A combined transition-metal-catalyzed and photopromoted process: Synthesis of 2,3-fused 4-phenylnaphthalen-1-yl carboxylates from 1,7-diaryl-1,6-diynes

Article abstract of DOI:10.1002/chem.201500978

2,3-Fused 4-phenylnaphthalen-1-yl carboxylates were synthesized in a step- and atom-economical manner using a ruthenium-catalyzed hydrocarboxylative cyclization of 1,7-diaryl-1,6-diynes and subsequent oxidative photocyclization. The scope of this novel two-step process was demonstrated by the construction of diverse structures from substrates with various tethers and terminal aryl groups. Late-stage C-H functionalizations of the arylnaphthalene product further enhance the synthetic potential of the developed process. Light a fuse! 2,3-Fused 4-phenylnaphthalen-1-yl carboxylates were synthesized in a step- and atom-economical manner using a ruthenium-catalyzed hydrocarboxylative cyclization of 1,7-diaryl-1,6-diynes and subsequent oxidative photocyclization. Late-stage C-H functionalizations of the obtained arylnaphthol derivatives further enhance the synthetic potential of the developed process (see scheme).

Full text of DOI:10.1002/chem.201500978

DOI: 10.1002/chem.201500978
Full Paper
&
Ruthenium Catalysis |Hot Paper|
A Combined Transition-Metal-Catalyzed and Photopromoted
Process: Synthesis of 2,3-Fused 4-Phenylnaphthalen-1-yl
Carboxylates from 1,7-Diaryl-1,6-diynes
Yoshihiko Yamamoto,* Shota Mori, and Masatoshi Shibuya[a]
Abstract: 2,3-Fused 4-phenylnaphthalen-1-yl carboxylates
were synthesized in a step- and atom-economical manner
using a ruthenium-catalyzed hydrocarboxylative cyclization
of 1,7-diaryl-1,6-diynes and subsequent oxidative photocycli-
zation. The scope of this novel two-step process was dem-
onstrated by the construction of diverse structures from sub-
strates with various tethers and terminal aryl groups. Late-
stage CÀH functionalizations of the arylnaphthalene product
further enhance the synthetic potential of the developed
process.
nase inhibitor (Figure 1).[9] Therefore, the development of
a novel expedient route to the diverse arylnaphthol derivatives
from readily available starting materials would make a signifi-
cant contribution to drug discovery.
Introduction
9-Phenyl-1,3-dihydronaphtho[2,3-c]furan-4-ol (1) is the basic
skeleton of arylnaphthalide lignan natural products, such as di-
phyllin,[1] justicidin A,[2] taiwanin E,[3] and neojusticidin A
(Figure 1).[4] Members of this lignan family display significant
Among the various methods for construction of the lignan-
type arylnaphthalene skeleton, intramolecular dehydro-Diels–
Alder reactions or Garratt–Braverman cyclization of 1,7-diaryl-
1,6-diynes, dehydro-Diels–Alder reactions/aromatization of 1,7-
diaryl-1,6-enynes, and gold-catalyzed cyclocondensation of a-
propargyloxy 2-propanones are the most straightforward.[10]
However, these routes are unable to introduce the naphthol
hydroxyl group. In contrast, Sato and co-workers achieved the
total synthesis of taiwanin E by using a palladium-catalyzed
[2+2+2] cycloaddition of a 1-aryl-1,6-diyne with a benzyne, al-
though multistep manipulations were also required for the in-
troduction of the naphthol hydroxyl group.[11] We envisaged
that the cycloaromatization of exocyclic 1,3-dienyl esters 3,
which can be prepared by the ruthenium-catalyzed stereose-
lective hydrocarboxylative cyclization of 1,7-diaryl-1,6-diynes 4
with carboxylic acids 5, would afford the desired arylnaphthol
esters 2 (Scheme 1a). This method would provide a novel
atom- and step-economical entry to the targeted arylnaphthol
scaffold because the required oxygen functionality can be in-
troduced by the catalytic hydrocarboxylative cyclization of
readily accessible 1,7-diaryl-1,6-diynes with carboxylic acids
before the cycloaromatization stage. In relation to this work,
Momose and co-workers have previously reported the photo-
chemical 6p-electrocyclization of a 2,3-dibenzylidenebutyrolac-
tone, which afforded apolignan 6a and its isomer
(Scheme 1b).[12] However, this method is limited to g-lactone
substrates. In addition, subsequent aromatization of 6a under
aerobic conditions led to an approximately 1:1 mixture of the
corresponding lignan 7a and its hydroxylated derivative 7b in
a moderate combined yield. Thus, the hydroxylation of the
naphthalene core is incomplete. Therefore, the efficient and
general methods to synthesize diverse-fused arylnaphthol de-
rivatives such as 2 have remained elusive.
Figure 1. Arylnaphthalide lignans and relevant bioactive molecules.
biological activities, including antiplatelet,[5] anti-inflammato-
ry,[6] antiviral,[7] and cytotoxic activities.[8] In addition, closely re-
lated natural and synthetic molecules with diverse significant
biological activities are also known, for example, antitumor/an-
timitotic podophyllotoxin, an anti-HIV agent, and a 5-lipoxyge-
[a] Prof. Dr. Y. Yamamoto, S. Mori, Dr. M. Shibuya
Department of Basic Medicinal Sciences
Graduate School of Pharmaceutical Sciences, Nagoya University
Chikusa, Nagoya 464-8601 (Japan)
E-mail: yamamoto-yoshi@ps.nagoya-u.ac.jp
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201500978.
Chem. Eur. J. 2015, 21, 9093 – 9100
9093
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
and selective protocol for the oxidative photocyclization lead-
ing to arylnaphthol derivatives. In fact, the oxidative photocyc-
lization of a stereoisomeric mixture of 3 resulted in a much
lower yield of the desired product (vide infra). However, the
transition-metal-catalyzed hydrocarboxylative cyclizations of
a,w-diynes have received scant attention and their synthetic
application has not yet been reported. During their study of
the ruthenium-catalyzed decarboxylative cyclization of 1,6-
diynes, Saµ and co-workers found that several 1,6- and 1,7-
diynes underwent hydrocarboxylative cyclization in acetic acid
as the solvent to afford exocyclic 1,3-dienyl acetates in various
yields and regio- and stereoselectivities.[16] Later, Tanaka and
co-workers developed a rhodium-catalyzed regio- and stereo-
selective hydrocarboxylative cyclization of unsymmetrical 1,6-
and 1,7-diynes that possess both internal and terminal alkynes,
affording similar exocyclic 1,3-dienyl carboxylates.[17] In contrast
to these examples, a different type of ruthenium-catalyzed hy-
drocarboxylative transformation of terminal 1,6-diynes to cyclo-
hexenylidene enol esters was reported by Lee and co-work-
ers.[18] They proposed a mechanism that involved alkylidene
carbene complexes as intermediates to explain this unique
chemoselectivity. Therefore, the chemo-, regio-, and stereose-
lectivity of the hydrocarboxylative cyclization of a,w-diynes
strongly depends on the substrate structure and catalyst used.
Therefore, at the outset of this study, we sought to identify op-
timal conditions for the unprecedented hydrocarboxylative
cyclization of 1,7-diaryl-1,6-diynes 4.
Scheme 1. a) Our synthetic plan for fused arylnaphthalenyl carboxylates 2
from diynes 4 and b) previous methods to synthesize lignans 7.
Although many methods have been devised for the synthe-
sis of substituted naphthalenes,[13] the oxidative photocycliza-
tion of 1-aryl-1,3-butadienes has scarcely been developed.[14]
This is in striking contrast to the similar oxidative photocycliza-
tion of stilbenes, which has been extensively studied and ap-
plied to the synthesis of phenanthrenes.[15] A seminal study by
Leznoff and Hayward reported that the oxidative photocycliza-
tion of (E,Z)-1,4-diaryl-1,3-butadienes, in which one of the two
aryl groups is a p-substituted benzene ring, was performed in
benzene for 3–6 days to afford a mixture of regioisomeric
products in low yields of 6–12%.[14a] This result suggests that
the 6p-electrocyclization is slower than facile E–Z alkene iso-
merization, which leads to a loss of regioselectivity. Later,
Olsen and co-workers introduced a fused-ring moiety to 1-
phenyl-1,3-butadiene to improve the cyclization efficacy.[14b] Ac-
cordingly, the E–Z isomerization was followed by cycloaromati-
zation to afford the desired fused naphthalene derivatives
within 6 h in acceptable yields of 45–74%. Therefore, the yield
and regioselectivity of the oxidative photocyclization of 1,4-
diaryl-1,3-butadienes has yet to be improved and its applica-
tion to the construction of fused arylnaphthalene scaffolds
such as 2 has not been attempted, to the best of our knowl-
edge. Herein, we report the results of our study on 1) the hy-
drocarboxylative cyclization of 1,7-diaryl-1,6-diynes 4, which
stereoselectively leads to exocyclic 1,3-dienyl carboxylates 3,
2) the oxidative photocyclization of 3, which selectively affords
fused arylnaphthyl esters 2, and 3) late-stage CÀH functionali-
zations of fused arylnaphthalene products.
Optimization of the hydrocarboxylative cyclization of 1,7-
diaryl-1,6-diynes
As a seminal study in this area, the prototypical hydrocarboxy-
lative dimerization of arylacetylenes was reported by Dixneuf
and co-workers.[19] This completely intermolecular three-com-
ponent reaction regio- and stereoselectively produces (1E,3E)-
1,4-diaryl-1,3-butadienyl carboxylates, and the distinct product
selectivity is as a result of a key ruthenacycle intermediate. To
shed light on each fundamental step that is involved in this
fascinating catalytic coupling, we carried out a theoretical
study on the mechanism using DFT calculations.[20] This study
theoretically confirmed the impact of electron-withdrawing or
-donating substituents of the terminal aryl groups and the
acidity of carboxylic acids on the reaction efficiency, which sup-
ported the mechanism proposed by Dixneuf and co-workers.
In addition, the feasibility of an elusive partially intramolecular
version of Dixneuf’s reaction, that is, the hydrocarboxylative
cyclization of 1,7-diphenyl-1,6-heptadiyne (4, X=CH2) with
acetic acid, was evaluated at the same level of theory, reveal-
ing that this partially intramolecular variant is less efficient
than the parent reaction of phenylacetylene, albeit with the
same stereoselectivity.
Taking the above facts in consideration, we first investigated
the reaction of ether-tethered diyne 4a as a representative
substrate with acetic acid (5a) to give the optimal reaction
conditions (Scheme 2). According to the report by Dixneuf,[19]
4a and 5a (1.2 equiv) were stirred in 1,4-dioxane in the pres-
ence of ruthenium catalyst [Cp*RuCl(cod)] (10 mol%) (Cp*=
Results and Discussion
With the above-mentioned background in mind, we consid-
ered that the development of a stereoselective protocol to
produce 3 would be highly beneficial to establish an efficient
Chem. Eur. J. 2015, 21, 9093 – 9100
www.chemeurj.org
9094
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
(Z,Z)-3aa. At a higher loading of 5a (10 equiv), the stereoselec-
tivity was almost completely lost (entry 5) and substantial
amounts of hydrochlorinative cyclization byproduct 8a were
formed under the neat conditions (entry 6). The erosion of the
stereoselectivity and formation of 8a are presumably responsi-
ble for the generation of a cationic ruthenium species by dis-
sociation of the chloride ligand in protic media. As previously
proposed,[19,20] the present hydrocarboxylative cyclization of 4a
is considered to proceed by protonation of one of the two
a carbon atoms in ruthenabicycle 11 a and subsequent addi-
tion of an acetate anion to the other a carbon atom of 12a
(Scheme 3). The resulting ruthenacyclopentene 13a subse-
Scheme 2. Hydrocarboxylative cyclization of diyne 4a with acetic acid (5a).
1,2,3,4,5-pentamethylcyclopentadienyl,
cod=1,5-cycloocta-
diene) at room temperature at varied concentrations of 4a
(Table 1, entries 1–3). As a result, the reaction carried out at
Table 1. Optimization of hydrocarboxylative cyclization.
Entry
[4a] [m]
5a [equiv]
T [8C], t [h]
3aa/8a/9a/10a[a]
1
2
3
4
5
6
7[d]
8[e]
9[e]
10[f]
1.0
0.5
0.25
0.5
0.5
0.5
0.5
0.5
0.5
0.5
1.2
1.2
1.2
5
RT, 24 h
RT, 20 h
RT, 24 h
RT, 3 h
RT, 1 h
RT, 2 h
908C, 24 h
RT, 24 h
508C, 3 h
708C, 4 h
95 (6:1):1:0:4[b]
90 (17:1):3:5:2
89 (44:1):0:2:9[b]
99 (3:1):1:0:0
99 (1:1):1:0:0
88 (2:1):12:0:0
99 (1:1):0:1:0[b]
86 (1:0):2:9:3
97 (1:0):0:0:3
94 (46:1):1:0:5
10
35[c]
35[c]
1.2
1.2
1.2
Scheme 3. Plausible mechanism for formation of stereoisomers of 3aa.
quently undergoes conrotatory ring opening in such a way
that the acetate substituent moves outward and the electron
repulsion between this substituent and the chloride ligand is
mitigated. As a result, the desired product (Z,Z)-3aa would be
formed. On the contrary, the absence of the chloride ligand
would lead to the less-selective formation of both isomers by
cationic ruthenacycles 14a and 15a. To corroborate this hy-
pothesis, the catalytic behavior of a cationic ruthenium catalyst
was examined under similar conditions. Because in situ ab-
straction of the chlorine ligand is a convenient method to gen-
erate a cationic catalyst species, we then performed the reac-
tion in the presence of [Cp*RuCl(cod)] (10 mol%) and AgPF6
(20 mol%), resulting in almost no reaction after 24 h. Thus, the
reaction was carried out at an elevated temperature (908C) for
24 h, resulting in the formations of (Z,Z)-3aa and (E,Z)-3aa in
a 1:1 ratio without formation of 8a (entry 7). These facts sug-
gest that at higher concentrations of acetic acid 5a, the chlo-
ride ligand of the catalyst dissociates to generate a less-catalyt-
ically active cationic species, which results in the lower reaction
efficiency and the lack of stereoselectivity.
[a] Product ratios were determined by 1H NMR spectroscopic analyses of
crude reaction mixtures. The Z,Z/E,Z-stereoisomer ratios of 3aa are given
in parenthesis. [b] Conversion of 4a was ca. 70% for entries 1 and 3, and
ca. 60% for entry 7. [c] Neat conditions. [d] AgPF6 (20 mol%) was added.
[e] TBAC (10 mol%) was added. [f] [Cp*RuCl(cod)] (5 mol%) and TBAC
(5 mol%) were used.
0.5m concentration proved to be optimal (entry 2) because
the reactions did not complete within 24 h at lower or higher
concentrations (entries 1 and 3). At 0.5m concentration, the
desired product (Z,Z)-3aa was isolated in 68% yield, although
small amounts of its stereoisomer (tentatively assigned as (E,Z)-
3aa), hydrochlorinative cyclization product 8a, bicyclic furan
1
9a, as well as diyne cyclodimer 10a were detected by H NMR
spectroscopic analysis of the crude reaction mixture. Notably,
single-crystal X-ray diffraction analysis of 8a revealed that hy-
drochlorinative cyclization occurred with opposite stereoselec-
tivity to the hydrocarboxylative cyclization to afford (Z,Z)-3aa
as the major isomer (Figure S1 in the Supporting Informa-
tion).[21] Bicyclic furan 9a was identical to the product obtained
from our previously reported transfer oxygenative cyclization
of 4a.[22]
To inhibit byproduct formation, the reactions were repeated
with increased loadings of acetic acid 5a (Table 1, entries 4
and 5). With five equivalents of 5a, the reaction reached com-
pletion after 3 h and the formation of 9a and 10a was effec-
tively suppressed (entry 4). However, the stereoselectivity of
the hydrocarboxylative cyclization decreased to 3:1 in favor of
Therefore, suppression of the chloride ligand dissociation is
the key to control the stereoselectivity. To this end, tetrabutyl-
ammonium chloride (nBu4NCl, TBAC) was used as an additive.
In the presence of [Cp*RuCl(cod)] (10 mol%) and TBAC
(10 mol%), 4a and 5a (1.2 equiv) were stirred in 1,4-dioxane at
room temperature (Table 1, entry 8). Although a longer reac-
tion time of 24 h was required for full conversion of 4a, hydro-
carboxylative cyclization product (Z,Z)-3aa was predominantly
formed, together with small amounts of 8a, 9a, and 10a. Pu-
rification with silica gel column chromatography afforded (Z,Z)-
Chem. Eur. J. 2015, 21, 9093 – 9100
www.chemeurj.org
9095
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
3aa in 64% isolated yield. Furthermore, the same reaction was
repeated at a higher temperature (508C) to shorten the reac-
tion time to 3 h (entry 9). As a result, byproduct formation was
effectively suppressed and the isolated yield of (Z,Z)-3aa was
improved to 76%. With a decreased catalyst loading (5 mol%),
the substrate conversion was about 90%, even after reaction
for 24 h. At an elevated temperature of 708C, the reaction
reached completion within 4 h (entry 10). However, slight ero-
sion of the stereoselectivity for hydrocarboxylative cyclization
and an increase in the formation of 10a resulted in a dimin-
ished isolated yield of (Z,Z)-3aa (72%). Therefore, the condi-
tions of described in Table 1, entry 9 are optimal.
Table 2. Scope of diyne substrates in hydrocarboxylative cyclization.
With the optimal reaction conditions in hand, we then inves-
tigated the reaction of diyne 4a with a variety of carboxylic
acids (see Table S1 in the Supporting Information). The hydro-
carboxylative cyclization proceeded uneventfully with aliphatic
carboxylic acids including dichloroacetic acid and N-tert-but-
oxycarbonyl (N-Boc) phenylalanine to afford the corresponding
products in 65–81% yields. Aromatic carboxylic acids and cin-
namic acid also proved to be effective, resulting in the forma-
tion of the corresponding products in good yields. It is note-
worthy that electron-donating or -withdrawing groups at the
para-position of the benzoic acid have no influence on the
product yields. However, the reaction was facilitated by more
acidic p-nitrobenzoic acid (5h). A single crystal of the corre-
sponding product (Z,Z)-3ah suitable for X-ray diffraction analy-
sis was obtained and the stereochemistry was unambiguously
confirmed (Figure S1 in the Supporting Information).
Scope of the diynes in the hydrocarboxylative cyclization
[a] NMR spectroscopic yield. [b] 5a (3.6 equiv). [c] Slow addition of 4m
for 1 h. [d] TBAC (20 mol%).
To establish the scope and limitations of the diyne substrate 4
on the reaction, the influence of the tether moiety was investi-
gated by performing the reactions of 1,7-diaryl-1,6-diynes 4b–
k with acetic acid (5a) as summarized in Table 2. Under the
standard conditions, diyne 4b with a tosylamide tether was
subjected to hydrocarboxylative cyclization. However, the re-
sulting byproduct from self-dimerization of 4b by intermolecu-
lar [2+2+2] cycloaddition hampered the separation of the de-
sired product (Z,Z)-3ba. Thus, the loading of 5a was increased
(3.6 equiv) to suppress the side reaction. As a result, the forma-
tion of the dimer of 4b was reduced and (Z,Z)-3ba was ob-
tained in 64% yield after recrystallization. The reaction of diyne
4c, which has an all-carbon tether with a malonate moiety,
was also found to be sluggish but an elevated reaction tem-
perature of 808C improved the efficiency without loss of selec-
tivity, affording (E,E)-3ca in 71% yield.
Similarly, the reactions of Meldrum’s acid derivative 4d, ace-
tylacetone derivative 4e, cyclohexa-1,3-dione derivative 4 f,
and barbituric acid derivative 4g were performed at 808C to
afford the corresponding products in 64–81% yields. The car-
bonyl group on the tether is unnecessary because fluorene de-
rivative 4h was also transformed into (E,E)-3ha in a comparable
yield to the previous examples. These results demonstrate the
tolerance of this protocol toward various functional groups
(ether, ester, ketone, amide, and sulfonamide). On the other
hand, diyne 4i with a sulfide tether was an inefficient substrate
due to the high coordinating ability of the sulfide moiety as
well as longer CÀS bonds: the reaction of 4i carried out with
[Cp*RuCl(cod)] (20 mol%) and TBAC (20 mol%) for 24 h led to
a complex mixture. The reaction of ester-tethered diyne 4j was
also sluggish at room temperature and heating at 508C led to
an intractable complex mixture. The reaction of 1,7-diyne 4k
was also examined under the standard conditions but resulted
in no reaction.
The hydrocarboxylative cyclization of diynes 4l–q was also
performed to investigate the influence of terminal aryl groups.
When the diyne substrate possessed a moderately electron-
withdrawing p-chlorophenyl terminal, the undesired cyclodi-
merization of 4l competed with hydrocarboxylative cyclization.
Therefore, the reaction of 4l was performed with an increased
loading of 5a (3.6 equiv) at 508C for 3 h, affording the desired
(Z,Z)-3la in 64% yield. Because cyclodimerization was more
pronounced in the reaction of diyne 4m, which had a highly
electron-withdrawing p-trifluoromethylphenyl terminal, slow
addition by a syringe pump for 1 h was required to selectively
Chem. Eur. J. 2015, 21, 9093 – 9100
www.chemeurj.org
9096
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
obtain (Z,Z)-3ma in a comparable yield. The reaction of diyne
4n, which had an electron-donating p-methoxyphenyl termi-
nal, was sluggish under the standard conditions. Thus, the re-
action temperature was raised to 808C and the loading of
TBAC was increased to minimize erosion of the stereoselectiv-
ity. As a result, (Z,Z)-3na was obtained in 62% yield deter-
mined by NMR spectroscopy. However, the purification of (Z,Z)-
3na was hampered by an inseparable and unidentifiable by-
product. In contrast, the reaction of 3,4,5-trimethoxyphenyl de-
rivative 4o and benzo[1,3]dioxol-5-yl derivative 4p under the
same conditions used for 4n afforded (Z,Z)-3oa and (Z,Z)-3pa
in 69 and 62% isolated yields. When the terminal group was
the smaller 2-thienyl ring, cyclodimerization became competi-
tive again. Thus, the reaction of 1,6-diyne 4q was conducted
with an increased amount of 5a (3.6 equiv). Although a pro-
longed reaction time of 10 h was required, the desired product
(Z,Z)-3qa was obtained in 54% yield.
report of Bedekar and co-workers.25 Thus, a solution of (Z,Z)-
3aa (0.03m) in a mixed solvent of toluene and acid scavenger
THF (20:1) was irradiated otherwise under the same conditions.
Although a longer irradiation time of 22 h was required, the
product yield was successfully improved to 86% (entry 2). In-
creasing the light intensity (300 W) shortened the reaction
time to 8 h and afforded a slightly lower yield of 2aa (entry 3).
Further examination of the amount of THF showed that the
optimal ratio of toluene/THF was 4:1 and 2aa was obtained in
89% yield (entries 4 and 5).
The importance of the fused (Z,Z)-dienyl acetate moiety was
clearly demonstrated by the following control experiments.
The yield of 2aa was considerably lower under the same reac-
tion conditions when an approximately 2:1 stereoisomeric mix-
ture of 3aa was used as the substrate (entry 6). In this case,
3aa was fully consumed and unidentifiable byproducts were
formed. The reaction of a Dixneuf’s product, derived from phe-
nylacetylene and acetic acid [(1E,3E)-1,4-diphenyl-1,3-butadien-
yl acetate], failed to afford the corresponding oxidative photo-
cyclization product, which indicated the importance of the
fused-ring moiety. Pivalate (Z,Z)-3ad also underwent oxidative
photocyclization in a shorter reaction time of 6 h under the
optimized conditions, affording 2ad, albeit in a diminished
yield of 70% (entry 7). In contrast, 2ag was formed in about
50% yield along with deacylation product 1 when benzoate
(Z,Z)-3ag was the substrate under the same conditions. These
results indicate that a less functionalized carboxylate moiety is
preferable for the oxidative photocyclization.
Oxidative photocyclization of the hydrocarboxylative cycli-
zation products
After having developed the stereoselective hydrocarboxylative
cyclization of 1,7-diaryl-1,6-diynes 4, we set out to explore the
cycloaromatization of the obtained products (Z,Z)-3. Because
photochemical 6p-electrocyclizations have been successfully
applied to the synthesis of polycyclic aromatic mole-
cules,[14,23,24] we decided to adapt known oxidative photocycli-
zation conditions to the hydrocarboxylative cyclization prod-
ucts (Z,Z)-3 (Table 3). In the presence of molecular iodine
The scope of the oxidative photocyclization was examined
under the optimal conditions (Table 4). Tosylamide derivative
(Z,Z)-3ba and malonate derivative (E,E)-3ca underwent smooth
oxidative photocyclization, albeit with a longer irradiation time
of 10 h, affording 2ba and 2ca in 87 and 82% yields, respec-
tively. The reaction of Meldrum’s acid derivative (E,E)-3da,
which is poorly soluble in the toluene/THF mixed solvent, led
to decomposition. Therefore, the reaction was performed in
CH2Cl2/THF for 8 h to give the expected product 2da in 42%
yield, along with the deacetalization side product 16 (22%).
Acetylacetone derivative (E,E)-3ea was successfully converted
to 2ea in 79% yield under the standard conditions, whereas
cyclohexane-1,3-dione derivative (E,E)-3 fa afforded 2 fa in
a low yield (29%). In the latter case, keto acid 17 was also ob-
tained in 51% yield. According to the report,[26] 17 was pre-
sumably formed from 2 fa by a Norrish type I fragmentation/
cyclization and subsequent hydrolysis of the resulting enol
ether (Scheme S1 in the Supporting Information). In contrast,
spirocyclic products 2ga and 2ha were obtained uneventfully
in 71 and 81% yields from barbituric acid derivative (E,E)-3ga
or fluorene derivative (E,E)-3ha, respectively.
Table 3. Optimization of oxidative photocyclization.
Entry
R
Conditions[a]
2, Yield [%][b]
1
2
3
4
Me
Me
Me
Me
Me
Me
tBu
Ph
benzene/PO (35:1), 75 W, 1 h
toluene/THF (20:1), 75 W, 22 h
toluene/THF (20:1), 300 W, 8 h
toluene/THF (4:1), 300 W, 8 h
toluene/THF (2:1), 300 W, 8 h
toluene/THF (4:1), 300 W, 8 h
toluene/THF (4:1), 300 W, 6 h
toluene/THF (4:1), 300 W, 8 h
2aa, 66
2aa, 86
2aa, 80
2aa, 89
2aa, 87
2aa, 57
2ad, 70
2ag[d]
5
6[c]
7
8
[a] PO=propylene oxide. [b] Isolated yields. [c] A 2:1 mixture of (Z,Z)-3aa
and (E,Z)-3aa was used. [d] 2ag was obtained in about 50% yield along
with 1.
Furthermore, the oxidative photocyclizations of the hydro-
carboxylative cyclization products with substituted phenyl
rings were investigated. Chlorinated and trifluoromethylated
naphthalene derivatives 2la and 2ma were obtained in 76 and
72% yields from (Z,Z)-3la and (Z,Z)-3ma, respectively. Com-
pound (Z,Z)-3na, which included a small amount of an insepa-
rable impurity, was directly irradiated to afford 2na in 24%
yield in two steps from diyne 4n. In contrast, the reaction of
(1.1 equiv) as the oxidant, a solution of (Z,Z)-3aa in a mixed
solvent of benzene and acid scavenger propylene oxide (35:1)
was irradiated using a xenon monochromatic light source
(75 W). As a result, (Z,Z)-3aa was consumed within 1 h and the
desired product 2aa was obtained in 66% yield after chroma-
tographic purification (Table 3, entry 1). Then, we performed
the reaction in a different mixed solvent according to the
Chem. Eur. J. 2015, 21, 9093 – 9100
www.chemeurj.org
9097
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 4. Scope of oxidative photocyclization.
Scheme 4. Benzylic CÀH oxidation of 1 and 2aa with oxoammonium salt 18.
products were determined by NMR spectroscopy. It was report-
ed that lactone methylene protons of type 20 resonate at
higher magnetic fields than those of type 19 due to the shield-
ing effect of the proximal phenyl ring (d=5.08–5.23 vs. 5.32–
5.52 ppm).[28] In fact, the methylene protons of 19 and 20 ap-
peared at d=5.35 and 5.22 ppm, respectively. The assignment
of 19 was also unambiguously confirmed by X-ray diffraction
(Figure S1 in the Supporting Information). The oxidation of
2aa was also carried out under different conditions. The use of
a combination of KBr (2 equiv) and oxone (4 equiv) as the oxi-
dant in aqueous acetonitrile at 458C for 5 h resulted in dimin-
ished yields of 19 and 20 (35 and 16% yields, respectively).[29a]
Similar methods that used iodosobenzene instead of oxone or
sodium o-iodobenzenesulfonate (10 mol%) instead of KBr
under phase-transfer conditions were also applied to the oxi-
dation of 2aa, resulting in lower yields and selectivity.[29b,c]
The oxidation of naphthol 1, which was obtained in a high
yield by deacetylation of 2aa by refluxing in ethanol with am-
monium acetate,[30] was also examined under the same condi-
tions. The reaction was complete in 2 h and lactone 22 was
obtained in 57% yield along with side product 23. The regioi-
someric lactone product 21, which can be obtained by deace-
tylation of 19, was not observed. The oxidation regiochemistry
was inferred by the fact that the chemical shift of the benzylic
proton of 22 (d=5.29 ppm) was similar to that of 20, and the
deacetylation product obtained from 20 was identical to 22.
Side product dial 23 was assigned by the observation of two
[a] CH2Cl2 was used as the co-solvent instead of toluene. [b] Compound
16 was obtained in 22% yield. [c] Compound 17 was obtained in 51%
yield. [d] Yield from diyne 4o.
3,4,5-trimethoxyphenyl analogue (Z,Z)-3oa led to the higher
yield (61%) formation of 2oa, although a longer irradiation
time was required for completion of the reaction. Similarly
benzo[1,3]dioxol-5-yl derivative (Z,Z)-3pa underwent oxidative
cyclization to afford an inseparable mixture of regioisomers
2pa in 49% yield with a 3:2 selectivity. 2-Thienyl analogue
(Z,Z)-3qa was also subjected to oxidative photocyclization for
24 h, affording benzothiophene derivative 2qa in 69% yield.
1
formyl groups as well as the phenylnaphthalene core in the H
and 13C NMR spectra.
To introduce substituents on the naphthalene nucleus, we
next investigated the transition-metal-catalyzed functionaliza-
tion of aromatic CÀH bonds with the hydroxyl directing group.
To this end, Ackermann’s protocol was applied to naphthol
1 (Scheme 5).[31] Thus, 1 was treated with diphenylacetylene
(1 equiv) in the presence of [(p-cymene)RuCl2]2 (2 mol%) and
Cu(OAc)2·H2O (2 equiv) in m-xylene at 808C for 23 h, affording
8,10-dihydrobenzo[de]furo[3,4 h]chromene derivative 24 in
68% yield. The oxidative Heck-type reaction of 1 was also per-
Late-stage CÀH functionalization of arylnaphthol derivatives
Finally, we investigated the derivatization of representative
product 2aa by several CÀH functionalizations. First, the syn-
thesis of arylnaphthalides by benzylic CÀH oxidation of 2aa
was investigated (Scheme 4). According to the report,[27] 2aa
was treated with oxoammonium salt 18 (2.5 equiv) in acetoni-
trile/H2O (9:1) at 508C for 8 h, affording lactones 19 and 20 in
58 and 18% yields, respectively. The regiochemistries of these
Chem. Eur. J. 2015, 21, 9093 – 9100
www.chemeurj.org
9098
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
1747 cmÀ1 (C=O); HRMS (ESI): m/z: calcd for C20H18O3·Na 329.1154;
found: 329.1169 [M+Na]+.
General procedure for oxidative photocyclization—Synthesis of
2aa
In a quarts tube, a solution of dienyl ester (Z,Z)-3aa (91.7 mg,
0.299 mmol) and I2 (83.5 mg, 0.329 mmol) in a dry degassed mixed
solvent of toluene (10 mL) and THF (2.5 mL) was irradiated with
a xenon monochromatic light source (300 W) under an argon at-
mosphere at room temperature for 8 h. The reaction mixture was
washed with sat. Na2S2O3 (10 mL) and the aqueous phase was ex-
tracted with AcOEt (3”10 mL). The combined organic layer was
washed with H2O (20 mL) and brine (20 mL), and then dried over
MgSO4. The solvents were evaporated in vacuo and the obtained
crude product was purified by silica gel column chromatography
(hexane/AcOEt=15:1) to afford 2aa (81.2 mg, 89% yield) as a color-
Scheme 5. Aromatic CÀH alkenylation of arylnaphthol 1.
1
less solid (m.p. 116.5–119.58C); H NMR (400 MHz, CDCl3, 258C): d=
formed using the protocol reported by Du, Li, and co-workers
(Scheme 5).[32] In the presence of [Cp*RhCl2]2 (3 mol%) and
Cu(OAc)2 (2.2 equiv), 1 and ethyl acrylate (2 equiv) were heated
in DMF at 1008C for 1 h to afford the desired product 25 in
74% yield.
2.49 (s, 3H), 5.03 (s, 2H), 5.19 (s, 2H), 7.33–7.53 (m, 7H), 7.68 (d, J=
8.4 Hz, 1H), 7.91 ppm (d, J=8.4 Hz, 1H); 13C NMR (100 MHz, CDCl3,
258C): d=20.7, 71.8, 73.3, 120.8, 126.00, 126.03, 126.2, 126.8, 127.7,
128.6, 129.5, 130.7, 133.6, 137.6, 137.7, 138.9, 168.4 ppm; IR (neat):
n˜ =1770 cmÀ1 (C=O); HRMS (DART): m/z: calcd for C20H16O3·NH4:
322.1443; found: 322.1429 [M+NH4]+.
Conclusion
Acknowledgements
We have developed a novel two-step process for the synthesis
of highly valuable 2,3-fused 4-phenylnaphthalen-1-yl carboxy-
lates from 1,7-diaryl-1,6-diynes. The first hydrocarboxylative
cyclization of 1,7-diaryl-1,6-diynes was optimized for the highly
stereoselective formation of exocyclic dienyl acetates by using
the ruthenium catalyst [Cp*RuCl(cod)] with nBu4NCl as an addi-
tive. A subsequent oxidative photocyclization of the resulting
exocyclic dienyl acetates efficiently afforded the desired 2,3-
fused 4-phenylnaphthalen-1-yl acetates when the reactions
were performed with I2 in a mixed solvent of toluene and THF.
The synthetic scope of this novel process was demonstrated
by the construction of diverse structures from substrates with
various tethers and terminal aryl groups. Moreover, the late-
stage CÀH functionalizations of the arylnaphthalene products
further increase the synthetic potential of the developed se-
quential process.
This work was partially supported by Grant-in-Aid for Scientific
Research on Platform for Drug Discovery, Informatics, and
Structural Life Sciences from the Ministry of Education, Culture,
sports and Technology, Japan, and the Naito foundation. We
thank Dr. K. Yamashita for X-ray crystallographic analysis.
Keywords: cyclization · lignan · naphthol · photoreaction ·
ruthenium catalysis
[1] a) T. R. Govindachari, S. S. Sathe, N. Viswanathan, B. R. Pai, M. Srinivasan,
Tetrahedron 1969, 25, 2815–2821; b) A. S. R. Anjaneyulu, P. A. Ramaiah,
L. R. Row, R. Venkateswarlu, Tetrahedron 1981, 37, 3641–3652.
[2] N. Fukamiya, K.-H. Lee, J. Nat. Prod. 1986, 49, 348–350.
[3] Y.-T. Lin, T.-B. Lo, K.-T. Wang, B. Weinstein, Tetrahedron Lett. 1967, 849–
852.
[4] C.-C. Chen, W.-C. Hsin, F.-N. Ko, Y.-L. Huang, J.-C. Ou, C.-M. Teng, J. Nat.
Prod. 1996, 59, 1149–1150.
[5] J.-R. Weng, H.-H. Ko, T.-L. Yeh, H.-C. Lin, C.-N. Lin, Arch. Pharm. Pharm.
Med. Chem. 2004, 337, 207–212.
Experimental Section
[6] S.-H. Day, N.-Y. Chiu, L.-T. Tsao, J.-P. Wang, C.-N. Lin, J. Nat. Prod. 2000,
63, 1560–1562.
General procedure for ruthenium-catalyzed hydrocarboxyla-
tive cyclization—Synthesis of (Z,Z)-3aa
[7] J. Asano, K. Chiba, M. Tada, T. Yoshii, Phytochemistry 1996, 42, 713–717.
[8] a) S.-H. Day, N.-Y. Chiu, S.-J. Won, C.-N. Lin, J. Nat. Prod. 1999, 62, 1056–
1058; b) S.-T. Chang, D. S.-Y. Wang, C.-L. Wu, S.-G. Shiah, Y.-H. Kuo, C.-J.
Chang, Phytochemistry 2000, 55, 227–232; c) S.-H. Day, Y.-C. Lin, M.-L.
Tsai, L.-T. Tsao, H.-H. Ko, M.-I. Chung, J.-C. Lee, J.-P. Wang, S.-J. Won, C.-N.
Lin, J. Nat. Prod. 2002, 65, 379–381; d) C. Ramesh, N. Ravinndranath,
T. S. Ram, B. Das, Chem. Pharm. Bull. 2003, 51, 1299–1300; e) S. Susplu-
gas, N. Van Hung, J. Bignon, O. Thoison, A. Kruczynski, T. SØvenet, F.
GuØritte, J. Nat. Prod. 2005, 68, 734–738; f) J.-C. Lee, C.-H. Lee, C.-L. Su,
C.-W. Huang, H.-S. Liu, C.-N. Lin, S.-J. Won, Carcinogenesis 2005, 26,
1716–1730; g) J. Hui, Y. Zhao, L. Zhu, Med. Chem. Res. 2012, 22, 3944–
4001.
[9] For reviews of lignan biological activities, see: a) W. D. MacRae, G. H. N.
Towers, Phytochemistry 1984, 23, 1207–1220; b) J. L. Charlton, J. Nat.
Prod. 1998, 61, 1447–1451; c) R. S. Ward, Nat. Prod. Rep. 1999, 16, 75–
96; d) M. Gordaliza, M. C. Castro, J. M. M. del Corral, A. S. Feliciano, Curr.
Pharm. Des. 2000, 6, 1811–1839.
A
solution of [Cp*RuCl(cod)] (19.2 mg, 0.051 mmol), nBu4NCl
(14.1 mg, 0.051 mmol), diyne 4a (123 mg, 0.50 mmol), and AcOH
(5a) (345 mL, 0.60 mmol) in dry degassed 1,4-dioxane (1 mL) was
stirred under an argon atmosphere at 508C for 3 h. After having
been cooled to room temperature, the reaction mixture was con-
centrated in vacuo and the crude material was purified by silica
gel column chromatography (hexane/AcOH=50:1–15:1) to give
(Z,Z)-3aa (116.2 mg, 76% yield) as a yellow solid (m.p. 139.8–
1
143.78C): H NMR (400 MHz, CDCl3, 258C): d=2.15 (s, 3H), 4.60 (s,
2H), 4.81 (d, J=2.8 Hz, 2H), 6.34 (t, J=2.8 Hz, 1H), 6.92 (d, J=
7.6 Hz, 2H), 7.16–7.21 (m, 1H), 7.25–7.30 (m, 2H), 7.35–7.44 (m,
3H), 7.55–7.59 ppm (m, 2H); 13C NMR (100 MHz, CDCl3, 258C): d=
20.7, 70.3, 72.6, 123.9, 127.2, 128.1, 128.36, 128.44, 128.8, 129.1,
129.2, 134.6, 134.8, 136.8, 139.9, 168.4 ppm; IR (neat): n˜ =
Chem. Eur. J. 2015, 21, 9093 – 9100
www.chemeurj.org
9099
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[10] For selected recent reports, see: a) P. Foley, N. Eghbali, P. T. Anastas,
Green Chem. 2010, 12, 888–892; b) S. Mondal, M. Maji, A. Basak, Tetra-
hedron Lett. 2011, 52, 1183–1186; c) T. Kudoh, A. Shihido, K. Ikeda, S.
Saito, T. Ishikawa, Synlett 2013, 24, 1509–1512; d) L. S. Kocsis, K. M.
Brummond, Org. Lett. 2014, 16, 4158–4161; e) V. Gudla, R. Balamuru-
gan, J. Org. Chem. 2011, 76, 9919–9933.
[11] a) Y. Sato, T. Tamura, M. Mori, Angew. Chem. Int. Ed. 2004, 43, 2436–
2440; Angew. Chem. 2004, 116, 2490–2494; b) Y. Sato, T. Tamura, A. Kin-
bara, M. Mori, Adv. Synth. Catal. 2007, 349, 647–661.
[12] a) T. Momose, K. Kanai, K. Hayashi, Chem. Pharm. Bull. 1978, 26, 3195–
3198; also, see: b) Y.-T. Lin, T.-B. Lo, K.-T. Wang, Tetrahedron Lett. 1976,
17, 50–51.
[13] C. B. de Koning, A. L. Rousseau, W. A. L. van Otterlo, Tetrahedron 2003,
59, 7–36.
[22] K. Yamashita, Y. Yamamoto, H. Nishiyama, J. Am. Chem. Soc. 2012, 134,
7660–7663.
[23] For reviews of photocyclizations, see: a) D. De Keukeleire, S.-L. He,
Chem. Rev. 1993, 93, 359–380; b) T. Bach, J. P. Hehn, Angew. Chem. Int.
Ed. 2011, 50, 1000–1045; Angew. Chem. 2011, 123, 1032–1077.
[24] For selected recent applications to natural product synthesis, see: a) A.
Fürstner, M. M. Domostoj, B. Scheiper, J. Am. Chem. Soc. 2006, 128,
8087–8094; b) Z. Lu, Y. Li, A. Li, Nat. Chem. 2013, 5, 679–684.
[25] H. R. Talele, M. J. Gohil, A. V. Bedekar, Bull. Chem. Soc. Jpn. 2009, 82,
1182–1186.
[26] H. Nozaki, Z. Yamaguchi, R. Noyori, Tetrahedron Lett. 1965, 6, 37–39.
[27] P. P. Pradhan, J. M. Bobbitt, W. F. Baily, J. Org. Chem. 2009, 74, 9524–
9527.
[28] Z. Horii, M. Tsujiuchi, K. Kanai, T. Momose, Chem. Pharm. Bull. 1977, 25,
1803–1808.
[14] a) C. C. Leznoff, R. J. Hayward, Can. J. Chem. 1970, 48, 1842–1849;
b) R. J. Olsen, J. C. Minniear, W. M. Overton, J. M. Sherrick, J. Org. Chem.
1991, 56, 989–991.
[15] a) F. B. Mallory, C. W. Mallory, Org. React. 1984, 30, 1–456; b) K. B. Jør-
gensen, Molecules 2010, 15, 4334–4358.
[29] a) L. Yin, J. Wu, J. Xiao, S. Cao, Tetrahedron Lett. 2012, 53, 4418–4421;
b) T. Dohi, N. Takenaga, A. Goto, H. Fujioka, Y. Kita, J. Org. Chem. 2008,
73, 7365–7368; c) L.-Q. Cui, K. Liu, C. Zhang, Org. Biomol. Chem. 2011,
9, 2258–2265.
[16] C. Gonzµlez-Rodríguez, J. A. Varela, L. Castedo, C. Saµ, J. Am. Chem. Soc.
2007, 129, 12916–12917.
[30] T. Kawamura, M. Hayashi, R. Mukai, J. Terao, H. Nemoto, Synthesis 2012,
44, 1308–1314.
[17] K. Tanaka, S. Saitoh, H. Hara, Y. Shibata, Org. Biomol. Chem. 2009, 7,
4817–4820.
[31] V. S. Thirunavukkarasu, M. Donati, L. Ackermann, Org. Lett. 2012, 14,
3416–3419.
[18] H. Kim, S. D. Goble, C. Lee, J. Am. Chem. Soc. 2007, 129, 1030–1031.
[19] a) J. Le Paih, S. DØrien, P. H. Dixneuf, Chem. Commun. 1999, 1437–1438;
b) J. Le Paih, F. Monnier, S. DØrien, P. H. Dixneuf, E. Clot, O. Eisenstein, J.
Am. Chem. Soc. 2003, 125, 11964–11975.
[32] F. Wang, G. Song, Z. Du, X. Li, J. Org. Chem. 2011, 76, 2926–2932.
[20] Y. Yamamoto, Organometallics 2013, 32, 5201–5211.
[21] The hydrochlorinative dimerization of arylalkynes was reported, see: H.
Klein, T. Roisnel, C. Bruneau, S. DØrien, Chem. Commun. 2012, 48,
11032–11034.
Received: March 11, 2015
Published online on May 15, 2015
Chem. Eur. J. 2015, 21, 9093 – 9100
www.chemeurj.org
9100
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim