Organic Letters
Letter
It is noteworthy that all products 3 were obtained in
exclusive E-selectivity of the trisubstituted double bond. In
addition, aryl halides are untouched in the reaction because
phosphines are inherently selective toward alkyl halides. To
explore the synthetic utility of the methodology, sequential
Heck-type cross-coupling reactions starting from 4-bromoben-
zyl bromide 1i were demonstrated (Scheme 2). The
1.0 equiv of D2O was introduced into the mixture of 1a and 2a
under the standard conditions (Scheme 4a), product 3a-d3 was
obtained in 80% yield with deuterium incorporated at both the
olefinic C−H hydrogen (38% D) and −CH2− protons
(15% D). For comparison, non-deuterium product 3a was
treated with D2O (1.0 equiv) under the same condition, which
afforded 3a-d2 in 85% yield with deuterium only at the −CH2−
unit (41% D) (Scheme 4b). These results suggest that the
olefinic C−H hydrogen was deuterated during the catalytic
cycle, while the −CH2− protons may be deuterated after the
reaction because of their acidity. Accordingly, it implies that
water is involved in the mechanism of cross-coupling.
Scheme 2. Sequential Heck-Type Cross-Coupling
To further elucidate the mechanism, DFT calculations for
the phosphine-catalyzed cross-coupling of benzyl bromide 1a
and fumarate 2a were performed (Figure 1). To simplify the
phosphine-catalyzed cross-coupling of 1i with fumarate 2a
provided 3i in 70% yield, which was followed by a Pd-catalyzed
Heck cross-coupling with alkenes 4, leading to the formation of
products 5a−c in 58−74% yields with excellent E,E-selectivity.
1
To clarify the mechanism, 31P and H NMR monitoring of
the reaction of benzyl bromide 1a, fumarate 2a, PBu3, and
Cs2CO3 was conducted (Scheme 3a), which suggested the
Figure 1. Gibbs free energy profile for the PMe3-catalyzed cross-
coupling of 1a and 2a. Relative energies in kilocalories per mole.
Scheme 3. Control Experiments
computation, trimethylphosphine is used instead of tribu-
tylphosphine. We computed that the formation of phospho-
nium 6a and phosphorus ylide A was exothermic with the
energies of −19.3 and −41.3 kcal mol−1, respectively. The
attack of ylide A on fumarate 2a to form adduct B was
calculated to bear an activation barrier of 18.7 kcal mol−1 (Ts-
1). The following hydrogen transfer to approach intermediate
C may proceed in a 1,2-H shift process (path a) or a tandem
1,3-H/1,2-H shift via ylide C′ (path b). However, both of
these processes show high activation barriers, that is, 43.5 kcal
mol−1 for the 1,2-H shift of path a (Ts-2) and 52.3 kcal mol−1
for the 1,2-H shift of path b (Ts-4). On the basis of the
deuteration experiments described above, we conceived that
the hydrogen transfer may be assisted by a trace amount of
water in the solvent.20 Therefore, for the water-assisted 1,2-H
shift of path a (Ts-6, colored pink), the activation barrier
decreases to 29.9 kcal mol−1; for the water-assisted 1,2-shift of
path b (Ts-7, colored blue), the activation energy is only 8.6
kcal mol−1. Thus, the water-assisted stepwise 1,3-H/1,2-H shift
via transition states TS-3 and TS-7 may account for the
hydrogen transfer step. This also corroborates the results of the
deuterium experiments in which the olefinic C−H hydrogen
was deuterated by D2O (Scheme 4a). Finally, it was shown
that the elimination of phosphine to form product 3a is fast
with an activation barrier of 2.0 kcal mol−1. The formation of
the E-product is computed to be both kinetically and
The overall reaction of the cross-coupling is highly exothermic
that releases 67.7 kcal mol−1 of energy.
formation of a phosphonium 6a (31P NMR δ 31.61) in a
tion). On the contrary, when phosphonium salt 6a was treated
with fumarate 2a in the presence of Cs2CO3, desired product
3a was obtained in 70% yield19 (Scheme 3b). The results
suggest that the phosphine attacks the benzyl bromide to form
phosphonium 6a to initiate the reaction.
To inspect the proton transfer step of the mechanism,
deuteration experiments were conducted (Scheme 4). When
Scheme 4. Deuteration Experiments
In summary, we have developed a phosphine-catalyzed
olefinic cross-coupling between benzyl bromides and fuma-
4572
Org. Lett. 2021, 23, 4570−4574