(S)-5a(1 mol%)
1a
(1 mol%)
(Z)-3a
(Z)-3a
(Z)-3a
no reaction
H2 (1 bar)
1,4-dioxane
80 °C, 1 h
+
4a
2a
(E)-3a
1,4-dioxane
80 °C, 16 h
97% yield 3% yield
(S)-5a (1 mol%)
(E)-3a
+
1,4-dioxane
80 °C, 1 h
Cl
H
3% yield
P
H
Ir
Cl-
Ir
6 (2 mol%)
P
P
(E)-3a
+
4a
Cl
H
1,4-dioxane
80 °C, 1 h
P
>99% yield <1% yield
Ir trihydride species
1M HCl/Et2O
(25 mol%)
5
no reaction
(Z)-3a
1,4-dioxane
80 °C, 1 h
Scheme 1. Semi-hydrogenation with iridium trihydride species.
Scheme 3. Catalytic activity for the isomerization of (Z)-3a.
6 (2 mol%)
H2 (1 bar)
4a
2a
2a
(E)-3a
+
insertion of 2a into an Ir-H bond of 6 gave (E)-alkenyl-iridium
species, whose reductive elimination resulted in the formation of
(Z)-3a, and 6 also worked as a catalyst for the isomerization of
(Z)-3a to (E)-3a. With respect to the isomerization, we tested the
catalytic performance of various iridium precursors, 1a, (S)-5,
and 6 along with HCl, which was expected to be formed in situ
from precursor 1a. Among them, 6 exhibited high catalytic
activity, whereas 1a, (S)-5a, and HCl exhibited almost no
activity for the isomerization (Scheme 3), suggesting that a
dihydride species catalyzed the isomerization, while monohy-
dride species had very low activity.
1,4-dioxane
80 °C, 16 h
51% yield 48% yield
6 (1.0 equiv)
(E)-3a
1,4-dioxane
80 °C, 1 h
82% yield
H
IrH2Cl((S)-binap)(pic)
N
H
P
Ir
P
6
Cl
Based on these experimental results, the plausible mecha-
nism can be divided into two cycles that involve the hydro-
genation of alkyne to (Z)-alkene (cycle I) and isomerization of
the (Z)-isomer to an (E)-isomer (cycle II) (Scheme 4). Under
optimal conditions, the dinuclear iridium complex 1a dissociates
to a monohydride iridium complex A, which reacts with H2 to
give a dihydride iridium complex B with the elimination of HCl.
Dihydride complex B works as a catalyst for both reaction
cycles. In cycle I, alkyne inserts into the Ir-H bond of B
followed by a reductive elimination to afford (Z)-alkene and
mononuclear Ir(I) complex C, whose exposure to hydrogen gas
regenerates B, although we cannot exclude another possible
mechanism of σ-bond metathesis-type direct cleavage of the
Ir-C(alkenyl) bond by H2 to regenerate B,15 as well as direct
semi-hydrogenation of alkyne to give (E)-alkene.9,16 In cycle II,
in situ-generated (Z)-alkene inserts into the Ir-H bond of B,
followed by rotation of the C-C bond and β-hydride elimination
to give the corresponding (E)-alkene and regenerate B as a
consequence of faster β-hydride elimination than reductive
elimination. The Ir(I) precursor (Table S2, Entry 9) and the
mononuclear iridium dihydride species 6 (Scheme 2) produced
more of the over-reduced product 4, and therefore, we propose
the potent scenario that the dinuclear iridium(III) species 1a and
5 were dormant species in equilibrium with B, and consumption
of alkynes shifted the equilibrium to dominant 5 together with
1a because of the different coordination abilities of alkynes
and alkenes to B. In fact, we conducted a semi-hydrogenation
of alkynes using a catalyst system of 6 treated with HCl
(25 mol %), which spontaneously generated the dinuclear
iridium(III) complex 1a, to significantly suppress the formation
of alkane 4a (less than 1%) while maintaining (E)-selectivity.
Scheme 2. Control experiments by iridium dihydride complex 6.
We conducted a time-course experiment of hydrogenating
2a under the optimized conditions, and each yield of (E)-3a, (Z)-
3a, and 4a was plotted over the reaction time. At the beginning,
(Z)-3a was formed, and the amount of (E)-3a gradually
increased with a decrease of (Z)-3a, indicating the isomerization
of (Z)-3a to (E)-3a. During the reaction, the yield of 4a was
suppressed and, after 16 h, only 3% of 4a was detected.
Additionally, under the optimized conditions, (Z)-3a was fully
converted to (E)-3a (95%) and a small amount of 4a (5% yield).
Measurement of the 1H NMR spectrum of the crude reaction
mixture after the hydrogenation of 2a revealed an iridium
trihydride complex 5a.13b Thus, we alternatively prepared the
corresponding iridium trihydride complex of (S)-BINAP [(S)-
5a], and examined its catalytic performance for the hydro-
genation of 2a, giving (E)-3a (97%) and 4a (3%) in almost the
same distribution as that of 1a (Scheme 1). Because complex 5a
was a product of recombination between a mononuclear iridium
dihydride species and a mononuclear iridium monohydride
species, both the dihydride species and monohydride species
were expected to be candidate active species. To investigate the
reaction mechanism, we performed some controlled experiments
using an isolated dihydride complex 6, whose monomeric
structure was supported by the coordination of α-picoline
(Scheme 2).13f We carried out the catalytic semi-hydrogenation
of 2a using 6 to afford (E)-3a in 51% yield along with the
over-reduction product 4a in 48% yield, and performed the
stoichiometric reaction of 6 with 2a under the conditions without
hydrogen gas to produce (E)-3a in 82% yield, indicating that
© 2016 The Chemical Society of Japan | 867