.
Angewandte
Communications
stabilize an adjacent cation than the terminal hydrogen atom
(7; Scheme 1c)
If our rationale for the stability of cationic intermediates is
operative, then it suggests that the regioselectivity of hydro-
stannylation can be reversed by changing the electronics of
the d substituent. To explore this concept, we carried out the
hydrostannylation of the propargylic alcohols 8 and 9
(Scheme 2). Strikingly, as a mechanism involving cationic
Scheme 2. Reversed regioselectivity in the reaction of 8 and 9.
intermediates would predict, the regiochemical outcomes of
the hydrostannylations of 8 and 9 were fully reversed! While
Scheme 1. Molecular oxygen effect on the AIBN-mediated hydrostanny-
lations of 1, 3, and 5.
the electron-donating d-CH group can better stabilize an
3
adjacent cation relative to the a-CH group which has an
2
Information for spectra of crude reaction mixture). To
confirm oxygen as the component of air that is responsible
for promoting this transformation, we again rigorously
excluded all air from one experiment and then carefully
oxygen substituent b to it (see 10), the electron-withdrawing
d-CF group more strongly disfavors adjacent cation forma-
3
tion, thus leading instead to 11. The three-center nature of the
intermediates 10 and 11 originated from our computational
assessment of the reaction surface of this transformation, and
added pure O (10 mol% based on the alkyne) to the reaction
2
[7]
and it proceeded to completion (Scheme 1a). These results
is discussed in detail (see below). On the basis of the tin
[3]
confirm the necessity of O and implies that only a catalytic
radical oxo-coordination hypothesis, which has been used
for decades to rationalize regiochemical outcomes in radical-
mediated hydrostannylation of internal propargylic alcohols,
both reactions in Scheme 2 should have given the same
regiochemical outcome, but they did not. Thus, an electronic
effect involving cationic intermediates can adequately explain
2
amount is needed for hydrostannylation to proceed to
completion. Similarly, low conversion (10%) of the free
propargylic alcohol 3 was observed in the absence of O2,
whereas full conversion into 4 occurred in air (Scheme 1b).
Most shockingly, the AIBN-mediated hydrostannylation of
the terminal propargylic ether 5 did not proceed at all to 6 in
this stark reversed regioselectivity.
[2,6]
119
the absence of O (Scheme 1c).
To see if this effect is
The Sn NMR spectrum of a 1:1 mixture of nBu SnH and
2
3
unique to AIBN in alkyne hydrostannylation, the radical
mediators p-methoxybenzene thiol and tert-butylbenzoper-
oxoate were examined and also found to be similarly O2
dependent (see the Supporting Information). Further, if the
radical initiator is omitted, hydrostannylation does not
proceed, hence the generation of a tin radical by the
mediators is crucial to the overall addition process, regardless
of the role of oxygen.
Following conventional pure radical mechanisms in the
literature, few would argue against suggestions that the
radical intermediate arising from tin radical addition to
a ground-state terminal alkyne would favor terminal place-
ment of the tin moiety for both steric and electronic reasons
AIBN in degassed [D ]benzene heated at 808C showed the
6
formation of Bu SnSnBu (d = ꢀ79 ppm) whose formation
3
3
can only be explained by the presence of nBu SnC (see the
3
[
8]
Supporting Information). Additionally, AIBN-mediated
deiodination of alkyl iodide with nBu SnH, ran under
3
identical reaction conditions as the hydrostannylation experi-
ments in this study, proceeded smoothly in the absence of O2
(see the Supporting Information). Also of note, the hydro-
stannylation of all the above-mentioned alkynes only pro-
ceeded in the presence of O in the dark, thus, excluding any
2
effect of light. Finally, isomerization of the kinetically formed
Z isomer by nBu SnC to the corresponding thermodynamic
3
E isomer (e.g., (Z)-2 to (E)-2) occurs in the absence of O (see
2
(
i.e., generating a secondary radical as opposed to a primary
the Supporting Information). Taken together, this suggests
that oxygenꢀs involvement relates to the alkynes, or their
resultant vinyl radical intermediates, and not anything related
to the radical initiators.
[
2,6]
one).
However, since the addition only proceeded in the
presence of O , and was faster in polar solvents, perhaps the
2
rational is in fact not based on steric or radical stability, but on
cation stability. In this case, the cation will primarily localize
on the b-carbon atom because the a-methylene can better
We reasoned that oxygenꢀs role, which is central in all of
[
9]
this, is most likely as a redox shuttle which can be involved
2
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 1 – 6
These are not the final page numbers!