Angewandte
Chemie
halides (bromides or chlorides) with diphenylzinc (1) in
[D6]benzene led to diaryl methanes 3a–i incorporating halide
(Cl, Br, or I), ether, nitro, vinyl, and trifluoromethyl
substituents in good to excellent yields (70–99%). Remark-
ably, in the majority of cases short reaction times were
sufficient (< 10 min) for complete conversion at ambient
temperature, with reaction completion indicated by the
precipitation of PhZnX (X = Br or Cl). After this time,
simple filtration of the crude reaction mixture typically
afforded the pure diaryl methane (as judged by GC–MS and
NMR spectroscopy) without the necessity for chromatogra-
phy or any further purification. It was also possible to scale up
the synthesis of 3a to 5.0 mmol without any significant
erosion of the yield of the isolated product (93%). In contrast
to the addition of even a stoichiometric amount of an aliphatic
ether, no significant reactivity suppression was observed with
aryl ethers, as indicated by the comparable rates for the
formation of 3a and 3b.
to form diaryl zinc reagents in THF from Grignard reagents,
and in our hands this approach resulted in the isolation of
ionic zincates. For example, attempts to form di(p-tolyl)zinc
from p-TolMgBr (1m in THF) and ZnBr2 in THF led instead
to the isolation of the zincate species 8 (identified by X-ray
crystallography, Scheme 2). In [D6]benzene, 8 exists as
A range of secondary alkyl halide (bromide or chloride)
substrates could also be effectively coupled with 1 at ambient
temperature to generate compounds 5a–i incorporating ester,
vinyl, and tert-butyloxycarbonyl-protected amine functional-
ities in 86–97% yield. Product 5j was also isolated in excellent
yield (96%, < 10 min at ambient temperature), thus demon-
strating the utility of this method for the generation of triaryl
methanes, which are ubiquitous motifs in medicinal/biological
chemistry.[33] The use of a secondary allylic electrophile, 3-
chlorobut-1-ene, and a tertiary propargylic electrophile, 3-
chloro-3-methyl-1-butyne, both led to SN2’ products, with the
formation of 5k and 6a, respectively. Although the latter two
classes of electrophile have been previously coupled with aryl
zinc reagents in THF, it is notable that under these etherate-
free conditions cross-coupling with 1 requires only a stoichio-
metric amount of the electrophile and is complete within
minutes versus 16 h with zincates in THF.[25] Tertiary alkyl
halides were also readily coupled with 1 at ambient temper-
ature in exceedingly short reaction times (< 5 min) to afford
compounds 6b–d bearing new quaternary centers in good to
excellent yields (72–91%). These results are remarkable, as
Scheme 2. Synthesis of zincate species 8 and its reactivity with 2
under the standard cross-coupling conditions. An ORTEP representa-
tion of part of the extended solid-state structure of 8 is shown with
ellipsoids at 50% probability. Hydrogen atoms and solvent molecules
(benzene) have been omitted for clarity; THF molecules have also
been simplified to oxygen atoms for clarity.
a single species displaying one set of THF and p-tolyl
1
resonances in the H NMR spectrum, which is comparable
to that previously reported for [Zn(p-tolyl)3]À[(Mg2(m-Cl)3-
(THF)6]+ by Hevia et al.[35] Although there is growing
evidence that zincates are active (and possibly crucial)[36–38]
nucleophiles in transition-metal-catalyzed cross-coupling in
more polar solvents (including THF), the combination of 3-
methoxybenzyl bromide (2; 2 equiv, 1 equiv per Zn atom) and
8 in [D6]benzene led to no cross-coupling even after extended
periods (20 h). This outcome is significant given that the
generation of diaryl zinc reagents from Grignard reagents
in situ in THF is a standard process.
2
3
À
C(sp ) C(sp ) cross-coupling with tertiary alkyl halides is
extremely challenging even using state-of-the-art transition-
metal-catalyzed procedures.[34] Finally, we examined the
propensity of primary alkyl halide substrates to undergo
heterocoupling transformations with 1. 1-Bromohexane could
be coupled with 1 to afford 7a in 74% yield, but this
transformation required heating at 808C for 20 h. After 20 h
the complete consumption of 1-bromohexane was observed
with the remaining mass balance (26%) made up by internal
isomerization cross-coupled products.[31] These reactions were
all performed in [D6]benzene. Clearly, the use of less toxic
solvents is highly desirable, and we re-emphasize that
chlorobenzene is also a viable solvent for this process (for
example, 3a, 5a, and 6c were all isolated in high yields when
chlorobenzene was used as the solvent with the reaction
carried out on a 1 mmol scale and at the higher concentration
of 1m).[31]
The lack of cross-coupling using zincate 8 led us to target
bona-fide examples of etherate-free diarylzinc reagents,
a family of compounds that are relatively scarce within the
literature.[39,40] Dimesitylzinc had been generated previously
in an etherate-free form from mesitylmagnesium bromide and
ZnCl2 in THF.[41] Pleasingly, the combination of dimesitylzinc
with 2 in [D6]benzene led to rapid cross-coupling to form 9a in
high yield (82%), with the rate of reaction not appreciably
diminished for this sterically hindered diaryl zinc reagent
(Scheme 1). Other etherate-free diaryl zinc reagents could be
generated by lithium/halogen exchange (with nBuLi) and
subsequent zincation with ZnCl2 in diethyl ether; all diethyl
ether was subsequently removed in vacuo (5 ꢀ 10À2 mbar,
ambient temperature, 2–5 h). Etherate-free diaryl zinc
reagents incorporating aromatic groups with electron-with-
We next sought to extend the range of nucleophilic diaryl
zinc reagents for this reaction by using 3-methoxybenzyl
bromide (2) as the model electrophile. We initially attempted
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3
These are not the final page numbers!