Silylation of Aryl Halides with Monoorganosilanes Activated by Lithium Alkoxide

Article abstract of DOI:10.1021/acs.orglett.8b00818

Lithium alkoxide activates a monoorganosilane to generate a transient LiH/alkoxysilane complex, which quickly reacts with aryl and alkenyl halides at 25 °C to deliver a diorganosilane product. Experimental and theoretical studies suggest that the reaction includes nucleophilic attack of LiH on the halogen atom of the organic halide to generate a transient organolithium/alkoxysilane intermediate, which undergoes quick carbon-silicon bond formation within the complex.

Full text of DOI:10.1021/acs.orglett.8b00818

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
Silylation of Aryl Halides with Monoorganosilanes Activated by
Lithium Alkoxide
Takumi Yoshida, Laurean Ilies,*,† and Eiichi Nakamura*
Department of Chemistry, School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
S
* Supporting Information
ABSTRACT: Lithium alkoxide activates a monoorganosilane to generate a transient
LiH/alkoxysilane complex, which quickly reacts with aryl and alkenyl halides at 25 °C to
deliver a diorganosilane product. Experimental and theoretical studies suggest that the
reaction includes nucleophilic attack of LiH on the halogen atom of the organic halide to
generate a transient organolithium/alkoxysilane intermediate, which undergoes quick
carbon−silicon bond formation within the complex.
ong regarded largely as a base, alkali metal hydrides are
and theoretical studies suggest an intriguing new possibility for
1
2
Lfinding new utility as a reducing agent. Our interest in generating a metal hydride species, including a LiH/
alkoxysilane complex (A) and a transient LiH (B), which
nucleophilically reduces the C−X bond in R2X. The reaction
bifurcates into silylation (a) and reduction (b) after a transient
R2−Li intermediate (C).
mild methods for the generation of reactive metal hydride
reagents led us to find a new way to cross-couple a hydrosilane
and an aryl and alkenyl halide to produce diorganosilanes, a
valuable synthetic intermediate for organic synthesis.3 However,
there is a paucity of selective synthetic routes to this class of
compounds4 due to the high reactivity of diorganosilanes
leading to disproportionation or overreaction.5,6 Herein we
describe that a transient LiH/alkoxysilane, which is generated
from two air stable reagents, monoorganosilane (R1SiH3) and
lithium tert-butoxide (LiOt-Bu),7,8 reacts with an aryl or alkenyl
halide (R2X) to produce a diorganosilane (Figure 1). The
synthetic merits include simplicity, scalability, and quickness
and mildness of the reaction often finishing within 10 min at 25
°C under dry air or argon. The reactive species is rather
insensitive to steric hindrance, and the reaction does not
produce tri- or tetraorganosilane side products. Experimental
Equation 1 shows a typical example of the silylation reaction
for 1-iodonaphthalene (1): To a solution of lithium tert-
butoxide (20 mmol, 1 M) in THF at 25 °C under argon were
added sequentially phenylsilane (20 mmol) and 1-iodonaph-
thalene (10 mmol), which was followed by hydrogen gas
1
evolution (identified by H NMR) and an exothermic reaction
for several minutes. As analyzed by GC, GC-MS, and 1H NMR,
the reaction produced 1-naphtylphenylsilane (2; 1.18 g, 51%
yield, isolated by silica gel chromatography), naphthalene (3;
43%), and di(tert-butoxy)phenylsilane (∼25% based on
phenylsilane; this product presumably forms by the reaction
of (tert-butoxy)phenylsilane, the formation of which was
detected by NMR (SI), and tert-butanol in the presence of
lithium tert-butoxide). The reaction under dry air9 took place
equally well in 50% yield, while the reaction was slightly
sensitive to moisture in air (44% yield). The silylation/
reduction ratio did not change much either by the increase/
decrease of PhSiH3 or by use of PhSiD3 (i.e., lack of kinetic
isotope effect; Supporting Information, SI).
The choice of the base much affects the reaction (Table 1;
additional data in SI). In the absence of a base, 1-
iodonaphthalene and phenylsilane alone did not react at all
(entry 1). Both secondary and tertiary lithium alkoxides reacted
equally well in THF (entries 2 and 3), while primary alkoxide
and phenoxide were unreactive (entries 4 and 5). 12-Crown-4-
ether did not significantly affect the reaction rate or product
selectivity (entry 6). Potassium and sodium tert-butoxide
Received: March 13, 2018
Figure 1. Reaction of monoorganosilane with aryl and alkenyl halide.
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.8b00818
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 1. Silylation of Aryl and Alkenyl Bromides or
a
Iodides
(entries 7, 8) gave a complex mixture of silylation and
reduction products as well as products due to silane
disproportionation (silane, diphenylsilane, and triphenylsilane).
No reaction took place in diethyl ether, dichloromethane,
toluene, and hexane (SI). Control experiments discounted the
possibility of catalysis by contaminated transition metals.10
We next investigated the scope of the reaction between aryl
and alkenyl halides, and monoorganosilanes (Scheme 1). All
reactions underwent 100% conversion (except for 23 (no
reaction) and some aryl bromides, 4−7, 9, 11), and the
reduction product accounts for the rest of the material balance.
The reduction products and alkoxysilane can be easily
separated by column chromatography. Iodobenzene and 4-
iodotoluene underwent silylation in about 50% yield (4 and 5),
while the corresponding bromides were unreactive. Radical
clock substrate 7 did not give any cyclization product.11
Interestingly, the reaction tolerates ortho-substitution, which
even promotes silylation over reduction (cf. 5 vs 6−11 and
iodide vs bromide). For instance, 2,4,6-triisopropyl-iodoben-
zene and mesityl iodide underwent silylation (10 and 11) in
higher yield than iodobenzene, although the reaction required a
higher temperature (50 °C instead of 25 °C). Here, the
intermediacy of benzyne can be discounted. 2,4,6-Tri-tert-
butyliodobenzene is apparently too hindered to be silylated, but
still underwent smooth C−I bond cleavage to give the
reduction product in 90% yield (12).
Electron-poor iodoarenes reacted faster but were silylated in
low yield (18, 19), while electron-rich iodoarenes reacted
slower but were silylated in higher yield (15−17). We discuss
further this interesting contrast in the next paragraph. The
reaction tolerates the presence of a sulfur group (20 and 25).
An alkylsilane also gave the silylated compounds (24, 25) albeit
in low yield. A trisubstituted alkenyl bromide reacted well to
give a tetrasubstituted alkenylsilane (21), while simple alkenyl
halides were very poor yielding. Fluoro, chloro, tosyl, and
trifluoromethanesulfonate compounds did not react at all
(unreacted substrates are listed in SI).12
a
Aryl iodide (0.5 mmol), silane (1.0 mmol), lithium tert-butoxide (1.0
mmol) in tetrahydrofuran (1.0 mL) at 25 °C under argon for 10 min,
unless mentioned otherwise. Aryl or alkenyl bromides: aryl bromide
(0.5 mmol), silane (1.5 mmol), lithium tert-butoxide (1.5 mmol) in
b
tetrahydrofuran (1.5 mL) at 70 °C under argon for 3 h. 25 °C, 3 h.
c
1
Yield determined by HNMR with 1,1,2,2-tetrachloroethane as an
d
e
f
g
internal standard. 50 °C, 1 h. 70 °C, 3 h. 1 mmol scale. 10 mmol
scale.
(Figure 2a). As mentioned for 15 and 16 vs 18 in (Scheme 1),
the 4-fluoro substituent favors reduction, while the 4-alkoxy
substituent favors silylation. The ratio of silylation/reduction
remained the same during the reaction, excluding proto-
desilylation of the product.
The electronegativity of the 4-substituent affects both the
reactivity and the silylation/reduction selectivity in an
interesting way. Thus, in a competitive reaction, 4-fluoroiodo-
benzene reacts markedly faster than 4-methoxyiodobenzene
(no reaction) with silylation/reduction selectivity as low as 1:4
A mechanistic issue is thus whether the C−X bond cleavage
includes electron donation directly to C−X σ* or indirectly to
π* of the aromatic ring. To probe this issue, we examined
competition between a 4-bromostilbene and 4-trifluoromethyl-
1-bromobenzene13 and found exclusive reaction of the latter
a
Table 1. Base Effects on Silylation of 1-Iodonaphthalene (1) with Phenylsilane
b
b
b
b
entry
base
none
2 (%)
3 (%)
entry
5
base
2 (%)
3 (%)
1
2
3
4
0
51
50
0
0
50
40
0
LiOPh
0
26
19
3
0
63
44
26
c
LiOt-Bu
LiOi-Pr
LiOEt
6
LiOt-Bu
KOt-Bu
NaOt-Bu
7
8
a
1-Iodonaphthalene (1, 0.20 mmol), phenylsilane (0.40 mmol), base (0.40 mmol) in tetrahydrofuran (0.40 mL) at 25 °C under argon for 10 min.
GC yield with tridecane as an internal standard. With 2.0 equiv of 12-crown-4-ether.
b
c
B
DOI: 10.1021/acs.orglett.8b00818
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Figure 2. Competition experiments as mechanistic probe. (a) Between
4-methoxyiodobenzene and 4-fluoroiodobenzene. (b) Between 4-
bromostilbene and 4-trifluoromethylbromobenzene. Orbital levels at
the B3LYP/6-31++G(d) level of theory, except SDD basis set and
ECP46MWB effective core potential for iodine atom. E0 determined
by DPV measurement.
Figure 3. Reaction pathway obtained by DFT calculations (M062X/6-
311++G(d,p)//B3LYP/6-31++G(d)). Gibbs energies (X = I) are
shown in kcal/mol. The numbers refer to bond length (Å) and natural
charges (underlined). (a) Generation of transient LiH. (b) Rate-
determining step: generation of aryllithiate complex from transient
LiH. (c) Product selectivity determining step: silylation vs reduction.
(Figure 2b). Thus, the C−Br bond cleavage is driven by
electron donation from LiH to the C−Br σ* bond rather than
to the aromatic π* orbital (B in Figure 1).
an electron-donating group destabilized TS1 and, hence,
decelerates the C−X bond cleavage as seen experimentally
(see above). This expectation was supported by theory; that is,
the calculated energy barrier increases in the order of 4-F, 4-H,
and 4-MeO as shown in Figure 4a.
We made additional observations. First, no EPR signals were
detected during the reaction, and a radical scavenger (TEMPO)
showed no effects on the reaction yield and rate. The tolerance
of air attests to the absence of radical species. Second, the
reaction rate examined for phenylsilane, lithium tert-butoxide,
and 1-iodonaphthalene showed 0.94th, 0.95th, and 0.48th order
dependence, respectively, indicating participation of all
reactants in the rate limiting transition state and probable
involvement of pre-equilibria before the formation of INT3.14
These observations were corroborated by a reaction pathway
found by the DFT calculation summarized in Figure 3. The
model here is arguably too simplified, yet provides insights into
the mechanism of reductive cleavage of the C−X bond. The
reaction comprises three stages, generation of a LiH reactive
species (a), C−X bond cleavage (b), and C−Si bond formation
(c).
In Figure 3a, the silane and the alkoxide first form a silicate
INT1 and then a LiH/alkoxysilane complex INT2 with energy
loss. In agreement with these energetics, we could not detect
these species by 29Si NMR or 1H NMR. In Figure 3b, LiH starts
donating electrons to the σ* orbital of the C−halide bond
(INT3), where a linear H−X−C geometry is characteristic, and
accounts for the observed low steric sensitivity of the reaction.
There was no sign of aromatic π-orbital participation as
suggested by the experiment in Figure 2b. The transition state
of the C−X bond cleavage is early and structurally similar to
INT3. It goes downhill to INT4, a ternary complex among
aryllithium, alkoxysilane, and HX. The −0.666 charge on the H
atom in INT3 becomes near zero in INT4, and the negative
charge accumulates on the aromatic ring.15 Here we expect that
Figure 4. Gibbs free energy diagrams for the reaction with 4-fluoro-1-
iodobenzene, iodobenzene, and 4-methoxy-1-iodobenzene (a) As to
TS1. (b) As to TS2.
INT4 is arguably an unstable species and decomposes
immediately to stable products in Figure 3c, where the
silylation/reduction selectivity is determined. The silylation
path via TS2 is an intramolecular substitution on the silicon
atom within INT4, and the reduction step via TS3 comprises
protonation of the aryllithium moiety of INT4. The calculated
C
DOI: 10.1021/acs.orglett.8b00818
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Chem. Soc. 2007, 129, 14164−14165. (d) Kuninobu, Y.; Yamauchi, K.;
Tamura, N.; Seiki, T.; Takai, K. Angew. Chem., Int. Ed. 2013, 52, 1520−
1522.
(4) Selected examples using organometallics: (a) Hirone, N.; Sanjiki,
H.; Tanaka, R.; Hata, T.; Urabe, H. Angew. Chem., Int. Ed. 2010, 49,
activation energy of the silylation path also shows correlation to
the 4-substituent (Figure 4b), indicating that an electron-
withdrawing substituent stabilized the aryllithium intermediate
and hence retarded the silylation relative to the protonation
pathway. This is opposite to the one found for the C−X
cleavage. The experimental observation in Figure 2a is thus
supported.16
In summary, we have generated a transient LiH species under
mild conditions from air stable reagents and discovered its
ability to reductively silylate aromatic and alkenyl halides. The
silylation reaction reported here shows some resemblance to
the C−H silylation reported by Stoltz and Grubbs,17 which
proceeds however under markedly different conditions and
presumably through different mechanisms.18 Several puzzling
reactivity profiles found in the synthetic experiments were
reconciled by the mechanism proposed with the aid of theory,
which will provide useful guidelines for future development of
new alkali metal hydride chemistry.
7762−7764. (b) Fridel−Crafts C−H silylation: Bahr, S.; Oestreich, D.
̈
Angew. Chem., Int. Ed. 2017, 56, 52−59 and references therein.
(c) Marciniec, B. Hydrosilylation: A Comprehensive Review on Recent
Advances; Springer Science; 2009, 1−418.
(5) Lesbani, A.; Kondo, H.; Yabusaki, Y.; Nakai, M.; Yamanoi, Y.;
Nishihara, H. Chem. - Eur. J. 2010, 16, 13519−13527.
(6) A review for the silylation reaction: (a) Xu, Z.; Xu, L. W.
ChemSusChem 2015, 8, 2176−2179. (b) Sharma, R.; Kumar, R.;
Kumar, I.; Singh, B.; Sharma, U. Synthesis 2015, 47, 2347−2366.
(c) Xu, Z.; Huang, W. S.; Zhang, J.; Xu, L. W. Synthesis 2015, 47,
3645−3668.
(7) Dehydrogenative silylation with solid bases including metal
alkoxides and hydrosilanes: (a) Calas, R.; Bourgeois, P. C. R. Acad. Sci.,
Paris, Ser. C 1969, 268, 3645−3668. (b) Baba, T.; Kato, A.; Ono, Y.
Catal. Today 1998, 44, 271−276. (c) Itoh, M.; Iwata, K.; Inoue, K. J.
Organomet. Chem. 1994, 476, C30−C31. (d) Ishikawa, J. I.; Inoue, K.;
Itoh, M. J. Organomet. Chem. 1998, 552, 303−311. (e) Ishikawa, J. I.;
Itoh, M. J. Organomet. Chem. 1998, 552, 303−311.
(8) Metal-alkoxide mediated silylation of aryl halides: Dervan, P. B.;
Shippey, M. A. J. Org. Chem. 1977, 42, 2654−2655.
(9) The reaction vessel was connected to atmospheric air through a
column of anhydrous CaCl2.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.8b00818.
(10) ICP analysis revealed only a trace amount of transition metal
contamination. Addition of a catalytic amount of a transition metal did
not accelerate the reaction. Use of several lithium tert-butoxides from
different sources gave the same results. See SI for details.
Experimental procedures and physical properties of the
compounds (PDF)
AUTHOR INFORMATION
Corresponding Authors
́
(11) (a) Pan, X.; Lacote, E.; Lalevee, J.; Curran, P. D. J. Am. Chem.
■
Soc. 2012, 134, 5669−5674. (b) Zhou, B.; Sato, H.; Ilies, L.;
Nakamura, E. ACS Catal. 2018, 8, 8−11.
*E-mail: laur@chem.s.u-tokyo.ac.jp.
*E-mail: nakamura@chem.s.u-tokyo.ac.jp.
ORCID
(12) For a list of substrates investigated, see SI.
(13) (a) Shirakawa, E.; Hayashi, Y.; Itoh, K.; Watabe, R.; Uchiyama,
N.; Konagaya, W.; Masui, S.; Hayashi, T. Angew. Chem., Int. Ed. 2012,
51, 218−221. (b) Yamamoto, E.; Ukigai, S.; Ito, H. Chem. Sci. 2015, 6,
2943−2951.
(14) For a detailed discussion, see SI.
(15) Attempts to trap this anionic species failed, maybe due to its
short lifetime. See Scheme S8 in SI.
Takumi Yoshida: 0000-0001-5006-3075
Laurean Ilies: 0000-0002-0514-2740
Eiichi Nakamura: 0000-0002-4192-1741
Present Address
†RIKEN Center for Sustainable Resource Science, 2-1
Hirosawa, Wako, Saitama 351-0198, Japan.
Notes
(16) The reactivity difference between different substrates (i.e.,
bromobenzene vs iodobenzene) is well explained by the proposed
mechanism. See Table S1 in the SI.
(17) (a) Fedorov, A.; Toutov, A. A.; Swisher, N. A.; Grubbs, R. H.
Chem. Sci. 2013, 4, 1640−1645. (b) Toutov, A. A.; Liu, W. B.; Betz, K.
N.; Fedorov, A.; Stoltz, B. M.; Grubbs, R. H. Nature 2015, 518, 80−84.
(c) Toutov, A. A.; Liu, W. B.; Betz, K. N.; Stoltz, B. M.; Grubbs, R. H.
Nat. Protoc. 2015, 10, 1897−1903. Toutov, A. A.; Salata, M.; Fedorov,
A.; Shabaker, J. W.; Houk, K. N.; Grubbs, R. H. Nat. Energy 2017, 2,
17008.
(18) (a) Liu, W.-B.; Schuman, D. P.; Yang, Y.-F.; Toutov, A. A.;
Liang, Y.; Klare, H. F. T.; Nesnas, N.; Oestreich, M.; Blackmond, D.
G.; Virgil, S. C.; Banerjee, S.; Zare, R. N.; Grubbs, R. H.; Houk, K. N.;
Stoltz, B. M. J. Am. Chem. Soc. 2017, 139, 6867−6879. (b) Banerjee,
S.; Yang, Y.-F.; Jenkins, I. D.; Liang, Y.; Toutov, A. A.; Liu, W.-B.;
Schuman, D. P.; Grubbs, R. H.; Stoltz, B. M.; Krenske, E. H.; Houk, D.
P.; Zare, R. N. J. Am. Chem. Soc. 2017, 139, 6880−6887.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank MEXT for financial support KAKENHI (15H05754
for E.N. and JP26708011 to L.I.). T.Y. thanks the University of
Tokyo’s “Evonik Scholars Fund” Scholarship. This work was
partially supported by CREST, JST (Molecular Technology).
We thank for Dr. Yasukawa Tomohiro (The University of
Tokyo) for assistance with the ICP-AES analysis.
REFERENCES
■
(1) (a) Too, P. C.; Chan, G. H.; Tnay, Y. L.; Hirao, H.; Chiba, S.
Angew. Chem., Int. Ed. 2016, 55, 3719−3723. (b) Hong, Z.; Ong, D. Y.;
Muduli, S. K.; Too, P. C.; Chan, G. H.; Tnay, Y. L.; Chiba, S.;
Nishiyama, Y.; Hirao, H.; Soo, H. S. Chem. - Eur. J. 2016, 22, 7108−
7114. (c) Ong, D. Y.; Tejo, C.; Xu, K.; Hirao, H.; Chiba, S. Angew.
Chem., Int. Ed. 2017, 56, 1840−1844.
(2) Ilies, L.; Yoshida, T.; Nakamura, E. J. Am. Chem. Soc. 2012, 134,
16951−16954.
(3) For example: (a) Corey, J. Y. Adv. Organomet. Chem. 2004, 51,
1−52. (b) Nakajima, Y.; Shimada, S. RSC Adv. 2015, 5, 20603−20616.
(c) Fukazawa, A.; Li, Y.; Yamaguchi, S.; Tsuji, H.; Tamao, K. J. Am.
D
DOI: 10.1021/acs.orglett.8b00818
Org. Lett. XXXX, XXX, XXX−XXX