Allyl-transfer reaction from photoexcited hypervalent allylsilicon reagent toward dicyanobenzenes

Article abstract of DOI:10.1246/cl.140940

Photoreaction between dicyanobenzenes and a hypervalent allylsilicon reagent us ing 2,3-dihydroxynaphtha lene as a ligand on silicon efficiently proceeded to afford a llyl-substituted benzonitriles in moderate to good yields. The present photoreaction was

Full text of DOI:10.1246/cl.140940

Received: October 11, 2014 | Accepted: November 8, 2014 | Web Released: November 18, 2014
CL-140940
Allyl-transfer Reaction from Photoexcited Hypervalent Allylsilicon Reagent
toward Dicyanobenzenes
Daisuke Matsuoka and Yutaka Nishigaichi*
Department of Material Science, Graduate School of Science and Engineering, Shimane University,
1060 Nishikawatsu-cho, Matsue, Shimane 690-8504
(E-mail: nishigai@riko.shimane-u.ac.jp)
Photoreaction between dicyanobenzenes and a hypervalent
allylsilicon reagent using 2,3-dihydroxynaphthalene as a ligand
on silicon efficiently proceeded to afford allyl-substituted
benzonitriles in moderate to good yields. The present photo-
reaction was found to occur based on the excitation of the
hypervalent allylsilicon reagent.
Table 1. Photosubstitution reaction of p-dicyanobenzenea
CN
Si
+
2
NC
Light ( > 280 nm)
1a
Photosensitizer
DMF, rt, 7 h
NC
3a
Photoinduced electron-transfer (PET) reaction, which gives
a radical ion pair (D + A ¼ D•+ + A•¹), is an important
reaction employed as a key step in organic photochemistry.
However, the competitive back electron transfer (BET; from A•¹
to D•+) decreases the efficiency of PET, thus various methods to
suppress BET have been developed.1 Among them, the use of
group 14 organometallic compounds has received considerable
attention due to their easy handling and C-M bond cleavage of
D•+.2-8 Since allylsilicon reagents such as allyltrimethylsilane
(2a) possess a relatively low oxidation potential, it is already
known to behave as an electron donor in the photoallylation
of iminium salts,4 aromatic imides,5 unsaturated nitriles,6 and
aromatic nitriles.7 We have reported that hypervalent allylsilicon
reagent 2b was found to facilitate the photoallylation reaction of
carbonyl compounds more effectively than allylsilane 2a.8 The
reason for the efficiency is much lower oxidation potential of 2b
than 2a.
Me4N
2a: Si = SiMe3 2b: Si =
O
O
Si
O O
Me4N
O
O
2c: Si =
Si
O O
Silicon
reagent
EOX of
2/V
Entry
Photosensitizer
Yield/%b
1
2
3
4c
5
2a
2b
2b
2c
2c
+1.64
+1.12
anthracene
anthracene
none
anthracene
none
0
82
8
85
78
+0.99
aReaction conditions:
a mixture of 1a (0.2 mmol), 2
(0.3 mmol), and photosensitizer (0.1 mmol) in DMF (10 mL)
in a Pyrex test tube was irradiated by a high-pressure mercury
lamp under N2. Isolated yield. Reaction time was 4 h.
Recently, we have also reported efficient photoreaction of
dicyanoarenes with 2b.9 In contrast to the photoreaction of p-
dicyanobenzene (1a) with 2a (Table 1, Entry 1),10 photoreaction
with 2b proceeded effectively to afford p-allylbenzonitrile (3a)
in 82% yield after 7 h (Table 1, Entry 2). The presence of a
photosensitizer was critical; without anthracene, only 8% of 3a
was obtained (Entry 3). When a novel hypervalent allylsilicon
reagent 2c, which was prepared from allyltrimethoxysilane, 2,3-
dihydroxynaphthalene, and tetramethylammonium hydroxide,
was employed under similar reaction conditions, 1a was
completely consumed after 4 h to afford the product 3a in
85% yield (Entry 4). Interestingly, photoreaction of 1a with 2c
in the absence of anthracene also proceeded to give 3a in 78%
yield (Entry 5), in contrast to the result in Entry 3. However, it
was difficult to explain the large difference of the reactivity for
this photoreaction by the small difference of oxidation potentials
between 2b and 2c.
b
c
2
2c
1
To explore the reason of the large enhancement of the
photoreaction with 2c, UV-vis spectra of hypervalent allylsili-
con reagents 2b and 2c along with p-dicyanobenzene (1a) were
measured (Figure 1). The UV-vis spectrum of 2c showed
characteristic absorption between 310 and 360 nm in contrast to
those of 2b and 1a. This observation prompted us to assume
that the photoreaction between 1a and 2c proceeded by the
preferential excitation of 2c. To demonstrate our hypothesis, we
2b
1a
0
280
300
320
340
360
380
Wavelength / nm
Figure 1. UV-vis spectra of 2b, 2c, and 1a (1.0 © 10¹4 M in
DMF).
Chem. Lett. 2015, 44, 163–165 | doi:10.1246/cl.140940
© 2015 The Chemical Society of Japan | 163

Table 2. Photosubstitution reaction of 1a under the irradiation
of various lighta
CN
CN
h
Me4N
2c*
PET
2c
Yield of 3a/%b
O
O
Entry
Silicon reagent
Light, -/nm
NC
NC
Si
1a
1
2
3
4
2c
2c
2b
2b
>310
>330
>310
>330
68
45
trace
trace
O O
2c
C–Si bond
cleavage
aReaction conditions: as shown in Table 1 with an appropriate
filter solution. Also see Supporting Information. bIsolated yield
of 3a.
1a
CN
– CN
NC
NC
3a
Scheme 1. Plausible reaction mechanism.
Table 3. Photosubstitution reaction of 1a with hypervalent
organosilicon reagents 2a
R
Me4N
O
O
1a
+
Si
O O
2
Figure 2. Stern-Volmer plot of 2c with 1a as a quencher.
R
Light
next investigated the photoreaction using light of - > 310 nm11
and > 330 nm,12 which were not absorbed by 1a. The results are
summarized in Table 2. As expected, the irradiation of light of
- > 310 nm afforded the product 3a in a good yield (Entry 1).
Furthermore, the irradiation at even longer wavelength also
afforded 3a in a moderate yield (Entry 2). In contrast, no
reaction proceeded with 2b under the irradiation of - > 310 nm
or > 330 nm light and 1a was recovered almost quantitatively
(Entries 3 and 4). These results clearly suggested that photo-
reaction as shown in Table 2 proceeded based on the excitation
of allylsilicon reagent 2c.
NC
DMF, rt, 7 h
3
Entry
R (2)
Light,
λ
/nm
Product Yield/%b
1
>280
>310
>330
>280
>310
>330
3b
58
48
38
53
21
10
2
2d
3
4c
5c
6c
3c
PhCH2
2e
This characteristic feature was further confirmed by the
fluorescence quenching experiment of 2c. No charge-transfer
interaction between silicon reagent 2c and 1a was detected in the
UV-vis spectra. The fluorescence maximum of 2c was observed
at 348 nm and efficiently quenched by 1a in acetonitrile. A
aReaction conditions: as shown in Table 1 with an appropriate
filter solution. Also see Supporting Information. bIsolated
c
yield. DMF (5 mL) was used.
¹1
Stern-Volmer plot showed good linearity with kq¸ = 105.9 M
(Figure 2). In addition, the free energy change ¦G for the single
electron transfer from excited 2c (S1 state) to 1a was calculated
to be negative (¹25.4 kcal mol¹1) in acetonitrile.13 These results
indicated the S1 state of 2c was quenched by PET to give 1•¹ and
2c•+, then allyl transfer occurred according to radical mechanism
accompanied by the elimination of silicate (C-Si bond cleavage)
(Scheme 1).9,14
Next, we investigated the photoreaction of analogous
hypervalent organosilicon reagents 2d and 2e with 1a under
irradiation with - > 280 nm, > 310 nm, and > 330 nm. The
results are shown in Table 3. β-Methallylsilicon reagent 2d
afforded the corresponding product 3b in the indicated yields
(Entries 1-3). Selective irradiation of 2d also promoted the
reaction (Entries 2 and 3). Benzylsilicon reagent 2e was also
attempted under the same conditions and was found to be
reactive toward 1a affording 3c (Entries 4-6). But the efficiency
was lower than those of 2c and 2d, especially when irradiated at
the longer wavelength. The reason is so far unclear, but stability
of the radical cation 2•+ may be involved (C-Si bond cleavage
step),15 because oxidation potentials of 2d (+0.80 V) and 2e
(+0.80 V) are lower than that of 2c and thus the efficiency of
PET step should be also high.
To confirm the generality of the photoreaction via excited 2,
we then examined the application to other isomers of dicyano-
benzene. Both o-dicyanobenzene (1b) and m-dicyanobenzene
(1c) underwent photoallylation with allylsilicon reagent 2c.
Here again, irradiations with three different wavelengths were
attempted (Table 4). In every case, the corresponding allylated
products were obtained. In the case of 1b, the cyano-substituted
164 | Chem. Lett. 2015, 44, 163–165 | doi:10.1246/cl.140940
© 2015 The Chemical Society of Japan

Table 4. Photosubstitution reaction of o- and m-dicyanobenzenes (1b and 1c) with hypervalent organosilicon reagents 2ca
Yield/%b
Entry
1
Light,
λ
/nm
Products
4, 6
56
5, 7
13
11
7
CN
CN
1
2
3
4
5
6
1b
>280
>310
>330
>280
>310
>330
33
CN
4
5
15
NC
NC
CN
1c
7
28
18
18
trace
trace
6
7
b
aReaction conditions: as shown in Table 1 with an appropriate filter solution. Also see Supporting Information. Isolated yield.
4 was mainly provided, accompanied by the hydrogen-substi-
tuted 5 (Entries 1-3). In contrast, the reaction with 1c hardly
afforded the cyano-substituted 6, while the hydrogen-substituted
7 was mainly obtained in low yield (Entries 4-6).
Acc. Chem. Res. 2011, 44, 204.
4
a) K. Ohga, P. S. Mariano, J. Am. Chem. Soc. 1982, 104,
617. b) A. J. Y. Lan, R. O. Heuckeroth, P. S. Mariano, J. Am.
Chem. Soc. 1987, 109, 2738. c) I.-S. Cho, C.-L. Tu, P. S.
Mariano, J. Am. Chem. Soc. 1990, 112, 3594. d) S.
Fukuzumi, M. Fujita, S. Noura, K. Ohkubo, T. Suenobu,
Y. Araki, O. Ito, J. Phys. Chem. A 2001, 105, 1857.
Y. Kubo, T. Imaoka, T. Shiragami, T. Araki, Chem. Lett.
1986, 1749.
K. Mizuno, M. Ikeda, Y. Otsuji, Chem. Lett. 1988, 1507.
a) K. Mizuno, K. Nakanishi, Y. Otsuji, Chem. Lett. 1988,
1833. b) M. Mella, E. Fasani, A. Albini, J. Org. Chem. 1992,
57, 6210. c) K. Nakanishi, K. Mizuno, Y. Otsuji, Bull. Chem.
Soc. Jpn. 1993, 66, 2371. d) T. Tamai, K. Mizuno, I.
Hashida, Y. Otsuji, Bull. Chem. Soc. Jpn. 1993, 66, 3747.
a) Y. Nishigaichi, A. Suzuki, T. Saito, A. Takuwa,
Tetrahedron Lett. 2005, 46, 5149. b) Y. Nishigaichi, A.
Suzuki, A. Takuwa, Tetrahedron Lett. 2007, 48, 211.
D. Matsuoka, Y. Nishigaichi, Chem. Lett. 2014, 43, 559.
In summary, it was found that photoreaction of dicynoben-
zenes 1 with hypervalent organosilicon reagents 2c using the
2,3-dihydroxynaphthalene as a ligand proceeded via photo-
excited 2c. To the best of our knowledge, the photopromoted
allylation reactions with organometallic reagents were generally
initiated by the excitation of the electron acceptor or the
sensitizer. This is the first example of photoallylation based on
the excitation of organosilicon reagent.16 The present approach
would enable photoreaction to proceed independent from the
absorption wavelength of the electron accepting substrate.
Further investigation for the application to a variety of substrates
using the present approach is ongoing in our laboratory.
5
6
7
8
9
The authors thank Dr. Takehiro Ohta and Prof. Yoshinori
Naruta, Institute for Materials Chemistry and Engineering,
Kyushu University, for their measurement of the ESI-MS
spectrum.
10 In ref 7c, Mizuno et al. reported a good yield of 3a in the
reaction between 1a and 2a, though large excess of 2a,
long reaction time, and phenanthrene as a sensitizer were
employed.
Supporting Information is available electronically on J-STAGE.
11 T. Sakurai, H. Sukegawa, H. Inoue, Bull. Chem. Soc. Jpn.
1985, 58, 2875.
References and Notes
1
2
a) A. G. Griesbeck, N. Hoffmann, K. Warzecha, Acc. Chem.
Res. 2007, 40, 128. b) N. Hoffmann, J. Photochem.
Photobiol., C 2008, 9, 43.
12 Acetone filter was used.
13 Oxidation potential of 2c was measured to be 0.99 V.
Reduction potential of 1a was ¹1.53 V.9 E0-0 value for 2c is
taken as 82.2 kcal mol¹1. ¦G is determined according to the
following literature. D. Rehm, A. Weller, Isr. J. Chem. 1970,
8, 259.
14 In the reaction mixture, hexacoordinate tris(2,3-naphthale-
nediolato)silicate was observed by means of 1H NMR. It was
probably derived by the ligand redistribution from the
eliminated silicon moiety. A. Suvitha, B. Varghese, M. N. S.
Rao, G. Sundararajan, B. Viswanathan, Indian J. Chem.
2006, 45A, 2193.
a) A. Takuwa, H. Tagawa, H. Iwamoto, O. Soga, K.
Maruyama, Chem. Lett. 1987, 1091. b) A. Takuwa, Y.
Nishigaichi, K. Yamashita, H. Iwamoto, Chem. Lett. 1990,
639. c) A. Takuwa, Y. Nishigaichi, K. Yamashita, H.
Iwamoto, Chem. Lett. 1990, 1761. d) A. Takuwa, Y.
Nishigaichi, H. Iwamoto, Chem. Lett. 1991, 1013. e) Y.
Nishigaichi, N. Yoshida, M. Matsuura, A. Takuwa, Chem.
Lett. 1999, 803. f) S. Kyushin, S. Otani, T. Takahashi, Y.
Nakadaira, M. Ohashi, Chem. Lett. 1991, 775.
3
a) E. Hasegawa, W. Xu, P. S. Mariano, U. C. Yoon, J. U.
Kim, J. Am. Chem. Soc. 1988, 110, 8099. b) U. C. Yoon,
Y. C. Kim, J. J. Choi, D. U. Kim, P. S. Mariano, I.-S. Cho,
Y. T. Jeon, J. Org. Chem. 1992, 57, 1422. c) E. Hasegawa,
M. A. Brumfield, P. S. Mariano, U.-C. Yoon, J. Org. Chem.
1988, 53, 5435. d) D. W. Cho, U. C. Yoon, P. S. Mariano,
15 K. P. Dockery, J. P. Dinnocenzo, S. Farid, J. L. Goodman,
I. R. Gould, W. P. Todd, J. Am. Chem. Soc. 1997, 119, 1876.
16 Photopromoted intramolecular alkyl-transfer of 6-coordinate
organosilicon reagents: a) J. Wagler, T. Doert, G. Roewer,
Angew. Chem., Int. Ed. 2004, 43, 2441. b) J. Wagler, G.
Roewer, D. Gerlach, Z. Anorg. Allg. Chem. 2009, 635, 1279.
Chem. Lett. 2015, 44, 163–165 | doi:10.1246/cl.140940
© 2015 The Chemical Society of Japan | 165