Journal of the American Chemical Society
Communication
a
Table 2. Regioselectivity: the present stannylation reaction
occurs with very high regiospecificity for terminal alkynes (no
regioisomer was detected), and the linear 2-functionalized
vinylstannanes were obtained exclusively in all cases. As
evidenced by alkynes 1c, 1d, and 1j, the stannylation process is
compatible with sterically demanding substituents on the
terminal alkynes. Stereoselectivity: the E/Z ratios of the resultant
vinylstannanes were usually high, with the E isomer as the
major component. Chemoselectivity: first, various polar func-
tional groups, including alkyl chloride (1e), propargyl ether
(1f), cyano (1g), alcohols (1h and 1j), and π-deficient
pyridines (1x and 1y), which are sensitive to stannyl anions or
stannyl metal reagents, are tolerated in the reaction. Second,
radical-susceptible moieties such as benzyl groups (1b and 1u),
cyclopropyl groups (1c), alkyl halide (1e), and pyridine (1x
and 1y) are compatible with this reaction,40 suggesting a very
different character from stannyl radicals. Terminal alkyne
specificity: aliphatic internal alkyne 4a and diphenyl acetylene
4b did not provide any stannylated product (5a and 5b). A
distal triple bond was untouched in the case of the conjugate
diyne 1i, affording the linear vinylstannane 3i exclusively. As
for (terminal) olefins, aryl olefin 4c gave the corresponding
phenethylstannane (5c) in high yield, while aliphatic olefin 4d
did not undergo the present stannylation reaction at all,
probably reflecting the different LUMO levels of the olefins.
A plausible mechanism is illustrated in Scheme 1A. In-situ-
generated stannyl anion (A) is smoothly photoexcited to the S1
state (1A*) by irradiation with blue LEDs, leading to the T1-
diradical stannyl reagent (3A*) through ISC due to the heavy-
atom effect. Then, SET from the higher-energy SOMO
Scheme 1. Mechanistic Studies
3
electron of A* to a terminal alkyne results in the formation
of the radical anion of the alkyne and stannyl radical species
(3B), and successive radical coupling between the Sn radical
(with the lower-energy SOMO electron) and the terminal
carbon radical affords the vinyl anion intermediate C-1.
Protonation of the intermediate C-1 with MeOH or H2O
provides the hydrostannylated product (D). To examine the
feasibility of this putative mechanism, we performed DFT
calculations (Scheme 1B). In the ground state, the stannylation
of the triple bond with a stannyl anion is kinetically
unfavorable (+34.3 kcal mol−1), and the deprotonation
reaction is favored by 3.7 kcal mol−1. However, judging from
the high activation barriers (over +30 kcal mol−1) and very
large endothermicities, neither of these reactions is expected to
proceed at ambient temperature. In contrast, in the excited state,
the “stannyl diradical” (3A*) generated in situ through
photoexcitation of a stannyl anion and successive ISC
facilitates SET to a terminal alkyne with a large energy gain
a
(B) Potential energies relative to A are shown in parentheses (kcal
mol−1; Gibbs free energies calculated at the (U)B3LYP/LANL2DZ
level for Sn and the 6-31+G* level for others. (C) Yields were
determined by 1H NMR analysis using mesitylene as an internal
standard.
extremely robust character [BDE(Ph−F) = 125.7 kcal
mol−1],42 undergoing cleavage only under harsh conditions.
Given the frequent occurrence of aromatic C−F bonds in
biologically active and functional molecules, developing
methods for the transformation of aromatic C−F bonds is an
important challenge. Aromatic nucleophilic substitution
(SNAr) reactions of aromatic C−F bonds on fluoroarenes
provide direct access to various aromatic molecules, but the
robustness of the C−F bond and the concomitant competing
reactions of ortho-deprotonation or benzyne formation limit
their applicability. To overcome this issue, we envisioned that a
SET-induced defluorostannylation of various fluoroarenes
might be realized by harnessing the powerful one-electron
reducing ability of the present photoexcited stannyl anion.
Two types of defluorostannylations of fluoroarenes utilizing
stannyl anions with/without transition-metal catalysts have
been documented. The nucleophilicity of the stannyl anion
reagents generated in these reactions significantly limits the
scope of the transformations, and efficient reactions require
3
(−22.8 kcal mol−1), affording an intermediary complex B,
which undergoes facile Sn−C bond formation with a
reasonably low activation energy (+6.9 kcal mol−1). On the
other hand, the deprotonation reaction is highly disfavored in
the triplet state due to the intrinsic instability of the alkynyl
radical (3[C−2]). These calculated results are in good
agreement with the experimental findings. Control experiments
also support the above mechanism; that is, the present
stannylation was significantly inhibited by a SET inhibitor
[nitrobenzene (6)], radical scavengers [9,10-dihydroanthra-
cene (7) and (2,2,6,6-tetramethylpiperidin-1-yl)oxy (TEMPO)
(8)], and triplet quenchers [Ni(acac)2 (9) and Cu(acac)2
(10)]41 (Scheme 1C).
Aromatic C−F bonds occupy an ambivalent position,
exhibiting unique physicochemical properties, but with
5631
J. Am. Chem. Soc. 2021, 143, 5629−5635