Organic Letters
Letter
alkenes (Scheme 1b, right).7,12 Although these issues have
been overcome through rational design of ligands by Sigman’s
group and others,6,7,13 it still remains unclear and challenging
under CAPT conditions, where the regioselectivity of the β-
hydride elimination as well as the migratory insertion mainly
relies on the chiral counteranion. Further, since the unactivated
acyclic internal alkenes generally exhibit lower reactivity in
Heck-type reactions compared to terminal and cyclic alkenes, it
is still questionable if sufficient reactivity could be achieved in
the CAPT catalysis scenario for these substrates. Despite these
challenges, to extend the application scenarios of CAPT,
herein, we report our efforts on the investigation of the chiral-
anion-mediated asymmetric Heck−Matsuda reaction of acyclic
alkenyl alcohols (Scheme 1c).
a
Table 1. Reaction Optimization
b
c
d
entry
8
additives solvent yield (3a, %)
er
3a/4a
1
2
3
4
5
6
7
8
9
10
11
12
8a
8a
8a
8a
8a
8a
8a
8a
8a
8a
8a
8a
8b
8c
none
none
none
none
6a
Et2O
MTBE
THF
DCM
Et2O
Et2O
Et2O
Et2O
Et2O
Et2O
Et2O
Et2O
Et2O
Et2O
41
45
44
trace
60
57
61
68
82
75
57
47
75
95:4
85.5:14,5
87.5:12.5
-
2.5:1
4:1
2.3:1
-
The hypothesized chiral-anion-mediated asymmetric Heck−
Matsuda reaction is outlined in Scheme 2 with phenyl
Scheme 2. Chiral Anion-Mediated Asymmetric Heck-
Matsuda Reaction of Acyclic Alkenyl Alcohols
94:6
2.6:1
2.8:1
3:1
>20:1
>20:1
>20:1
>20:1
5:1
6b
6c
93.5:6.5
93.5:6.5
91:9
94.5:5.5
95:5
6a, 7a
6a, 7a
6a, 7b
6a, 7c
6a, 7d
6a, 7a
6a, 7a
e
e
e
95:5
95:5
e
e
13
14
87.5:12.5
95:5
>20:1
>20:1
e
f
82(75 )
a
Unless noted otherwise, the reaction of 1a (0.1 mmol) with 2a (0.2
mmol) was carried out with Pd2dba3 (0.005 mmol), 8 (0.013 mmol),
Na2CO3 (0.2 mmol), additive 6 (0 or 0.015 mmol), and additive 7 (0
or 0.016 mmol) in 2.0 mL of solvent at 25 °C for 12 h; a trace amount
b
c
of 5a was observed. NMR yield. Determined by HPLC analysis.
d
Determined by 1H NMR analysis of the crude reaction mixture.
The reaction was carried out at 0 °C. Isolation yield.
e
f
dazonium salt 1a and allylic alcohol 2a as the typical substrates
and chiral phosphoric acid 8a as the source of the chiral anion.
In this strategy, insoluble diazonium salt is transferred into the
solution phase to give ion-pair I9a,g−j,10 via anion exchange
with sodium phosphate generated from 8a and sodium
carbonate. The ion pair then undergoes oxidative addition by
the Pd(0) species to give Pd(II) intermediate II3,9b with the
chiral phosphate as the counteranion, which generates
intermediate III after migratory insertion of alkene 2a. In
this stage, two possible β-H elimination processes are possible
to form either byproduct 5a or intermediate IV. It requires the
chiral anion to govern the selectivity for a preferred β-Hb
elimination process. Olefin dissociation and tautomerization of
the enol afford the desired chiral ketone 3a and Pd hydride
species V, which then undergoes formal reductive elimination
and regenerates Pd(0) and sodium phosphate as suggested by
Toste and co-workers.9h On the other hand, the alkene
isomerization issue has been previously observed in Heck−
Matsuda reactions.14,9h In this case, terminal alkene 9 can be
formed and undergoes another Heck reaction to produce
byproduct 4a.
observed as the main byproduct, and only a trace amount of 5a
was detected. Employing methyl tert-butyl ether as the solvent
to furnish the reaction provided the product with a slightly
higher chemoselectivity, moderate yield, and lower enantiose-
lectivity (entry 2). Switching the solvent to THF led to no
better results, and the reaction turned completely unproductive
when using DCM as reaction media (entries 3 and 4). Inspired
by previous observations by Fairlamb,15 Trost,9a and Sigman13j
that dibenzylideneacetone (dba) derivatives can possibly
stabilize palladium species along the catalytic cycle and thus
affect the catalytic performance, several dibenzylideneacetone
(dba) derivatives (6a−6c) were tested as additives for the
reaction in ether (entries 5−7). It was found that in all these
reactions a significant boost in the yield was observed but with
lower enantioselectivity, and the generation of byproduct 4a
could not be surpressed (vs entry 1). To further optimize the
reaction and inhibite the side reaction, we noticed that
dimethyl sulfoxide (DMSO) was frequently used as the solvent
or ligand for the palladium-catalyzed reaction.16 Beller and co-
workers found that DMSO as a cosolvent could tune the
selectivity of the Heck arylation of cyclohexene with aryl
bromides.16g It was found that the addition of 15 mol %
DMSO (7a) in the reaction sufficiently suppressed the
generation of 4a and led to a slight increase in the yield,
albeit that slightly lower enantioselectivity was obtained (entry
8 vs entry 5). Lowering the temperature to 0 °C rendered the
reaction with 82% yield, excellent chemoselectivity, and a
94.5:5.5 er (entry 9). At this stage, due to the significant
To verify the feasibility of the hypothesized reaction,
phenyldiazoniumtetrafluoroborate 1a and (E)-pent-3-en-2-ol
2a were subjected to 5 mol % of Pd2dba3 and 13 mol % of
chiral phosphoric acid 8a with Na2CO3 in ether (Table 1, entry
1). Encouragingly, the reaction did afford chiral ketone 3a with
moderate yield and excellent enantioselectivity. At the same
time, a notable amount of compound 4a was surprisingly
1474
Org. Lett. 2021, 23, 1473−1477