Rhodium-Catalyzed Cascade Synthesis of Benzofuranylmethylidene-Benzoxasiloles: Elucidating Reaction Mechanism and Efficient Solid-State Fluorescence

Article abstract of DOI:10.1002/chem.201800381

A new synthetic route to highly fluorescent benzofuranylmethylidenebenzoxasiloles through cationic rhodium(I)/binap complex-catalyzed cascade cycloisomerization of bis(2-ethynylphenol)silanes has been developed involving 1,2-silicon and 1,3-carbon (alkyne) migrations followed by oxycyclization. The present synthesis requires only three steps, starting from commercially available dichlorodiisopropylsilane, which is markedly shorter than our previous synthesis (eight steps starting from commercially available chlorodiisopropylsilane). Theoretical calculations elucidated the mechanism of the above cascade cycloisomerization. This reaction is initiated by the formation of a rhodium vinylidene not through direct 1,2-silicon migration but rather through an unprecedented stepwise 1,5-silicon migration followed by C?Si bond-forming cyclization from a dearomatized allenylrhodium complex. Subsequent 1,3-carbon (alkyne) migration leading to a η³-allenyl/propargyl-rhodium complex followed by oxycyclization through π-bond (alkyne) activation with the cationic rhodium(I) complex affords the benzofuranylmethylidenebenzoxasilole product. The structure–fluorescence property relationships of the thus obtained benzofuranylmethylidenebenzoxasiloles were investigated, which revealed that good fluorescence quantum yields were generated in the solution state (φ_F=69–87 %) by introduction of electron-donating alkyl and phenyl groups on two phenoxy groups. In the powder state, 4-methyl- and 4-methoxy-phenoxy derivatives exhibited efficient blue fluorescence (φ_F=52 % and 46 %, respectively). Especially, the 4-methylphenoxy derivative was thermally stable, and exhibited strong narrow-band fluorescence in the film state (blue, φ_F=95 %) and redshifted strong narrow-band fluorescence (green, φ_F=90 %) in the crystalline state as a result of the formation of an offset π-stacked dimer; the latter was confirmed by X-ray crystallographic analysis and by theoretical calculations.

Full text of DOI:10.1002/chem.201800381

A Journal of
Accepted Article
Title: Rhodium-Catalyzed Cascade Synthesis of
Benzofuranylmethylidene-Benzoxasiloles: Elucidating Reaction
Mechanism and Efficient Solid-State Fluorescence
Authors: Tomoya Namba, Yoshihiro Hayashi, Susumu Kawauchi, Yu
Shibata, and Ken Tanaka
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201800381
Link to VoR: http://dx.doi.org/10.1002/chem.201800381
Supported by

10.1002/chem.201800381
Chemistry - A European Journal
FULL PAPER
Rhodium-Catalyzed Cascade Synthesis of
Benzofuranylmethylidene-Benzoxasiloles: Elucidating Reaction
Mechanism and Efficient Solid-State Fluorescence
Tomoya Namba,[a] Yoshihiro Hayashi,[a] Susumu Kawauchi,[a] Yu Shibata,[a] and Ken Tanaka*,[a]
Abstract: It has been developed a new synthetic route to highly
fluorescent benzofuranylmethylidene-benzoxasiloles via the cationic
rhodium(I)/binap complex-catalyzed cascade cycloisomerization of
bis(2-ethynylphenol)silanes involving the 1,2-silicon and 1,3-carbon
(alkyne) migrations followed by oxycyclization. The present
synthesis requires only three steps, starting from commercially
available dichlorodiisopropylsilane, which is markedly shorter than
our previous synthesis (eight steps starting from commercially
available chlorodiisopropylsilane). Theoretical calculation elucidated
the mechanism of the above cascade cycloisomerization. This
reaction is initiated by the formation of a rhodium vinylidene via not
the direct 1,2-silicon migration but the unprecedented stepwise 1,5-
silicon migration followed by C-Si bond-forming cyclization from a
dearomatized allenylrhodium complex. Subsequent 1,3-carbon
(alkyne) migration leading to a η3-allenyl/propargyl-rhodium complex
followed by oxycyclization through π-bond (alkyne) activation with
the σ*–π* conjugation of the σ* orbital of the Si–C bond and the
π* orbital of the π electron system lowers the LUMO energy and
increases the fluorescence quantum yield as a result of increase
in the oscillator strength at the π–π* transition.[3] As another
important effect of the silicon introduction is the improvement of
solid-state fluorescence. In general, fluorescent molecules
exhibit intense emission in solution state, but the fluorescence
quantum yield decreases sharply in solid state due to self-
absorption, intermolecular energy transfer, formation of non-
luminescent aggregates, and so on.[4] Introducing bulky
substituents on the silicon atom of the fluorescent molecule
suppresses intermolecular interactions in solid state, which
improves the fluorescence quantum yield.[5]
Previous work
iPr
iPr
4 steps
Si
RhI+
iPr
iPr
R2
Si
R1
Cl
H
1,2-Si-migration
&
1,3-C-migration
the
cationic
rhodium(I)
complex
affords
the
OH
Me
benzofuranylmethylidene-benzoxasilole product. The structure-
fluorescence property relationships of the thus obtained
benzofuranylmethylidene-benzoxasiloles were investigated, which
revealed that good fluorescence quantum yields were observed in
solution state (φF = 69–87%) by introduction of electron-donating
alkyl and phenyl groups on two phenoxy groups. In powder state, 4-
methyl- and 4-methoxy-phenoxy derivatives exhibited efficient blue
fluorescence (φF = 52% and 46%, respectively). Especially, the 4-
methylphenoxy derivative was thermally stable, and exhibited strong
narrow-band fluorescence in film state (blue, φF = 95%) and the red-
shifted strong narrow-band fluorescence (green, φF = 90%) in
crystalline state as a result of the formation of an offset π−stacked
dimer, which was confirmed by an X-ray crystallographic analysis
and theoretical calculation.
R1
3 steps
Me
Si
R2
O
Si
R1 = Me
O
iPr
O
iPr
R2 = SiMe3
iPr
iPr
(2c) (8 steps)
highly fluorescent
This work
iPr
iPr
Si
2 steps
iPr
iPr
Si
R
R
R
Cl
Cl
OH
HO
1
RhI+
1,2-Si-migration
O
Si
&
O
iPr
R
1,3-C-migration
&
Oxycyclization
iPr
2 (3 steps)
strong narrow-band
fluorescence in film and crystalline-states
(R = p-Me)
Introduction
Scheme
1.
Rhodium-catalyzed
synthesis
of
highly
fluorescent
From the viewpoint of the development of organic functional
materials not using rare elements, the synthesis of π-conjugated
molecules containing naturally abundant silicon has been
extensively studied for decades.[1] These studies revealed that
introduction of silicon into the π electron system develops
characteristic electronic and optical properties.[2] For example,
benzofuranylmethylidene-benzoxasiloles 2.
On the other hand, our research group recently reported the
novel method for the synthesis of highly fluorescent
organosilicon compounds. In 2016, we reported the rhodium(I)-
catalyzed cycloisomerization of 2-silylethynyl phenols, leading to
3-silylbenzofurans, via 1,2-silicon migration through the alkyne-
vinylidene isomerization.[6] As previously reported reactions
involving the metal vinylidene formation via the 1,2-silicon
migration require high temperature (≥90 °C),[7] the above
reaction proceeding under mild conditions (room temperature–
80 °C) is noteworthy. Subsequently, we reported the synthesis
[a]
T. Namba, Dr. Y. Hayashi, Prof. Dr. S. Kawauchi, Dr. Y. Shibata,
Prof. Dr. K. Tanaka
Department of Chemical Science and Engineering, Tokyo Institute
of Technology, O-okayama, Meguro-ku, Tokyo 152-8550, Japan
E-mail: ktanaka@apc.titech.ac.jp
of
alkynylmethylidene-benzoxasiloles[8]
from
2-
http://www.apc.titech.ac.jp/~ktanaka/
(alkynylsilylethynyl)phenols via the 1,2-silicon migration to form
rhodium vinylidenes followed by 1,3-carbon (alkyne) migration
(Scheme 1, top).[9] Although there are several examples of the
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/chem.201800381
Chemistry - A European Journal
FULL PAPER
transition-metal-catalyzed
cycloisomerization
reactions[10,11]
methanol afforded 1a in 79% yield with single isolation by a
involving silicon[12,13] or carbon migration,[14,15] the transition-
metal-catalyzed cycloisomerization reaction involving both the
silicon and carbon migrations has not been reported to date.
Pleasingly, the thus obtained trimethylsilylethynylmethylidene-
benzoxasilole[16] was transformed into benzofuranylmethylidene-
benzoxasilole 2c, which showed high fluorescence quantum
yields in both solution and powder states (Scheme 1, top).[9]
However, the synthesis of 2c requires many reaction steps (8
steps from commercially available chlorodiisopropylsilane),
although different substituents can be introduced on two
benzene rings of benzofuranylmethylidene-benzoxasiloles 2,
and the detailed mechanism of the 1,2-silicon migration that
proceeds under surprisingly mild conditions has not been
elucidated. Furthermore, although introducing bulky substituents
on the silicon atom would suppress undesired intermolecular
interactions in solid state, the detailed reason why the observed
efficient solid-state fluorescence of 2c appears is unknown.
In this article, we have succeeded the synthesis of
benzofuranylmethylidene-benzoxasiloles 2 in only three steps
from commercially available dichlorodiisopropylsilane via the
silica gel column chromatography.[20]
Table 1. Optimization of reaction conditions for synthesis of 2a from 1a.[a]
O
Si
O
iPr
iPr
10–20 mol %
[Rh(cod)2]BF4/Ligand
additive (0.1–1.5 equiv)
iPr
iPr
Si
2a
+
(CH2Cl)2, 60 °C, 17 h
OH
HO
1a
Si
O
iPr
iPr
HO
3a
Entry
Ligand
Catalyst
[mol %]
Additive
[equiv]
Yield
[%][b]
2a
3a
–
1
2
(S)-binap
(S)-binap
(S)-binap
(S)-binap
(S)-binap
(S)-binap/cod
(S)-segphos
(S)-H8-binap
biphep
20
–
45
52
60
72
74
0
20
20
20
20
20
20
20
20
20
10
10
20
THF (1.5)
MeOH (1.5)
MeOH (0.5)
MeOH (0.1)
MeOH (0.1)
MeOH (0.1)
MeOH (0.1)
MeOH (0.1)
MeOH (0.1)
MeOH (0.1)
MeOH (0.1)
MeOH (0.1)
–
3
–
4
–
cationic
rhodium(I)/binap
complex-catalyzed
cascade
5
6[c]
–
cycloisomerization of bis(2-ethynylphenol)silanes 1 (Scheme 1,
bottom). The detailed mechanism of the above catalysis was
fully clarified by both experimental and theoretical studies.
Moreover, structure-fluorescence property relationship of 2 was
examined and elucidated by X-ray crystallographic analyses and
theoretical calculations.
0
7
74
52
72
7
–
8
13
–
9
10[d]
11
12
13[e]
dppe
36
–
(S)-binap
(S)-segphos
(S)-binap
73
19
9
40
63
Results and Discussion
[a] [Rh(cod)2]BF4 (0.010–0.020 mmol), ligand (0.010–0.020 mmol), 1a (0.10
mmol), additive (0.010–0.15 mmol), and (CH2Cl)2 (1.0 mL) were used. Active
catalysts are prepared through hydrogenation (1 atm, RT). [b] Isolated yield.
[c] Catalysts were used without hydrogenation. [d] [Rh(nbd)2]BF4 was used
instead of [Rh(cod)2]BF4. [e] At RT.
Our research group previously reported the catalytic cascade
isomerization reactions of allyl propargyl ethers to allenic
aldehydes and dienals, in which the olefin isomerization, the
propargyl Claisen rearrangement, and the carbonyl migration
reactions are catalyzed through both the σ- and π-bond
activations with
a
single cationic rhodium(I) complex.[17]
O
Therefore, we envisioned that the cationic rhodium(I) complex
would catalyze oxycyclization[18,19] through π-bond activation as
well as the 1,2-silicon and 1,3-carbon (alkyne) migrations of the
bis(2-ethynylphenol)silane 1, which would be readily prepared
from commercially available dichlorodiisopropylsilane, to
benzofuranylmethylidene-benzoxasilole 2 (Scheme 1, bottom).
O
O
PPh2
PPh2
PPh2
PPh2
PPh2
PPh2
O
(S)-binap
(S)-segphos
(S)-H8-binap
PPh2
PPh2
PPh2
PPh2
dppe
OTHP
biphep
iPr
iPr
(2.1 equiv)
Si
1) nBuLi (2.1 equiv), Et2O
Figure 1. Structures of bisphosphine ligands.
0 ˚C to RT
iPr
iPr
Si
The reaction of the thus obtained bis(2-ethynylphenol)silane
1a to benzofuranylmethylidene-benzoxasilole 2a was examined
as shown in Table 1. In this study, active catalysts were
prepared through removal of the cod ligand by hydrogenation as
reported in our previous report.[9] We were pleased to find that
the desired cascade cycloiosmerization proceeds to give 2a by
using a cationic rhodium(I)/(S)-binap complex (20 mol %) at
60 °C, while the yield of 2a was moderate due to the formation
of various unidentified by-products as well as incomplete
conversion of 1a (entry 1). In our previous report, coordinating
additives suppressed the formation of by-products as well as
increasing the substrate conversion in the cases of highly
coordinative substrates. As the present bisphenol substrate 1a is
Cl
Cl
2) 20 mol % PPTS, MeOH, RT
OH
HO
1a / 79% (2 steps, single isolation)
Scheme 2. Synthesis of bis(2-ethynylphenol)silane 1a.
Thus, we first examined the synthesis of bis(2-
ethynylphenol)silane 1a. After examining various synthetic
routes, we found that this compound can be readily synthesized
starting from commercially available dichlorodiisopropylsilane as
shown in Scheme 2. The reaction of lithiated THP-protected 2-
ethynylphenol and dichlorodiisopropylsilane followed by
deprotection with pyridinium p-toluenesulfonate (PPTS) and
This article is protected by copyright. All rights reserved.

10.1002/chem.201800381
Chemistry - A European Journal
FULL PAPER
highly coordinative, we tested the coordinating additives. We
found that the addition of THF and MeOH (1.5 equiv) increases
the yield of 2a (entries 2 and 3), and MeOH is the best additive
(entry 3). Reduction of the amount of MeOH was also examined
(entries 3–5), which revealed that the use of 0.1 equiv of MeOH
affords 2a in the highest yield (entry 5). When the catalyst was
used without hydrogenation, no reaction was observed (entry 6).
Screening of bisphosphine ligands (entries 7–10) revealed that
the use of (S)-binap and (S)-segphos affords 2a in the highest
yield (entries 6 and 7). Reduction of catalyst loadings to 10
mol % revealed that the use of (S)-binap affords 2a in
acceptable yield (entry 11), in contrast, the use of (S)-segphos
afforded not the desired product 2a but phenol 3a as a major
product (entry12). Thus, we selected (S)-binap as the best
ligand. At room temperature, 1a was consumed thoroughly, but
the desired product 2a was obtained as a minor product and the
phenol 3a was obtained as a major product (entry 13).
ethynyl-2-naphthol derivative 1i proceeded to give the desired
cycloisomerization product 2i in low yield by employing a high
catalyst loading (20 mol %) and elevated temperature (80 °C).
To gain mechanistic insight into the reaction, the time course
of the reaction of 1a was examined as shown in Table 2. To our
surprise, 1a was consumed thoroughly after only 1 hour to
generate alkynylphenol 3a as a major product and the desired
product 2a as a minor product (entry 1). After 3 hours, 3a
decreased and 2a became the major product (entry 2). Finally
after 17 hours, 3a disappeared and 2a became the sole product
(entry 3). This result suggests the intermediacy of alkynylphenol
3a in the formation of benzofuran 2a.
Table 2. Time-course of the reaction of 1a.[a]
O
Si
O
iPr
10 mol %
[Rh(cod)2]BF4/(S)-binap
MeOH (0.1 equiv)
iPr
iPr
iPr
R2
Si
2a
+
R1
10 mol %
[Rh(cod)2]BF4/(S)-binap
MeOH (0.1 equiv)
iPr
iPr
R1
Si
R1
R3
R1
(CH2Cl)2, 60 °C
R2
R3
R2
R3
OH
HO
R2
R3
R4
(CH2Cl)2, 60 °C
17 h
O
Si
1a
O
Si
O
iPr
iPr
OH
HO
iPr
iPr
R4
R4
R4
HO
1
2
3a
MeO
Me
Time (h)
2a / Yield [%][b]
3a / Yield [%][b]
1
3
13
52
74
65
24
0
Si
O
Si
O
Si
O
O
O
OMe
O
Me
iPr
iPr
iPr
iPr
iPr
iPr
2a / 73%
2b / 81%
2c / 75%
17
t-Bu
Ph
F3C
[a] [Rh(cod)2]BF4 (0.010 mmol), (S)-binap (0.010 mmol), 1a (0.10 mmol),
MeOH (0.010 mmol), and (CH2Cl)2 (1.0 mL) were used. Active catalysts were
prepared through hydrogenation (1 atm, RT). [b] Isolated yield.
Si
O
Si
O
Si
O
O
t-Bu
O
Ph
O
CF3
iPr
iPr
iPr
iPr
iPr
iPr
2d / 42%[a]
2e / 78%
Me
2f / 60%
On the other hand, in the reaction of 1e, a small amount of
chromenylidene-benzoxasilole 4e[21] was isolated as a by-
product (Scheme 4). The same by-products 4 were generated
from all bisphenol substrates 1, although these by-products
Ac
Si
O
O
iPr
iPr
Si
O
iPr
O
Ac
iPr
Me
could not be isolated in pure forms.
2g / 43%
2h / 72%
Ph
Si
O
O
Ph
iPr
iPr
10 mol %
[Rh(cod)2]BF4/(S)-binap
MeOH (0.1 equiv)
iPr
iPr
Me
Si
O
O
iPr
iPr
Si
Si
O
O
2e / 78%
+
iPr
Ph
Ph
Me
iPr
(CH2Cl)2, 60 °C, 17 h
Ph
2i / 26%[a,b]
2j / 0%
OH
HO
Ph
1e
O
Scheme 3. Synthesis of benzofuranylmethylidene-benzoxasiloles 2.
[Rh(cod)2]BF4 (0.010 mmol), (S)-binap (0.010 mmol), 1 (0.10 mmol), MeOH
(0.010 mmol), and (CH2Cl)2 (1.0 mL) were used. Active catalysts were
prepared through hydrogenation (1 atm, RT). [a] [Rh(cod)2]BF4 (0.020 mmol)
and (S)-binap (0.020 mmol) were used. [b] At 80 °C for 16 h.
Si
O
iPr
iPr
4e / 6%
Scheme 4. Isolated by-product 4e in the reaction of 1e.
With the optimized reaction conditions in hand, we explored
the scope of this cycloisomerization as shown in Scheme 3.
Electronically and sterically diverse functional groups could be
incorporated on the phenol moiety. The yields of the desired
cycloisomerization products 2a–c and 2h, obtained from
electron-rich substrates 1a–c and 1h, were higher than those of
2f and 2g, obtained from electron-deficient substrates 1f and 1g.
Unfortunately, the reaction of sterically demanding tert-butyl
substituted substrate 1d required a high catalyst loading (20
mol %) to reach complete conversion. The reactions of sterically
demanding 2-naphthol derivatives 1i and 1j were also examined.
Although 1-ethynyl-2-naphthol derivative 1j was not converted to
the desired cycloisomerization product 2j, the reaction of 3-
As shown in Scheme 5, the reaction of 1a was conducted in
the presence of CH3OD. External deuteriums were incorporated
in the product 2a at the vinyl position and the C-3 position of the
benzofuran moiety.
Db
10 mol %
[Rh(cod)2]BF4/(S)-binap
iPr
iPr
Da
Si
CH3OD (1.5 equiv)
O
Si
O
(CH2Cl)2, 60 °C, 17 h
iPr
OH
HO
iPr
1a
2a / 64%
Da : (15% D)
Db : (16% D)
Scheme 5. Reaction of 1a in the presence of CH3OD.
This article is protected by copyright. All rights reserved.

10.1002/chem.201800381
Chemistry - A European Journal
FULL PAPER
via η3-butadienyl form H,[24] although energy levels of G and H
are very close (0.4 kcal/mol, see Supporting Information).
Furthermore, the direct cyclization pathway from B without
protonation via the attack of the hydroxy group to the central
carbon furnishing η3-butadienyl intermediate H would also be
possible.[25] However, this pathway may be excluded because
protonated intermediate 3a was observed in the time-course
study (Table 2). On the other hand, E is the only intermediate to
be capable of affording 4 via vinylrhodium intermediate I, which
is slightly lower energy than G and H (these relative energies
respect to I are 1.3 and 0.9 kcal/mol, respectively).
R
R
[Rh]+
HO
Si
[Rh]+
–H+
•
O
Si
R
R
OH
HO
1
A
[Rh]
Si
O
R
HO
R
B
H+
The detailed explanation of conversion of triene E to not
benzofuran 2 but chromene 4 is given by the calculated Gibbs
free energy diagram as shown in Figure 2. Complex J, derived
from E by deprotonation, would undergo oxycyclization to give H
and I via transition states JH_TS and JI_TS, respectively.
Significantly lower energy (ca. 11 kcal/mol) of transition state
JH_TS than transition state JI_TS strongly supports the
conversion of triene E to chromene 4, although energy levels of
cyclized complexes H and I are very close.
H
[Rh]+
•
H
R
OH
[Rh]+
HO
•
•
[Rh]+
•
H
Si
R
O
HO
[Rh]+
Si
O
R
HO
Si
R
O
R
Si
O
R
R
R
D
C
E
F
0 kcol/mol
(R = Me)
8.7 kcol/mol
(R = Me)
0.5 kcol/mol
(R = Me)
11.1 kcol/mol
(R = Me)
–H+
–H+
[Rh]
H
R
[Rh]
H
H
[Rh]+
O
•
[Rh]
•
O
Si
R
O
H
O
Si
R
O
H
R
Si [Rh]
HO
Si
O
R
O
Me
G
H
I
R
O
Si
Me
Me
Me
O
–[Rh]+
H+
E
JH_TS
H
16.8
H
R
O
R
H+
H+
–[Rh]+
H
O
"
–[Rh]+
Si
R
O
Si
R
H
–H+
5.9
O
4
2
0.0
1
-
Scheme 6. Possible reaction mechanism for formation of 2 and 4 from 1.
Gibbs free energies of intermediates C–F were obtained by the ONIOM
calculations. [Rh] = Rh-biphep.
l
o
[Rh]
H
O
m
l
a
c
[Rh]
H
k
/
H
O
Si
Si
O
O
Me
Me [Rh]
G
Me
JI_TS
Based on the above observations, a possible mechanism for
the rhodium-catalyzed cascade cycloisomerization of bis(2-
Me
!
O
Si
O
Me
J
Me
H
ethynylphenol)silanes
1
into
benzofuranylmethylidene-
–25.3
–26.2
benzoxasiloles via the concomitant silicon and carbon
2
migration, and the oxycyclization is shown in Scheme 6. In our
previous report of the rhodium-catalyzed 1,2-silicon and 1,3-
carbon migration, the formation of η3-allenyl/propargyl-
O
rhodium[22]
is
proposed.
In
the
present
cascade
H
Si
[Rh]
O
Me
Me
cycloisomerization, the reaction of diyne 1 with the cationic
rhodium(I) complex would also generate η3-allenyl/propargyl-
rhodium species B via rhodium vinylidene A. Protonation of B
gives four possible intermediates C–F, which undergo
oxycyclization to give 2 and 4 via intermediates G–I. In order to
determine the most plausible pathways, theoretical investigation
of the oxycyclization was conducted by the two-layer (QM:QM)
ONIOM calculations[23] using a simplified substrate (R = Me) and
I
Figure 2. Gibbs free energy diagram (in kcal/mol) of the rhodium-catalyzed
oxycyclization pathway obtained by the ONIOM calculations. [Rh] = Rh-biphep.
We also conducted the theoretical study of the detailed
pathway of the 1,2-silicon migration. The calculated Gibbs free
energy diagram is shown in Figure 3. Once deprotonation of
rhodium(I)-bisphenol
complex
K
affords
dearomatized
rhodium catalyst (ligand
= biphep). The ONIOM method
allenylrhodium complex L.[26] Nucleophilic attack of the carbonyl
oxygen to the silicon atom affords rhodium acetylide M, which is
lower energy than L (ca. 18 kcal/mol), via not expected 1,2-
silicon migration but unprecedented 1,5-silicon migration[27]
through transition state LM_TS. Then, interaction between the
basic β-carbon of acetylide and the silicon atom affords rhodium
vinylidene A[28,29] through transition state MA_TS. Rhodium
vinylidene A is transformed into η3-allenyl/propargyl-rhodium
complex B via 1,3-carbon (alkyne) migration through transition
state AB_TS. On the other hand, 1,2-silicon migration without
deprotonation would afford rhodium vinylidene N via transition
state KN_TS. Subsequent deprotonation would also afford the
subdivides the molecular model into two layers, with each being
described at high and low levels of theory. In this study, the
density functional theory using the ωB97X-D functional with the
SDD+f basis sets for the Rh atom and with the 6-31+G(d,p)
basis sets for the other atoms was selected for the high level of
theory. The semi-empirical PM6 was selected for the low level of
theory. Gibbs free energies of intermediates are compared with
intermediate C (0 kcal/mol). The calculations revealed that
enynes C and D are lower energy than trienes E and F (ca. 8–11
kcal/mol), so that the formation of 2 from C and D via
vinylrhodium intermediate G is more plausible than that from E
This article is protected by copyright. All rights reserved.

10.1002/chem.201800381
Chemistry - A European Journal
FULL PAPER
above mentioned rhodium vinylidene A. However, activation
energy of the step from L to M via LM_TS (6.5 kcal/mol) is ca.
10 kcal/mol lower than that of the step from K to N via KN_TS
(16.9 kcal/mol), which suggests that the former pathway is more
likely to occur. We previously advocated that rhodium vinylidene
A might be generated from 1 via the 1,2-silicon migration. The
present calculation, however, suggested that A would be
generated from 1 via the unprecedented stepwise 1,5-silicon
migration followed by C-Si bond-forming cyclization. This
stepwise mechanism lowers energy barrier of the silicon
migration to form the rhodium vinylidene and enables the
present low-temperature reaction.
[Rh]+
OH
Me
Si
•
[Rh]+
HO
Me
HO
KN_TS
16.9
Si
HO
Me
Me
[Rh]+
N
14.4
OH
Si
OH
Me Me
[Rh]+
HO
[Rh]
K
•
1
-
•
l
0.0
o
[Rh]
Si–
Me
Si
O
m
HO
l
O
Me
Me
a
Me
"
Si
c
O
k
LM_TS
AB_TS
Me
Me
/
HO
–H+
"
6.5
1.2
–H+
G
0.0
[Rh]
Si
!
MA_TS
–0.5
–1.9
•
•
[Rh]+
HO
[Rh]
OH
O
Me Me
Si–
Me Me
–17.4
[Rh]
O
L
Si
O
Me
HO
Me
A
B
–43.7
O
Me
Si
Me
HO
M
Figure 3. Gibbs free energy diagram (in kcal/mol) of the rhodium-catalyzed oxycyclization pathway the reaction pathways in the rhodium-catalyzed
cycloisomerization of bis(2-ethynylphenol)silanes 1, leading to benzoxasiloles 2, obtained by the ONIOM calculations. [Rh] = Rh-biphep.
Photophysical
properties
of
the
present
cascade
in solution state. Interestingly, fluorescence quantum yields in
solid state highly depended on the substituents on two phenoxy
groups. Products 2e–i showed significantly lower fluorescence
quantum yields in powder state than those in solution state. In
particular, it was markedly decreased in 4-phenylphenoxy
product 2e, which showed excellent quantum yield in solution
state. On the other hand, none-substituted product 2a and 4-tert-
butylphenoxy product 2d showed moderate fluorescence
quantum yields, and 4-methoxylphenoxy and 4-methylphenoxy
products 2b and 2c showed good fluorescence quantum yields.
Normalized fluorescence spectra of the representative products
(2c–e) in solution and powder states are shown in Figures 4a–c.
cycloisomerization products 2a–i are shown in Table 3 and
Figures S2–S10.[20] Substitution with electron-donating groups
(2b–d and h) and additional benzene rings (2e and 2i) on two
phenoxy groups showed red shifts of absorption and emission
maxima as compared to a non-substituted product 2a. In
contrast, substitution with electron-withdrawing groups (2f and
2g) showed little change. Pleasingly, all products showed
moderate to high fluorescence quantum yields in solution state.
Particularly, products with alkyl or phenyl groups (2c–e and 2h)
on two phenoxy groups showed high fluorescence quantum
yields. Among them, the quantum yield of tert-butyl substituted
one 2d was the highest.
Table 4. Photophysical properties of benzofuranylmethylidene-benzoxasiloles 2
in powder, crystal, and film states.[a]
Table 3. Photophysical properties of benzofuranylmethylidene-benzoxasiloles 2
in solution state.[a]
fluorescence
φF
(nm[c])
compound
state
λmax / nm[b]
[b]
absorption
fluorescence
φF
compound
(nm[c])
λmax / nm[b]
λmax / nm[b]
(nm[c])
(nm[c])
2a
2b
2c
2d
2e
2f
powder
powder
powder
powder
powder
powder
powder
powder
powder
crystal
film
487, 518 (300)
509, 539 (340)
497, 526 (300)
501, 520 (300)
492 (340)
0.274 (420)
0.463 (360)
0.517 (430)
0.361 (340)
0.115 (330)
0.113 (300)
0.239 (400)
0.190 (300)
0.072 (320)
0.904 (450)
0.947 (450)
2a
2b
2c
2d
2e
2f
376, 397
401
411, 430 (300)
480 (340)
426, 445 (300)
423,443 (300)
460 (340)
0.756 (360)
0.691 (340)
0.843 (400)
0.866 (330)
0.844 (340)
0.606 (340)
0.603 (300)
0.803 (360)
0.442 (370)
384, 405
383, 404
389, 409
372, 393
377, 398
381, 403
392, 413
551 (300)
2g
2h
2i
2c
2c
471 (300)
404, 424 (300)
409, 430 (300)
416, 437 (300)
488 (340)
506, 536 (300)
575 (340)
498 (350)
2g
2h
2i
440 (350)
[a] Measured at 25 °C in CHCl3. [b] At 1.25 ×10–5 M. [c] Excitation wavelength.
[a] Measured at 25 °C in CHCl3. [b] At 1.25 ×10–5 M. [c] Excitation wavelength.
Fluorescence data of 2a–i in solid state are summarized in
Table 4 and Figures S2–S10.[20] Fluorescence maxima for all
products in powder state were red-shifted as compared to those
As the 4-methylphenoxy product 2c showed high fluorescence
quantum yields in both solution and powder states, its
This article is protected by copyright. All rights reserved.

10.1002/chem.201800381
Chemistry - A European Journal
FULL PAPER
fluorescence properties were further examined in a cast film of
poly(methyl methacrylate) (PMMA) and a single crystal. The
powder state (Figure 4e) and blue-green emission in crystalline
state (Figure 4g)]. Furthermore, fluorescence quantum yields in
both film and crystalline states were higher than those in solution
and powder states (Table 4; 0.947 in film state and 0.904 in
crystalline state vs. 0.843 in solution state and 0.517 in powder
state).
Electronic properties of the representative products 2a and
2c–e were examined by DFT calculations using Gaussian 09 at
the B3LYP/6-31G(d) level.[20] Frontier molecular orbitals (MOs)
are depicted in Figure S11. In 2a, 2c, and 2d, both HOMO and
LUMO orbitals are delocalized and spread over whole molecules
except substituents (Figures S11a–e). In 2e, HOMO orbital
spreads to the additional benzene ring but LUMO orbital does
not (Figure S11d). The energy diagrams of 2a and 2c–e reveals
that 2d shows the highest HOMO and LUMO levels and 2c
shows the smallest HOMO-LUMO energy gap (Figure S11f). We
also examined carbon analogue of 2c (2k, Figure S11e), in
which the silicon atom is replaced to the carbon atom, to see the
effect of silicon on the electronic properties. Importantly, no
significant difference was observed between 2c and 2k in terms
of the HOMO and LUMO orbitals as well as the HOMO-LUMO
energy gaps (Figures S11b, S11e, and S11f). Therefore, the
observed efficient fluorescence properties, especially in solid-
state, may not arise from the electronic nature of silicon.[30]
Figure 4. Normalized fluorescence spectra in solution (blue line), in film (light
blue line), in powder (red line), and in crystal (green line) of 2c (a), and
normalized fluorescence spectra in solution (blue line), and in powder (red
line) of 2d (b) and 2e (c). Photographs showing the fluorescence in the CHCl3
solution (d), powders (e), film (PMMA) (f), and crystal (g) of 2c with irradiation
at 365 nm.
Figure 5. X-ray crystal structure of 2a. (a) Packing structure and (b)
interactions of two adjacent molecules through π–π stacking and CH–π
interaction.
Single-crystal X-ray structure determination was performed for
2a and 2c–e and their packing structures are shown in Figures
5–8.[20] These compounds 2a and 2c–e have nearly planar
conformations with dihedral angles (between two double bonds)
of 3.50–11.4° (Figures S15, S16b, S18, and S21). In the
unsubstituted compound 2a, there are two columns (green and
blue, Figure 5a), in which there are two distinct intermolecular
interactions. In the green column, an offset π-stacked dimmer
was observed, in which the C=C double bond of the furan moiety
is located above that of the adjacent molecule with an
interatomic distance of 3.30 Å and the diisopropylsilyl groups
face the opposite direction to avoid steric repulsion (Figure 5b).
The green and blue columns interact with each other through
PMMA film showed a narrow-band single emission maximum
peak and the observed fluorescence color was similar to that in
solution state [Table 4 and Figure 4a; blue emission in solution
(Figure 4d) and in cast film (Figure 4f)]. Fortunately, 2c was a
highly crystalline compound and very large prismatic crystals (ca.
7 mm) could be readily prepared by recrystallization from n-
hexane. Although 2b also showed high fluorescence quantum
yields in both solution and powder states, this compound
afforded small needle crystals. The thus obtained single crystal
of 2c showed a narrow-band single emission maximum peak
and the observed fluorescence color was slightly different from
that in powder state [Table 4 and Figure 4a; green emission in
This article is protected by copyright. All rights reserved.

10.1002/chem.201800381
Chemistry - A European Journal
FULL PAPER
CH–π interactions between the hydrogen atom at the para
position of the benzoxasilole moiety with respect to oxygen and
the benzofuran moiety with an interatomic distance of 2.89 Å
(Figure 5b). As a result of these intermolecular interactions, the
molecules pack tightly, which may decrease the solid-state
fluorescence quantum yield.
In the same way as 2a, the p-methyl-substituted compound 2c
also forms an offset π-stacked dimer with the nearest interatomic
distance A of 3.43 Å and the diisopropylsilyl groups face the
opposite direction (Figure 6a). Importantly, methyl substitution of
the para hydrogen atom of the benzoxasilole moiety blocks the
CH–π interactions between the columns and results in the
dimers to stack with the nearest interatomic distance B of 3.50 Å,
which is longer than the distance A (Figure 6b). Additionally,
interactions between two methyl groups (2.16 Å) are observed
as shown in Figure 6b.
On the other hand, no overlap of the π conjugated moiety
between the two adjacent molecules is observed in the p-tert-
butyl-substituted compound 2d presumably due to steric
hindrance of the tert-butyl group, but instead, CH–π interactions
between the diisopropylsilyl group and the benzofuran moiety
with an interatomic distance of 2.85 Å are observed in the side
view (Figure 7a). Additionally, in the top view, C-H...O
interactions between the diisopropylsilyl group and the
benzofuran moiety with an interatomic distance of 2.52 Å are
observed (Figure 7b). Thus, molecules are spatially separated
from neighbors, which may prevent concentration quenching of
fluorescence in solid state.
On the contrary, in the p-phenyl-substituted compound 2e, π-π
stacking interactions between the benzofuran and benzoxasilole
moieties with an interatomic distance of 3.30 Å, and the
benzoxasilole moiety and the phenyl substituent of the
benzofuran moiety with a longer interatomic distance of 3.37 Å
are observed in the side view (Figure 8a). Additionally, in the top
view, C-H...O interactions between the phenyl substituent of the
benzofuran moiety and the benzoxasilole moiety with an
interatomic distance of 2.45 Å are observed (Figure 8b).
Figure 6. X-ray crystal structure of 2c. (a) Interaction of two adjacent
molecules through π–π stacking and (b) packing structure.
Figure 8. X-ray crystal structure of 2e. (a) Side view and (b) top view of
packing structure.
As seen in Figures 6 and 8, two types of dimers can be
extracted from the packing structures of 2c and 2e. These
dimers are named as 2cA (interplanar distance A = 3.43 Å) and
2cB (interplanar distance B = 3.50 Å) extracted from 2c, and
2eA (interplanar distance A = 3.30 Å) and 2eB (interplanar
distance B = 3.37 Å) extracted from 2e. Electronic structures of
the above dimers were examined by DFT calculation using
Gaussian 09 at the B3LYP/6-31G(d) level. Frontier molecular
orbitals (MOs) and energy diagrams of 2cA, 2cB, 2eA, and 2cB
are depicted in Figure 9. The interaction of LUMO orbitals in 2cA
is markedly larger than that in 2cB. In contrast, large overlaps of
LUMO orbitals are observed in both 2eA and 2eB.
Figure 7. X-ray crystal structure of 2d. (a) Side view and (b) top view of
packing structure.
This article is protected by copyright. All rights reserved.

10.1002/chem.201800381
Chemistry - A European Journal
FULL PAPER
Figure 9. Pictorial representation of frontier molecular orbitals of dimer 2cA (a),
dimer 2cB (b), dimer 2eA (c), and dimer 2eB (d) calculated at the B3LYP/6-
31G(d) level of theory.
From the above X-ray crystallographic analyses and
theoretical calculations, we concluded that 2c shows monomer
emission in solution and film states, but dimer emission in
crystalline state arising from the offset π-stacking.[31] In solution
and film states, molecules are distant from each other; therefore,
those emissions are attributed to monomers. High quantum
yields were observed in these states because molecules are
separated from each other and concentration quenching is
prevented.[5] In powder and crystal states, dimers are formed in
the short interplanar distance A of 3.43 Å with the interaction of
LUMO orbitals and these dimers are separated in the long
interplanar distance B of 3.50 Å. This intermolecular LUMO
orbital interaction of the adjacent molecules resulted in a
decrease in the HOMO–LUMO energy gap and an increase in
the oscillator strength for the lowest-energy transition as
compared to the isolated monomer.[32] These observations
indicate strong exciton coupling in 2c, which probably caused
the red shift and the quantum yield increase in the crystal-state
fluorescence. On the other hand, 2e densely stacks in the
interplanar distances A and B (3.30 Å and 3.37 Å, respectively)
with the interactions of LUMO orbitals, which resulted in
concentration quenching of fluorescence in solid state.[4]
Finally, thermal stability of the most emissive product 2c was
examined by thermogravimetric and differential thermal analysis
(TG-DTA) under a helium atmosphere.[33] The TG-DTA curves,
shown in Figure 10, revealed that 2c is thermally stable with
◦
keeping constant TG (less than 5% weight loss) up to 256 C.
Figure 10. TG-DTA curves of 2c.
Conclusions
In conclusion, we have developed a new synthetic route to
highly fluorescent benzofuranylmethylidene-benzoxasiloles via
the cationic rhodium(I)/(S)-binap complex-catalyzed cascade
cycloisomerization of bis(2-ethynylphenol)silanes involving the
1,2-silicon and 1,3-carbon (alkyne) migrations followed by
oxycyclization. The present synthesis requires only three steps,
starting form commercially available dichlorodiisopropylsilane,
which is markedly shorter than our previous synthesis (eight
steps
starting
from
commercially
available
chlorodiisopropylsilane). Theoretical calculation elucidated the
mechanism of the above cascade cycloisomerization. This
reaction is initiated by the formation of a rhodium vinylidene via
not the direct 1,2-silicon migration but the unprecedented
stepwise 1,5-silicon migration followed by C-Si bond-forming
cyclization from
a
dearomatized allenylrhodium complex.
Subsequent 1,3-carbon (alkyne) migration leading to a η3-
allenyl/propargyl-rhodium complex followed by oxycyclization
through π-bond (alkyne) activation with the cationic rhodium(I)
complex affords the benzofuranylmethylidene-benzoxasilole
product. The structure-fluorescence property relationships of the
thus obtained benzofuranylmethylidene-benzoxasiloles were
investigated, which revealed that good fluorescence quantum
yields were observed in solution state (φF = 69–87%) by
introduction of electron-donating alkyl and phenyl groups on two
phenoxy groups. In powder state, 4-methylphenoxy and 4-
methoxyphenoxy
derivatives
exhibited
efficient
blue
fluorescence (φF = 52% and 46%, respectively). Especially, the
4-methylphenoxy derivative was thermally stable, and exhibited
strong narrow-band fluorescence in film state (blue, φF = 95%)
and the red-shifted strong narrow-band fluorescence (green, φF
= 90%) in crystalline state as a result of the formation of an
offset π−stacked dimer, which was confirmed by an X-ray
crystallographic analysis and theoretical calculation. Because of
This article is protected by copyright. All rights reserved.

10.1002/chem.201800381
Chemistry - A European Journal
FULL PAPER
Matsumoto, H. Horiuchi, H. Hiratsuka, Chem. Lett. 2006, 35, 64; f) H.
Horiuchi, T. Tanaka, K. Yoshimura, K. Sato, S. Kyushin, H. Matsumoto,
H. Hiratsuka, Chem. Lett. 2006, 35, 662.
highly efficient fluorescence properties, high thermal stability,
and facile and rapid synthesis, this compound would be an
attractive candidate for OLED materials.[34]
[4]
[5]
For a review on solid state fluorescence, see: a) Y. Hong, J. W. Lam, B.
Z. Tang, Chem. Soc. Rev. 2011, 40, 5361; b) M. Shimizu, T. Hiyama,
Chem. Asian J. 2010, 5, 1516; c) J. Mei, Y. Hong, J. W. Lam, A. Qin, Y.
Tang, B. Z. Tang, Adv. Mater. 2014, 26, 5429.
Acknowledgements
a) K. Mouri, A. Wakamiya, H. Yamada, T. Kajiwara, S. Yamaguchi, Org.
Lett. 2007, 9, 93; b) H. Yamada, C. H. Xu, A. Fukazawa, A. Wakamiya,
S. Yamaguchi, Macromol. Chem. Phys. 2009, 210, 904; c) A. Iida, S.
Yamaguchi, Chem. Commun. 2009, 3002.
This work was supported partly by ACT-C (No.
JPMJCR1122YR) from Japan Science and Technology Agency
(JST), Japan and a Grant-in-Aid for Scientific Research on
Innovative Areas "π-System Figuration: Control of Electron and
[6]
[7]
H. Kanno, K. Nakamura, K. Noguchi, Y. Shibata, K. Tanaka, Org. Lett.
2016, 18, 1654.
Structural Dynamism for Innovative
Functions" (No.
a) H. Sakurai, Y. Nakadaira, A. Hosomi, Y. Eriyama, K. Hirama, C.
Kabuto, J. Am. Chem. Soc. 1984, 106, 8315; b) J. W. Dankwardt,
Tetrahedron Lett. 2001, 42, 5809; c) H. Kim, C. Lee, J. Am. Chem. Soc.
2005, 127, 10180.
JP26102004) from Japan Society for the Promotion of Science
(JSPS), Japan. T.N. thanks JSPS research fellowship for young
scientists (No. 17J10223). We are grateful to Prof. Takashi
Ishizone and Dr. Raita Goseki (Tokyo Institute of Technology)
for their assistance in the preparation of the cast film of 2c, Prof.
Ichiro Yamanaka and Mr. Yuki Semba (Tokyo Institute of
Technology) for the TG-DTA analysis, and Mr. Hiroshi Kanno
and Mr. Hideaki Oike (Tokyo University of Agriculture and
Technology) for their preliminary experiments. We also thank
Takasago International Corporation for the gift of segphos and
H8-binap, and Umicore for generous support in supplying the
rhodium and palladium complexes. The computational work was
supported partly by a Grant-in-Aid for Young Scientists (B)
(JSPS KAKENHI Grant Number 17K17720), a Grant-in-Aid for
Specially promoted Research (JSPS KAKENHI Grant Number
17H06092), and CREST (Grant Number JPMJCR1522) from
Japan Science and Technology Agency (JST), Japan. We are
grateful to Mr. Hikaru Deguchi and Mr. Ryo Nishikubo (Tokyo
Institute of Technology) for their assistance for the
computational work. The numerical calculations were carried out
on the TSUBAME3.0 supercomputer in the Tokyo Institute of
Technology, Tokyo, Japan and on the supercomputer in the
Research Center for Computational Science, Okazaki, Japan.
[8]
For examples of the synthesis of substituted methylidenebenzoxasiloles,
see: a) S. C. Shim, S. K. Park, Tetrahedron Lett. 1998, 39, 6891; b) S.
K. Park, J. Photochem. Photobiol. 2001, 144, 167; c) T. Matsuda, Y.
Ichioka, Org. Biomol. Chem. 2012, 10, 3175.
[9]
T. Namba, S. Kawauchi, Y. Shibata, H. Kanno, K. Tanaka, Angew.
Chem. Int. Ed. 2017, 56, 3004; Angew. Chem. 2017. 129. 3050.
[10] For selected recent reviews of the transition-metal-catalyzed
cycloisomerization, see: a) G. Fang, X. Bi, Chem. Soc. Rev. 2015, 44,
8124; b) R. K. Neff, D. E. Frantz, Tetrahedron 2015, 71, 7; c) V.
Michelet, In Comprehensive Organic Synthesis II, (Eds.: Knochel, P., G.
A. Molander), Elsevier, 2014, Vol. 5, P 1483; d) A. Fürstner, Acc. Chem.
Res. 2014, 47, 925; e) J. Keilitz, H. A. Malik, M. Lautens, Top.
Heterocycl. Chem. 2013, 32, 187; f) Y. Yamamoto, Chem. Rev. 2012,
112, 4736; g) C. Fehr, Synlett 2012, 23, 990.
[11] For selected recent reviews of the transition-metal-catalyzed
intramolecular heterocyclization involving molecular rearrangements,
see: a) N. Jane, T. G. Driver, Org. Biomol. Chem. 2015, 13, 9720; b) M.
Egi, S. Akai, Heterocycles 2015, 31, 931; c) M. Tőrincsi, G. Hornyánsky,
P. Kolonits, L. Novák, Period. Polytech. Chem. Eng. 2015, 59, 16; d) L.
Fensterbank, M. Malacria, Acc. Chem. Res. 2014, 47, 953; e) B.
Alcaida, P. Almendros, Acc. Chem. Res. 2014, 47, 939; f) C. Obradors,
A. M. Echavarren, Acc. Chem. Res. 2014, 47, 902; g) R. K. Shiroodi, V.
Gevorgyan, Chem. Soc. Rev. 2013, 42, 4991.
[12] a) K. Mochida, M. Shimizu, T. Hiyama, J. Am. Chem. Soc. 2009, 131,
8350; b) A. S. Dudnik, Y. Xia, Y. Li, V. Gevorgyan, J. Am. Chem. Soc.
2010, 132, 7645; c) Y. Xia, A. S. Dudnik, Y. Li, V. Gevorgyan, Org. Lett.
2010, 12, 5538; d) R. K. Shiroodi, C. I. Vera, A. S. Dudnik, V.
Gevorgyan, Tetrahedron Lett. 2015, 56, 3251.
Keywords: Organosilicon Compounds • Cycloisomerization •
Solid-State Fluorescence • Migration • Rhodium
[13] For an example of the transition-metal-catalyzed intramolecular
heterocyclization involving 1,2-germanium migration, see: I. V. Seregin,
V. Gevorgyan, J. Am. Chem. Soc. 2006, 128, 12050.
[1]
For a review on π-conjugated ladder type molecules containing silicon
and typical elements, see: a) S. Yamaguchi, C. Xu, J. Synth. Org.
Chem. Jpn. 2005, 63, 1115; b) A. Fukazawa, S. Yamaguchi, Chem.
Asian J. 2009, 4, 1386. For applications to transporting materials for
OLED, see: c) K. Tamao, K. Nakamura, H. Ishii, S. Yamaguchi, M.
Shiro, J. Am. Chem. Soc. 1996, 118, 12469; d) M. Uchida, T.
Izumizawa, T. Nakano, S. Yamaguchi, K. Tamao, K. Furukawa, Chem.
Mater. 2001, 13, 2680. For poly siloles, see: e) S. Yamaguchi, R.-Z. Jin,
Y. Itami, T. Goto, K. Tamao, J. Am. Chem. Soc. 1999, 121, 10420. For
applications to aggregation-induced emission (AIE) materials, see: f) J.
Luo, Z. Xie, J. W. Y. Lam, L. Cheng, B. Z. Tang, H. Chen, C. Qiu, H. S.
Kwok, X. Zhan, Y. Liu, D. Zhu, Chem. Commun. 2001, 1740.
[14] a) Y. Hoshimoto, H. Yabuki, R. Kumar, H. Suzuki, M. Ohashi, S. Ogoshi,
J. Am. Chem. Soc. 2014, 136, 16752; b) R. Kumar, Y. Hoshimoto, H.
Yabuki, M. Ohashi, S. Ogoshi, J. Am. Chem. Soc. 2015, 137, 11838.
[15] Very recently, the ruthenium-catalyzed cycloisomerization of 2-
alkynylanilides to produce 3-substituted indoles involving 1,2-carbon
migration has been reported. See: T. Watanabe, Y. Mutoh, S. Saito, J.
Am. Chem. Soc. 2017, 139, 7749.
[16] For the synthesis of fluorescent benzofuran derivatives, see: a) S.
Akiyama, H. Akimoto, S. i. Nakatsuji, K. Nakashima, Bull. Chem. Soc.
Jpn. 1985, 58, 2192; b) K. Nakashima, H. Akimoto, K. i. Nishida, S. i.
Nakatsuji, S. Akiyama, Talanta 1985, 32, 167; c) H. Tsuji, G. M. O.
Favier, C. Mitsui, S. Lee, D. Hashizume, E. Nakamura, Chem. Lett.
2011, 40, 576; d) D. Chand, P. Mandali, Synthesis 2015, 47, 1661; e) L.
Zhang, Z. Peng, Q. Wen, X. Li, P. Lu, Y. Wang, Org. Lett. 2016, 18,
728; f) G. Galvani, K. H. Vardhan Reddy, C. Beauvineau, N. Ghermani,
F. Mahuteau-Betzer, M. Alami, S. Messaoudi, Org. Lett. 2017, 19, 910.
[17] a) K. Tanaka, E. Okazaki, Y. Shibata, J. Am. Chem. Soc. 2009, 131,
10822; b) E. Okazaki, R. Okamoto, Y. Shibata, K. Noguchi, K. Tanaka,
Angew. Chem. Int. Ed. 2012, 51, 6722; Angew. Chem. 2012, 124, 6826.
[18] For oxycyclization catalyzed by rhdoium, see: a) N. Isono, M. Lautens,
Org. Lett. 2009, 11, 1329; b) A. Boyer, N. Isono, S. Lackner, M.
Lautens, Tetrahedron 2010, 66, 6468.
[2]
[3]
a) M. Kira, T. Miyazawa, N. Mikami, H. Sakurai, Organoparallics 1991,
10, 3793; b) C. G. Pitt, H. Bock, J. Chem. Soc. Chem. Commun. 1972,
28; c) S. Yamaguchi, K. Tamao, Chem. Lett. 2005, 34, 2; d) H. Sakurai,
M. Kumada, Bull. Chem. Soc. Jpn. 1964, 37, 1894.
a) S. Kyushin, M. Ikarugi, M. Goto, H. Hiratsuka, H. Matsumoto,
Organometallics 1996, 15, 1067; b) H. Hiratsuka, M. Tanaka, H.
Horiuchi, Naris, T. Yoshinaga, M. Oba, K. Nishiyama, J. Organomet.
Chem. 2000, 611, 71; c) H. Hiratsuka, H. Horiuchi, Y. Furukawa, H.
Watanabe, A. Ishihara, T. Okutsu, S. Tobita, T. Yoshinaga, A.
Shinohara, N. Tokitoh, M. Oba, K. Nishiyama, J. Phys. Chem. A 2006,
110, 3868; d) S. Kyushin, N. Takemasa, H. Matsumoto, H. Horiuchi, H.
Hiratsuka, Chem. Lett. 2003, 32, 1048; e) S. Kyushin, Y. Ishikita, H.
This article is protected by copyright. All rights reserved.

10.1002/chem.201800381
Chemistry - A European Journal
FULL PAPER
[19] For oxycyclization catalyzed by other transition metals, see: a) P.
Compain, J. Goré, J.-M. Vatèle, Tetrahedron 1996, 52, 10405; b) F. E.
McDonald, M. M. Gleason, J. Am. Chem. Soc. 1996, 118, 6648; c) V.
Dalla, P. Pale, New J. Chem. 1999, 23, 803; d) P. Pale, J. Chuche, Eur.
J. Org. Chem. 2000, 2000, 1019; e) E. Genin, S. Antoniotti, V. Michelet,
J. P. Genet, Angew. Chem. Int. Ed. 2005, 44, 4949; Angew. Chem.
2005, 117, 5029; f) B. Alcaide, P. Almendros, J. M. Alonso, Org. Biomol.
Chem. 2011, 9, 4405.
Ao, J. Deng, H. Yan, Angew. Chem. Int. Ed. 2017, 56, 13722; Angew.
Chem. 2017, 129, 13910.
[27] For examples of [1,5]-Brook rearrangements, see: a) K. Takeda, H.
Haraguchi, Y. Okamoto, Org. Lett. 2003, 5, 3705; b) Q. Liu, Y. Chen, X.
Zhang, K. N. Houk, Y. Liang, A. B. Smith, III, J. Am. Chem. Soc. 2017,
139, 8710.
[28] For examples of the catalytic C-C bond forming cyclization reactions via
formation of transition-metal vinylidene complexes, see: a) J. M. Joo, Y.
Yuan, C. Lee, J. Am. Chem. Soc. 2006, 128, 14818; b) L. Nunes Dos
Santos Comprido, J. E. Klein, G. Knizia, J. Kastner, A. S. Hashmi,
Chem. Eur. J. 2016, 22, 2892.
[20] See the supporting information.
[21] CCDC 1819209 (4e), CCDC 1819206 (2a), CCDC 1524363 (2c),
CCDC 1819207 (2d), and CCDC 1819208 (2e) contain the
supplementary crystallographic data for this paper. These data can be
obtained free of charge from The Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
[22] For examples of the synthesis of η3-allenyl/propargyl-rhodium
complexes via rhodium vinylidene complexes, see: S. Martínez de
Salinas, J. Díez, L. R. Falvello, J. González, M. P. Gamasa, E. Lastra,
Organometallics 2016, 35, 2793.
[29] For an example of the synthesis of a disilyl substitued gold vinylidene
complex via C-Si bond forming cyclization, see: a) R. J. Harris, R. A.
Widenhoefer, Angew. Chem. Int. Ed. 2015, 54, 6867; Angew. Chem.
2015, 127, 6971; b) R. G. Carden, N. Lam, R. A. Widenhoefer, Chem.
Eur. J. 2017, 23, 17992.
[30] A recent report disclosed that comparison of photophysical properties
between six-membered silacycles that contain multiple rings and their
carbon analogues revealed that silicon does not influence absorption
and fluorescence properties but drastically increase the fluorescence
quantum yields. See: S. Yoshioka, Y. Fujii, H. Tsujino, T. Uno, H.
Fujioka, M. Arisawa, Chem. Commun. 2017, 53, 5970.
[23] M. Svensson, S. Humbel, R. D. J. Froese, T. Matsubara, S. Sieber, K.
Morokuma, J. Phys. Chem.1996, 100, 19357.
[24] a) M. Ogasawara, A. Okada, S. Watanabe, L. Fan, K. Uetake, K.
Nakajima, T. Takahashi, Organometallics 2007, 26, 5025.; b) S. A.
Benyunes, L. Brandt, A. Fries, M. Green, M. F. Mahon, T. M. T.
Papworth, J. Chem. Soc. Dalton Trans. 1993, 3785; c) S. A. Benyunes,
L. Brandt, M. Green, A. W. Parkins, Organometallics 1991, 10, 57.
[31] For an example of the compounds, which show monomer emission in
solution and film state, and dimer smission in crystalline state, see: S.
Hisamatsu, H. Masu, M. Takahashi, K. Kishikawa, S. Kohmoto, Cryst.
Growth Des. 2015, 15, 2291.
[25] a) V. V. Krivykh, E. S. Taits, P. V. Petrovskii, Y. T. Struchkov, A. I.
Yanovski, Mendeleev Commun. 1991, 1, 103; b) T.-M, Huang, J.-T.
Chen, G.-H. Lee, Y. Wang, J. Am. Chem. Soc. 1993, 115, 1170.
[32] For an example, see: A. Fukazawa, D. Kishi, Y. Tanaka, S. Seki, S.
Yamaguchi, Angew. Chem. Int. Ed. 2013, 52, 12091; Angew. Chem.
2013, 125, 12313.
[26] For examples of cyclization via vinylidene ortho-quinone methides a) M.
Furusawa, K. Arita, T. Imahori, K. Igawa, K. Tomooka, R. Irie,
Tetrahedron Lett. 2013, 54, 7107; b) S. Beppu, S. Arae, M. Furusawa,
K. Arita, H. Fujimoto, M. Sumimoto, T. Imahori, K. Igawa, K. Tomooka,
R. Irie, Eur. J. Org. Chem. 2017, 6914; c) X. Wu, L. Xue, D. Li, S. Jia, J.
[33] For an example of investigation of thermal stability of silicon cotaining
fluorescent compounds by TG-DTA, see: Y. Yamanoi, T. Nakashima, M.
Shimada, H. Maeda, H. Nishihara, Molecules 2016, 21, 1173.
[34] For a recent review on OLED, see: Y. Liu, X. Tao, F. Wang, X. Dang, D.
Zou, Y. Ren, M. Jiang, J. Phys. Chem. C 2008, 112, 3975.
This article is protected by copyright. All rights reserved.

10.1002/chem.201800381
Chemistry - A European Journal
FULL PAPER
FULL PAPER
10 mol %
[Rh(cod)2]BF4/(S)-binap
MeOH (0.1 equiv)
R
iPr
iPr
Tomoya Namba, Yoshihiro Hayashi,
Si
Susumu Kawauchi, Yu Shibata, and Ken
Tanaka*
(CH2Cl)2, 60 °C
O
Si
R
R
O
iPr
R
OH
HO
1,2-Si-migration
iPr
&
up to 81% yield
highly fluorescent
strong narrow-band
1,3-C-migration
&
Oxycyclization
Page No. – Page No.
fluorescence in film and solid-states
(R = p-Me)
Rhodium-Catalyzed Cascade
Synthesis of
Benzofuranylmethylidene-
Benzoxasiloles: Elucidating Reaction
Mechanism and Efficient Solid-State
Fluorescence
A new three-step route to highly fluorescent benzofuranylmethylidenebenzoxa-
siloles via the cationic rhodium(I) complex-catalyzed cascade cycloisomerization of
bis(2-ethynylphenol)silanes has been developed. The thus obtained benzofuranyl-
methylidene-benzoxasiloles were highly emissive, especially, 4-methylphenoxy
derivative exhibited strong narrow-band fluorescence in film state (blue, φF = 95%)
arising from the monomer and the red-shifted strong narrow-band fluorescence in
crystal state (green, φF = 90%) arising from formation of an offset π−stacked dimer.
This article is protected by copyright. All rights reserved.