine in CDCl3 (Scheme 5).[17] Monitor-
1
ing of the reaction by H NMR spectroscopy showed the
presence of the postulated intermediate dienone 9a which
could be obtained as a component of a mixture with ben-
zofuran 6a (70%, 9a:6a=2:1), after 7h of reaction. It was
shown that this mixture of 9a and 6a rapidly evolves in
CDCl3 (under visible light irradiation or not) to give ben-
zofuran 6a as the sole product.
1
[a] Substrate concentration: 0.2 M. [b] Ratio determined by H NMR
spectroscopy. [c] Isolated yield.[a] Substrate concentration: 0.2 M.
[b] Isolated yield.
It should be noted that 9a could not be detected in the
reaction mixture when 5a was irradiated in CDCl3 in the
absence of triethylamine.[17] Additionally, in order to
obtain dienone 9a, careful purification by flash chroma-
tography on silica gel basified with triethylamine was re-
quired. When triethylamine was omitted, only the benzo-
furan 6a could be isolated.[18] These observations led us to
hypothesize that the cyclization of dienone 9a into benzo-
furan 6a should be favored under acidic conditions (H+
traces in CHCl3 or SiOH groups in silica), probably via
activation of the alkylidene moiety in 9a. On this basis,
and since the isomerization of chromane 8 into chromene
strate 4f (Table 1, entry 6) and the naphthalene derivative
4i (Table 1, entry 9). While a 1 mol% loading of catalyst 7
was generally sufficient to efficiently perform the cycliza-
tion, a higher loading was necessary in the cases of sub-
strates 4d, g, h, k. The Brønsted acid catalyzed isomeriza-
tion of 8 into 5 was generally rapid (<6 h), with the ex-
ception of chromane 8g, for which an extended reaction
time was required (Table 1, entry 7). For chromanes 8e
and 8i, the isomerization could not be completed, and
918
www.ijc.wiley-vch.de
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Isr. J. Chem. 2013, 53, 915 – 922

Synthesis of Chromenes and Benzofurans
Scheme 4. Mechanistic proposal for the photochemical rearrangement of chromenes 5.
Scheme 5. Probing the intermediacy of dienone 9 in the formation of benzofuran 6.
5 was performed under acidic conditions, we envisaged
the production of benzofurans 6a–k by a more convenient
one-pot, 3-step, sequential procedure. The substrates
were first cyclized under gold-catalysis conditions (step
1), the resulting chromanes then isomerized under
Brønsted acid conditions (step 2), and the corresponding
chromenes finally isomerized into benzofurans under visi-
ble light irradiation (step 3). As seen from the results
compiled in Table 3, this approach was successful. Benzo-
furans 6a–k could be produced in yields at least compara-
ble to those obtained when chromene intermediates 5a–k
were isolated. This proved that the photochemical rear-
rangement was compatible with the presence of both the
gold catalyst and the Brønsted acid, and that no inter-
mediate purification of the chromenes 5a–k was required
in order to obtain benzofurans 6a–k in high yields.
Finally, we envisaged rendering the procedure even
more practical by reacting the substrates with the gold
catalyst 7 in the presence of p-TsOH, under visible light
irradiation, in chloroform, at 608C. This was attempted
on substrates 4h and 4j. As seen in Table 4, these experi-
mental conditions were suitable, and the corresponding
benzofurans 6h and 6j could be isolated in good yields.
This procedure is extremely advantageous, as it does not
Scheme 6. Unsuitable substrates for the production of benzofurans.
Isr. J. Chem. 2013, 53, 915 – 922
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.ijc.wiley-vch.de
919

Communication
F. Gagosz et al.
Table 3. Synthesis of benzofurans 6a–k by a one-pot, three-step, sequential proce-
dure.[a]
1
[a] Substrate concentration: 0.2 M. [b] Ratio determined by H NMR spectroscopy.
[c] Isolated yield.[a] Substrate concentration: 0.2 M. [b] Isolated yield.[a] Substrate
concentration: 0.2 M. [b] Isolated yield. In parentheses: overall yield of 6 when
chromene 5 was intermediately isolated after step 2 (see Tables 1 and 2).
require the sequential addition of the reagents, while it
produces the desired benzofurans in yields comparable to
those previously obtained (see Tables 1, 2, and 3).
served when derivative 4l was used as the substrate. This
result might be explained by the strong electron with-
drawing character of the cyano group, which would disfa-
vor the nucleophilic addition of the aromatic nucleus to
the gold-activated alkyne. Additionally, no cyclization
was observed when a substituent such as a vinyl (4m) or
While synthetically interesting, this new approach to
benzofuran
synthesis
showed
some
limitations
(Scheme 6). No gold-catalyzed cyclization could be ob-
920
www.ijc.wiley-vch.de
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Isr. J. Chem. 2013, 53, 915 – 922

Synthesis of Chromenes and Benzofurans
Table 4. Synthesis of benzofurans 6h and 6j by a one-pot, non-sequential proce-
dure.[a]
1
[a] Substrate concentration: 0.2 M. [b] Ratio determined by H NMR spectroscopy.
[c] Isolated yield.[a] Substrate concentration: 0.2 M. [b] Isolated yield.[a] Substrate
concentration: 0.2 M. [b] Isolated yield. In parentheses: overall yield of 6 when
chromene 5 was intermediately isolated after step 2 (see Tables 1 and 2).[a] Sub-
strate concentration: 0.2 M. [b] Isolated yield. [c] Yield obtained by employing the
one-pot, 3-step, sequential procedure (see Table 3). [d] Overall yield of 6 when
chromene 5 was intermediately isolated after step 2 (see Tables 1 and 2).
phenyl group (4n) was present at the terminus of the
alkyne, probably the result of an unfavorable steric inter-
action in the cyclizing transition state.
light irradiation. From a mechanistic point of view, it was
also proven that the photochemical rearrangement of
chromenes involves the formation of intermediate dien-
ones that subsequently undergo ring closure to produce
the benzofurans.
Regarding the nature of the linker, the procedure
could also be applied to substrate 4o, which possesses
a single ester moiety at the position alpha to the oxygen
atom (eq. 5). By comparison with the use of substrate 4a,
a lower yield was obtained for the formation of the inter-
mediate chromene (5o), but the photochemically induced
ring contraction leading to the benzofuran (6o) proved to
be equally efficient.
References
[1] For chromenes, see: a) K. C. Nicolaou, J. A. Pfefferkorn, S.
Barluenga, H. J. Mitchell, A. J. Roecker, G. Q. Cao, J. Am.
Chem. Soc. 2000, 122, 9968; b) K. C. Nicolaou, J. A. Pfeffer-
korn, S. Barluenga, H. J. Mitchell, A. J. Roecker, G. Q. Cao,
J. Am. Chem. Soc. 2000, 122, 9939; c) K. C. Nicolaou, J. A.
Pfefferkorn, S. Barluenga, H. J. Mitchell, A. J. Roecker,
G. Q. Cao, J. Am. Chem. Soc. 2000, 122, 9954, and referen-
ces cited therein. For benzofurans, see: d) B. A. Keay, P. W.
Dibble, in Comprehensive Heterocyclic Chemistry II (Eds.:
A. R. Katritzky, C. W. Rees, E. F. V. Scriven), vol. 2, Perga-
mon Press, Oxford, 1996, pp. 395–436.
[2] For chromenes, see: a) Y. Shi, M. Shi, Org. Biomol. Chem.
2007, 5, 1499. For benzofurans, see: b) X.-L. Hou, Z. Yang,
H. N. C. Wong, “Furans and Benzofurans,” in Progress in
Heterocyclic Chemistry (Eds.: G. W. Gribble, T. L. Gil-
christ), vol. 14, Pergamon Press, Oxford, 2002, pp. 139–179.
In conclusion, we have developed a synthetic sequence
– gold-catalyzed cyclization, followed by Brønsted acid-
catalyzed isomerization and photochemical rearrange-
ment – which allows new access to functionalized benzo-
furans from aryloxy propargyl malonates. This sequence
could be performed stepwise, thus allowing the possible
isolation of the intermediate chromanes and chromenes,
or in a one-pot manner. It was also shown that the reac-
tion conditions employed for each step of the sequence
were compatible with each other. The benzofurans could
therefore be efficiently produced following a convenient
one-pot protocol where the substrates simply reacted
with the gold catalyst and the Brønsted acid under visible
Isr. J. Chem. 2013, 53, 915 – 922
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.ijc.wiley-vch.de
921

Communication
F. Gagosz et al.
[3] For selected contributions, see: a) A. Wetzel, F. Gagosz,
Angew. Chem. 2011, 123, 7492; Angew. Chem. Int. Ed. 2011,
50, 7354; b) B. Bolte, F. Gagosz, J. Am. Chem. Soc. 2011,
133, 7696; c) B. Bolte, Y. Odabachian, F. Gagosz J. Am.
Chem. Soc. 2010, 132, 7294; d) N. Mꢀzailles, L. Ricard, F.
Gagosz, Org. Lett. 2005, 7, 4133.
[4] a) C. Gronnier, Y. Odabachian, F. Gagosz, Chem. Commun.
(Cambridge, U.K.) 2011, 47, 218. See also : b) I. D. Jurberg,
F. Gagosz, J. Organomet. Chem. 2011, 696, 37.
[5] For selected reviews on gold catalysis, see: a) A. S. K.
Hashmi, M. Rudolph, Chem. Soc. Rev. 2008, 37, 1766; b) E.
Jimꢀnez-NfflÇez, A. M. Echavarren, Chem. Rev. (Washing-
ton, DC, U.S.) 2008, 108, 3326; c) Z. Li, C. Brower, C. He,
Chem. Rev. (Washington, DC, U.S.) 2008, 108, 3239; d) D. J.
Gorin, F. D. Toste, Chem. Rev. (Washington, DC, U.S.) 2008,
108, 3351; e) A. S. K. Hashmi, Chem. Rev. (Washington, DC,
U.S.) 2007, 107, 3180; f) A. Fürstner, P. W. Davies, Angew.
Chem. 2007, 119, 3478; Angew. Chem. Int. Ed. 2007, 46,
3410.
[6] For examples of gold-catalyzed formation of chromene and
benzofurans, see: a) I. N. Lykakis, C. Efe, C. Gryparis, M.
Stratakis, Eur. J. Org. Chem. 2011, 2334; b) C. Nevado,
A. M. Echavarren, Chem.–Eur. J. 2005, 11, 3155; c) A.
Arcadi, F. Blesi, S. Cacchi, G. Fabrizi, A. Goggiamani, F.
Marinelli, Org. Biomol. Chem. 2012, 10, 9700; d) R. S.
Menon, A. D. Findlay, A. C. Bissember, M. G. Banwell, J.
Org. Chem. 2009, 74, 8901; e) A. S. K. Hashmi, W. Yang, F.
Rominger, Chem.–Eur. J. 2012, 18, 6576; f) Y. Liu, J. Qian,
S. Lou, Z. Xu, J. Org. Chem. 2010, 75, 6300; g) A. Hashmi,
M. Rudolph, J. Bats, W. Frey, F. Rominger, T. Oeser, Chem.
– Eur. J. 2008, 14, 6672.
[9] For instance [(Ph3P)Au]NTf2 was completely inactive with
substrates 4c and 4g.
[10] No reaction occurred in the absence of a gold catalyst.
[11] p-TsOH did not catalyze the 6-exo cyclization.
[12] The isomerization of the double bond should be more diffi-
cult in the case of 8e and 8i, due to the generation of unfav-
orable steric interactions between the methyl at the vinylic
position and the methyl on the aromatic nucleus (for 8e) or
between the methyl at the vinylic position and the naphtha-
lene moiety (for 8i).
[13] Samples were irradiated at 300 W using an Osram ULTRA-
VITALUXꢁ Sun Lamp.
[14] For entries 5 and 9 (Table 2), the residual acidity of the
chloroform should catalyze the in situ isomerization of 8e
into 5e and 8i into 5i. Compounds 5e and 5i are then photo-
isomerized into benzofurans 6e and 6i, respectively.
[15] No ring contraction was observed in the absence of irradia-
tion.
[16] For selected examples, see: a) E. A. Shilova, G. P›pe, A.
Samat, C. Moustrou, Tetrahedron 2008, 64, 9977; b) C. D.
Gabbutt, B. M. Heron, S. B. Kolla, C. Kilner, S. J. Coles,
P. N. Hortonc, M. B. Hursthouse, Org. Biomol. Chem. 2008,
6, 3096; c) J. N. Moorthy, P. Venkatakrishnan, S. Sengupta,
M. Baidya, Org. Lett. 2006, 8, 4891; d) J. N. Moorthy, P. Ven-
katakrishnan, S. Samanta, D. K. Kumar, Org. Lett. 2007, 9,
919.
[17] This observation tends to prove that the residual acidity of
the chloroform catalyzes the ring closure of 9 into 6, poten-
tially via activation of the diester moiety.
[18] In this case, 9a was converted into 6a during purification by
flash chromatography on silica gel.
[7] See Supporting Information for details.
[8] C. H. M. Amijs, V. López-Carrillo, V. M. Raducan, P. Pꢀrez-
Galµn, C. Ferrer, A. M. Echavarren, J. Org. Chem. 2008, 73,
7721.
Received: March 24, 2013
Accepted: May 13, 2013
Published online: July 2, 2013
922
www.ijc.wiley-vch.de
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Isr. J. Chem. 2013, 53, 915 – 922