Potassium Base-Catalyzed Michael Additions of Allylic Alcohols to α,β-Unsaturated Amides: Scope and Mechanistic Insights

Article abstract of DOI:10.1002/adsc.202100272

We report herein the first KHMDS-catalyzed Michael additions of allylic alcohols to α,β-unsaturated amides through allylic isomerization. The reaction proceeds smoothly in the presence of only 5 mol% of KHMDS to afford a variety of 1,5-ketoamides in high yields. Mechanistic investigations, including experimental and computational studies, reveal that the KHMDS-catalyzed in-situ generation of the enolate from the allylic alcohol through a tunneling-assisted 1,2-hydride shift is the key to the success of this transformation. (Figure presented.).

Full text of DOI:10.1002/adsc.202100272

Title: Potassium Base-Catalyzed Michael Additions of Allylic Alcohols
to α,β-Unsaturated Amides: Scope and Mechanistic Insights
Authors: Masahiro Sai and Hiroaki Kurouchi
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Adv. Synth. Catal. 10.1002/adsc.202100272
Link to VoR: https://doi.org/10.1002/adsc.202100272

10.1002/adsc.202100272
Advanced Synthesis & Catalysis
FULL PAPER
DOI: 10.1002/adsc.202((will be filled in by the editorial staff))
Potassium Base-Catalyzed Michael Additions of Allylic Alcohols
to ,-Unsaturated Amides: Scope and Mechanistic Insights
Masahiro Saia,b,* and Hiroaki Kurouchib
a
Department of Chemistry and Biomolecular Science, Faculty of Engineering, Gifu University, 1-1 Yanagido, Gifu 501-
1193, Japan
Phone: (+81)-58-293-2796; fax: (+81)-58-293-2564; e-mail: msai@gifu-u.ac.jp
Research Foundation ITSUU Laboratory, C1232 Kanagawa Science Park R & D Building, 3-2-1 Sakado, Takatsu-ku,
b
Kawasaki, Kanagawa 213-0012, Japan
Received: ((will be filled in by the editorial staff))
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201######.
unsaturated alcohols[8] prompted us to develop a
Abstract. We report herein the first KHMDS-catalyzed
Michael additions of allylic alcohols to ,-unsaturated base-catalyzed
tandem
allylic
amides through allylic isomerization. The reaction proceeds
smoothly in the presence of only 5 mol% of KHMDS to
afford a variety of 1,5-ketoamides in high yields.
Mechanistic investigations, including experimental and
computational studies, reveal that the KHMDS-catalyzed
in-situ generation of the enolate from the allylic alcohol
through a tunneling-assisted 1,2-hydride shift is the key to
the success of this transformation.
isomerization/electrophilic trapping reaction. As
outlined in Scheme 1, our working hypothesis for a
base-catalyzed tandem reaction relies on the use of
,-unsaturated carbonyl compounds as electrophiles.
Enolate B, which is generated from alkoxide A
through allylic isomerization, undergoes Michael
addition to the ,-unsaturated carbonyl compound
to form enolate C, which is sufficiently basic to
smoothly abstract the OH proton of the next substrate
1, to give the desired Michael adduct and regenerate
A. Thus, a catalytic amount of a base (M+B–) drives
the catalytic cycle even though the base itself is not
regenerated.[9] In this paper, we describe the first
base-catalyzed tandem allylic isomerization/Michael
Keywords: allylic alcohols; enolates; Michael addition;
hydride shift; tunneling
addition
reaction
under
transition-metal-free
Introduction
conditions. The mechanism of the allylic
isomerization process is also discussed based on
experimental and computational studies.
Allylic alcohols are versatile synthetic precursors that
can participate in a variety of transformations.[1]
Among them, isomerization is particularly important
because it provides carbonyl compounds in an atom-
economical manner. Therefore, a variety of catalytic
systems that enable such transformations have been
developed, but most use transition-metal catalysts.[2]
Although the development of transition-metal-free
base-mediated protocols is highly desirable from the
viewpoint of green chemistry, they have only been
O–M+
O–M+
cat. base (M+B–)
deprotonation
R2
OH
allylic
isomerization
R1
R1
R2
R1
R2
alkoxide A
enolate B
allylic alcohol 1
protonation
O
O
O
O
O–M+
R3
R3
Michael addition
R1
R3
R1
R2
R2
enolate C
studied sporadically, and most require
a
product
stoichiometric or excess amount of a base.[3] In 2015,
Kang et al. first reported the NaOtBu/phenanthroline-
catalyzed allylic isomerization under transition-
metal-free conditions;[4a] they also conducted a
detailed mechanistic study and revealed that the
reaction proceeds through a radical pathway. The
Scheme 1. Working hypothesis for a base-catalyzed
tandem allylic-isomerization/Michael-addition sequence.
base-mediated allylic isomerization is also applicable Results and Discussion
to tandem reactions; i.e., allylic isomerization
followed by electrophilic trapping of the in-situ
generated enolate.[5,6] However, this type of reaction
still requires stoichiometric amounts of base, and a
catalytic variant remains unexplored.[7] Our
continuing interest in the transformations of
Initially, we aimed to identify the appropriate base for
the isomerization of an allylic alcohol to the
corresponding enolate or homoenolate. To this end,
allylic alcohol 1a was reacted with various bases for
1 h and then trapped with benzyl bromide. The
1
This article is protected by copyright. All rights reserved.

10.1002/adsc.202100272
Advanced Synthesis & Catalysis
1
product distribution was monitored by H NMR
spectroscopy (Table 1). The use of lithium bases led
to no observable reaction (entries 1 and 2). When
sodium bases were used, large amounts of 1a were
again recovered, while O-benzylated ether 2 and C2-
benzylated ketone 3 were obtained in low yields
(entries 3 and 4). In contrast, we found that 3 was
formed in high yields with almost complete
consumption of 1a when potassium bases were
employed (entries 5 and 6).[10] In particular, KHMDS
was effective for the allylic isomerization of 1a to the
corresponding enolate and its subsequent reaction
with benzyl bromide. No C3-benzylated ketone 4 was
observed in any reaction.
the alkoxide of 1a and provides an opportunity for
allylic isomerization. It should be noted that this
represents the first example of the base-catalyzed C-
addition of an allylic alcohol to an electrophile other
than a proton.
O
Ph
O
KHMDS (10 mol%)
Ph
Ph
THF, r.t., 15.5 h
O
5, 75%
StBu
Ph
O
KHMDS (10 mol%)
Ph
O
O
StBu
THF, r.t., 15.5 h
O
6, 47%
OH
O
Ph
Ph
Table 1. Benzylation of allylic alcohol 1a in the presence
Ph
OMe
1a (1.2 equiv.)
of various bases.[a]
OMe
Bn
KHMDS (10 mol%)
7, 25%
Ph
O
THF, r.t., 12 h
Bn
O
Ph
O
OMe
O
8, 14%
Ph
Ph
O
O
NMe2
9a
2
base (1.0 equiv.)
THF, 0 ˚C to r.t., 1 h
OH
O-benzylation
KHMDS (10 mol%)
THF, r.t., 13.5 h
Ph
NMe2
O
then BnBr (1.2 equiv.)
r.t., 1 h
Bn
Ph
Ph
O
Bn
1a
2
10aa, 92%
Ph
Ph
Ph
3
Ph
Bn
3
4
Scheme 2. The KHMDS-catalyzed tandem allylic-
C2-benzylation
C3-benzylation
isomerization/Michael-addition reaction.
Yield [%][b]
Entry
Base
1a
2
3
We next investigated the allylic alcohol substrate
scope (Table 2). Optimization studies revealed that
the reaction of 1a with 9a was complete in 1.5 h in
the presence of only 5 mol% of KHMDS.[12] A wide
range of allylic alcohols 1 underwent successive
allylic isomerization and Michael addition to 9a, to
provide the corresponding 1,5-ketoamides 10 in
good-to-excellent yields. This reaction tolerates a
variety of functionalities, such as ether, halo,
trifluoromethyl, cyano, furyl, thienyl, and pyridyl
groups.[13] Although the reaction of 1q devoid of
substituent at the 3-position was sluggish, giving
10qa in only 20% yield,[14] the yield improved to
88% when the aryl group was changed from phenyl
to 2-pyridyl, as in 10ra. Methyl-substituted allylic
1
2
3
4
5
nBuLi
LHMDS
NaH[c]
NaHMDS
KH[c]
KHMDS
>99
94
49
77
0
0
0
17
15
3
0
0
17
4
71
94
6
[a]
1
3
Conditions: 1a (0.2 mmol) and base (0.2 mmol) in THF
(2 mL) at room temperature for 1 h followed by trapping
[b]
with benzyl bromide (0.24 mmol) for 1 h.
Determined
by 1H NMR analysis. [c] With 1.5 equiv. of base.
We next explored the feasibility of the working
hypothesis outlined in Scheme 1, using various ,-
unsaturated carbonyl compounds (Scheme 2). When
an ,-unsaturated ketone was selected as the
electrophile, no Michael adduct was observed, and
only ketone 5 derived from 1a was obtained in 75%
yield. Ester 6 was formed in 47% yield when the ,-
unsaturated thioester was used; 6 was formed by the
1,2-addition of the alkoxide of 1a to the thioester
followed by the thia-Michael addition of KStBu to the
in-situ-generated ,-unsaturated ester. While the
reaction of 1a with the ,-unsaturated ester provided
a 25% yield of the desired Michael adduct 7 for the
first time, inseparable oxa-Michael adduct 8 was also
generated. Gratifyingly, the use of ,-unsaturated
amide 9a delivered Michael adduct 10aa in 92%
isolated yield without the formation of any side
products. The high yield and chemoselective
formation of 10aa is likely due to the lower reactivity
of 9a,[11] which retards the oxa-Michael addition of
alcohol 1s was also
a
suitable substrate.
Unfortunately, allylic alcohols bearing an alkyl group
at the C1 position could not be employed in this
reaction because such substrates did not undergo the
allylic isomerization.[15]
Table 2. Substrate scope of allylic alcohols 1.[a]
2
This article is protected by copyright. All rights reserved.

10.1002/adsc.202100272
Advanced Synthesis & Catalysis
O
O
[a]
For general reaction conditions, see Table 2. Yields of
O
OH
KHMDS (5 mol%)
THF, r.t., 1.5 h
R1
NMe2
isolated products are shown. [b] With 7.5 mol% of KHMDS.
R1
R2
NMe2
[c] With 2.0 equiv. of 1a and 20 mol% of KHMDS.
R2
10
1 (1.2 equiv.)
9a
O
O
O
O
O
O
Ph
NMe2
Ph
NMe2
The key step in this transformation is the
isomerization of the allylic alcohol to the
corresponding enolate. Therefore, to elucidate the
isomerization mechanism, we performed a series of
control experiments (Scheme 3). First, 1a was reacted
Ph
NMe2
Ph
10aa, 92%
Me
10ba, 82%
10ca, 93%
O
O
O
O
O
O
Ph
NMe2
O
Ph
NMe2
Ph
NMe2
OMe
with
9a
in
the
presence
of
(2,2,6,6-
to
tetramethylpiperidin-1-yl)oxyl
(TEMPO)
X
O
investigate if the mechanism is radical-based as
proposed by Kang and co-workers.[4a] The reaction
afforded 10aa in 80% yield even in the presence of
TEMPO. In addition, radical-clock substrate 1t was
reacted with 9a, which afforded 10ta in 67% yield
without any formation of ring-opening byproducts.
These results provide strong evidence against a
radical mechanism. Because a direct 1,2-proton shift
is symmetry-forbidden and unlikely to occur,[16,17] we
believe that the allylic isomerization proceeds via a
hydride-shift mechanism. We performed a deuterium-
labeling experiment to distinguish between 1,2- and
1,3-hydride shifts.[3e] When 1a-d was reacted with 10
mol% of KHMDS, 31% deuterium was incorporated
at the C2 position in 5-d2, and only 4% was
incorporated at the C3 position. Because the erosion
of deuterium content at the C2 position can be
explained by base-catalyzed D/H exchange,[18,19] this
result indicates that a 1,2-hydride shift is likely to be
operative in the allylic-isomerization process. Next,
we reacted 1a-d with 9a in the presence of 5 mol%
KHMDS, which furnished a 5.3:1 mixture of 10aa-d
(26% D) and 11aa-d (99% D). The detection of oxa-
Michael adduct 11aa-d was in sharp contrast to the
outcome observed with 1a and suggests that the
incorporation of deuterium significantly retards the
1,2-hydride-shift process. An alternative pathway that
involves direct C1–H deprotonation is conceivable
because the difference between the pKa values of the
OH and the C1–H in 1a may not be significant in
organic solvents. To test this hypothesis, 1a-d was
reacted with a stoichiometric amount of KH. If KH
directly abstracts the C1–D in 1a-d, the reaction
should not afford deuterated ketone 5-d because the
deuterium at the C1 position is lost from the reaction
system as HD gas. Surprisingly, 76% of the
deuterium was transferred to the C3 position of
ketone 5-d3 although no deuterium incorporation at
the C2 position was observed. A similar reaction
using KHMDS as the base produced 5-d3 (82% D at
C3). The fact that deuterium content was well
preserved during the reaction ruled out a direct C1–H
deprotonation pathway. An explanation for the
change in position of the deuteration that is
dependent on the amount of base is presented as
follows (Scheme 4): For cases with a catalytic
amount of a base, homoenolate B is generated from
alkoxide A through a 1,2-deuteride shift undergoing
rapid C3-protonation by the OH group of 1a-d, thus
producing C2-deuterated ketone 5-d2, which is
10ea (X = O), 90%
10fa (X = S), 91%
10da, 89%
OMe
10ga, 92%
O
O
O
O
10ha (X = OMe), 79%
10ia (X = F), 92%
10ja (X = Cl), 93%
10ka (X = Br), 53%
10la (X = CF3), 93%
10ma (X = CN), 96%
Ph
NMe2
Ph
NMe2
X
N
10na, 95%
O
O
O
O
O
O
X
N
NMe2
NMe2
NMe2
X
Me
Et
Ph
10oa (X = O), 90%
10pa (X = S), 90%
10qa (X = CH), 20%
10ra (X = N), 88%
10sa, 55%[b]
[a] Conditions: 1 (0.36 mmol), 9a (0.3 mmol), and KHMDS
(5 mol%, 0.5 M in toluene) in THF (3 mL) at room
temperature for 1.5 h. Yields of isolated products are
shown. [b] With 20 mol% of KHMDS for 2 h.
We also examined the generality of reactions
involving ,-unsaturated amides 9 (Table 3), and
various substituents on the nitrogen atom were
evaluated. N,N-Dialkylamides reacted well to afford
the desired products 10ab and 10ac. N,N-diallyl-,
N,N-dibenzyl-, and N,N-di(p-methoxybenzyl)amides,
with possible subsequent deprotection in mind, were
well tolerated and gave the corresponding products
10ad–10af in high yields. Weinreb-type amide 9g
and N-arylamide 9h were also suitable substrates.
,-Unsaturated amides 9i and 9j bearing alkyl or
aryl substituents at their -positions could also be
employed in this reaction.
Table 3. Substrate scope of ,-unsaturated amides 9.[a]
O
R1
O
OH
O
KHMDS (5 mol%)
NR2R3
THF, r.t., 1.5 h
Ph
O
NR2R3
Ph
Ph
R1
Bn
1a (1.2 equiv.)
9
10
O
O
O
O
O
Ph
N
Ph
Ph
N
Ph
N
Bn
Bn
10ac, 83%
O
Bn
10ab, 91%
10ad, 84%
O
O
O
O
O
O
OMe
Ph
Ph
NBn2
N(PMB)2 Ph
N
Bn
Bn
Bn
10ag, 68%
Ph O
Me
10ae, 73%
10af, 80%
Me
O
O
O
O
O
Ph
N
Ph
NPh2
Ph
NPh2
Bn
Me
Bn
Bn
10ah, 88%
10ai, 62%[b]
10aj, 57%[c]
dr = 1.5:1
dr = 1.6:1
3
This article is protected by copyright. All rights reserved.

10.1002/adsc.202100272
Advanced Synthesis & Catalysis
with a catalytic amount of a base
with a stoichiometric amount of a base
susceptible to base-catalyzed C2–D/H exchange.
However, when a stoichiometric amount of base is
used, homoenolate B abstracts the C1–D of alkoxide
A affording dianion C because there is no OH group
of 1a-d. Dianion C also abstracts the C1–D of
alkoxide A to give enolate D and regenerate C.
Enolate D is converted into C3-deuterated ketone 5-
d3 after aqueous work-up.[20,21]
K+
O
K+
O
2
3
1,2-deuteride Ph
Ph
2
shift
slow
1,2-deuteride
shift
3
D
Ph
Ph
homoenolate B
fast
D
O–K+
1
slow
O
D
3
homoenolate B
C3-deuteration
Ph
2
Ph
Ph
Ph
Ph
OH
D
C3-protonation
alkoxide A
D
O–K+
Ph
Ph
D/H
exchange
K+
3
D
1a-d
Ph
Ph
O
D
3
dianion C
O
H
O–K+
D
3
2
fast
Ph
Ph
Ph
C3-deuteration
(1) Investigation of a radical mechanism
D/H
D
5-d2
Ph
Ph
Ph
KHMDS (5 mol%)
TEMPO (20 mol%)
enolate D
O
O
O
OH
D/H
exchange
H+ (work-up)
D
Ph
NMe2
O
THF, r.t., 1.5 h
NMe2
Ph
Ph
Bn
O
O
1a (1.2 equiv.)
9a
10aa, 80%
2
Ph
Ph
3
Ph
O
5-d3
OH
D/H
N
O
KHMDS (20 mol%)
N
NMe2
THF, 65 ºC, 2 h
NMe2
Scheme 4. A plausible mechanism of the base-catalyzed
isomerization of allylic alcohol 1a-d.
9a
1t (1.2 equiv.)
10ta, 67%
(2) Investigation of 1,2- and 1,3-hydride shifts
OH
K+
3
O
O
H
O–K+
Ph
Ph
D
We turned to kinetics studies to acquire more
information on this allylic-isomerization process. We
first reacted 1a with 5 mol% of KHMDS to measure
rate constant kH1 (Scheme 5). This reaction is
catalytic; hence the concentration of alkoxide A is
essentially constant and the same as that of the added
KHMDS. However, we found that the reaction
clearly slowed as the reaction proceeded, which is
probably due to the equilibrium between A, ketone 5,
1a, and the corresponding enolate, which decreases
the concentration of A. Accordingly, we performed
initial-rate kinetics studies to measure kH1. On the
basis of these experiments,[22] kH1 was determined to
be 3.33 × 10–3 s–1. We also conducted similar kinetics
experiments using 1a-d bearing 99% deuterium at the
1-position and consequently observed a large KIE of
7.6 (kD1 = 4.39 × 10–4 s–1). These results indicate that
the 1,2-hydride shift (C–H bond cleavage step) is
rate-determining and that tunneling[23] is involved in
this event.
2
2
2
1
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
1
3
1,2-hydride
shift
C3-protonation
Ph
D
D
5-d2
–A
D
alkoxide A
OH
O
D
3
O–K+
1
Ph
Ph
O
1
D
3
D
3
Ph
Ph
Ph
2
1,3-hydride
shift
C2-protonation
Ph
Ph
2
D
5-d3
–A
H
K+
alkoxide A
O
1
D (4%)
OH
KHMDS (10 mol%)
THF, r.t., 3 h
2
3
Ph
Ph
D (99%)
D (31%)
5-d2, 38%
1a-d
(3) The reaction of 1a-d with 9a
O
O
Ph
NMe2
9a
OH
KHMDS (5 mol%) (26%) D Bn
Ph
10aa-d
O
Ph
THF, r.t., 1.5 h
Ph D (99%)
D (99%)
1a-d
(1.2 equiv.)
Ph
O
NMe2
11aa-d
72% (10aa-d/11aa-d = 5.3:1)
(4) Investigation of a direct C1-H deprotonation mechanism
OH
OH/O–K+
O
KH (stoichiometric)
K+
1
Ph
Ph
–HD gas
direct C1-H
deprotonation
Ph
Ph
Ph
Ph
KH
D
5
OH
O
D (76%)
KH (2 equiv.)
THF, r.t., 3 h
2
Ph
O
OH
3
Ph
Ph
Ph
Ph
Ph
OH
D (94%)
1a-d
5-d3, 89%
D (<2%)
Ph
Ph
Ph
Ph
Ph
Ph
Ph
OH
O
D (82%)
5
1a
D (99%)
KHMDS (2 equiv.)
THF, r.t., 3 h
2
Ph
1a-d
3
Ph
D (99%)
1a-d
O–K+
O
K+
5-d3, 90%
D (<2%)
kH1
kH1/kD1 = 7.6
Ph 1,2-hydride
shift
Ph
Ph
H
Scheme 3. Control experiments.
alkoxide A
homoenolate B
Scheme 5. Kinetics studies for the KHMDS-catalyzed
isomerizations of allylic alcohols 1a and 1a-d.
To further probe the validity of the 1,2-hydride-
shift mechanism, the rate of the reaction and KIE
were determined computationally using canonical
variational transition state theory (CVT) with small-
curvature tunneling (SCT) correction. These
calculations were carried out on the M06-2X/6-
31+G*
potential
energy
surface
using
GAUSSRATE[24] as the interface between
4
This article is protected by copyright. All rights reserved.

10.1002/adsc.202100272
Advanced Synthesis & Catalysis
POLYRATE[25] and Gaussian16.[26] The SMD model
was used to include solvent effects.[27] Of all possible
allylic-isomerization processes, the 1,2-hydride shift
was found to be the fastest, both with and without
considering tunneling (Figure 1). The CVT/SCT-
predicted reaction rate constant at 294.15 K of 1.57 ×
10-3 s-1 is in good agreement with the experimentally
determined rate constant (3.33 × 10-3 s-1). The KIE
was calculated by comparing the calculated reaction
rate constant for the 1,2-deuteride shift of 1.94 × 10-4
s-1 (CVT/SCT) with the above-mentioned value for
the analogous hydride shift; the CVT/SCT KIE value
of 8.1 is close to the experimental value. The CVT
KIE value of 4.2 (without tunneling correction) is
significantly different from the experimental value.
These computational results support the notion that
the 1,2-hydride shift is the rate-determining step in
this reaction and that tunneling accelerates this
hydride shift.
R1
O
OH
KHMDS
R2
O
NR3
2
R1
2
R2
O
O
oxa-Michael adduct
1
9
slow
R1
NR3
HMDS
R2
O–K+
1
R2
10
O–K+
R1
H
R2
protonation
O
2
R1
alkoxide A
tunneling-assisted
1,2-hydride shift
O–K+
NR3
rate-determining step
R1
K+
2
O
R2
R1
R2
3
homoenolate B
E
1
Michael addition
O
C3-protonation
O–K+
O
A
NR3
2
2
9
R1
R2
R1
R2
enolate D
C
1
A
C2-deprotonation
Scheme 6. Proposed catalytic cycle.
–O
VTS1
22.0
22.8 Ph
k
H,CVT/SCT = 5.24
H/D
Conclusion
kD,CVT/SCT = 2.74
∆GH
∆GD
Ph
(kcal/mol)
In conclusion, we developed novel KHMDS-
catalyzed Michael-addition chemistry of allylic
alcohols with ,-unsaturated amides that involves
allylic isomerization. A series of mechanistic studies,
including radical-inhibition, deuterium-labeling,
kinetics, and KIE experiments, along with a
computational study led us to propose that the in-situ
generation of the enolate from the allylic alcohol
through a tunneling-assisted 1,2-hydride shift is the
key step in this transformation. Further investigations
into the reaction mechanism and applications of this
protocol to the use of allylic alcohols as
homoenolates are currently underway in our
laboratory.
experimental KIE = 7.6
CVT/SCT KIE = 8.1
CVT KIE = 4.2
0.3
0.3
0.0
0.0
O
–O
Ph
H/D
Ph
Ph
Ph
H/D
PD
SM
M06-2X/6-31+G*/SMD(THF)
Figure 1. Free energy profile for the 1,2-hydride/deuteride
shift in the alkoxide of 1a. Transmission coefficient values
() are also shown. VTS = variational transition state.
Experimental Section
On the basis of the experimental and computaional
studies, we propose the reaction mechanism shown in
Scheme 6. The OH proton of 1 is initially abstracted
by KHMDS to form alkoxide A. Due to the low
electrophilicity of the ,-unsaturated amide 9, A is
reluctant to undergo oxa-Michael addition to 9 and is
therefore transformed into the homoenolate B
through a 1,2-hydride shift, which is the rate-
determining step and in which tunneling is involved.
B is rapidly protonated by 1 prior to reaction with 9,
and the -proton of the resulting ketone C is
deprotonated by A to generate enolate D, which then
undergoes Michael addition to 9 to give enolate E,
which is protonated by 1 to afford product 10 and
regenerate A.
Allylic alcohol 1 (0.36 mmol) was placed in an oven-dried
vial equipped with a magnetic stir bar. The vial was
flushed with argon and sealed with a rubber septum. To the
vial were added THF (3 mL), ,-unsaturated amide 9
(0.30 mmol), and KHMDS (0.5 M in toluene, 0.015 mmol,
30 L), and the mixture was stirred at room temperature
for 1.5 h. The reaction was quenched with saturated aq.
NH4Cl solution and extracted with Et2O. The organic layer
was dried over Na2SO4 and concentrated under reduced
pressure. The residue was purified by flash
chromatography on silica gel (hexane/EtOAc) to give the
corresponding product 10.
Acknowledgements
We are grateful to Prof. Shuji Akai (Osaka University) for his
kind and helpful discussions. We also thank Prof. Masayuki Inoue
(The University of Tokyo), Prof. Takeo Kawabata (Kyoto
University), Prof. Tomohiko Ohwada (The University of Tokyo),
Dr. Mitsuaki Ohtani (ITSUU Laboratory), and Dr. Kin-ichi
Tadano (ITSUU Laboratory) for helpful discussions. We thank
Dr. Masayoshi Ogawa (Shionogi Pharmaceutical Research
Center) for the X-ray crystallographic analysis. The
computations were performed at the Research Center for
Computational Science, Okazaki, Japan. This work was
5
This article is protected by copyright. All rights reserved.

10.1002/adsc.202100272
Advanced Synthesis & Catalysis
supported by JSPS KAKENHI (19K15576 to M.S. and 18K14205
to H.K.).
J. S. Johnston, M. G. McLaughlin, J. P. Reid, M. J.
Cook, Org. Biomol. Chem. 2013, 11, 7662–7666; b) B.
Suchand, G. Satyanarayana, Eur. J. Org. Chem. 2017,
3886–3895.
References
[6] For examples of base-promoted allylic isomerization to
homoenolates, see: a) T. Cuvigny, M. Julia, L. Jullien,
C. Rolando, Tetrahedron Lett. 1987, 28, 2587–2590; b)
W.-B. Wang, E. J. Roskamp, Tetrahedron Lett. 1992,
33, 7631–7634; c) M. Rehan, S. Maity, L. K. Morya, K.
Pal, P. Ghorai, Angew. Chem. Int. Ed. 2016, 55, 7728–
7732.
[1] For recent reviews, see: a) E. Emer, R. Sinisi, M. G.
Capdevila, D. Petruzziello, F. De Vincentiis, P. G.
Cozzi, Eur. J. Org. Chem. 2011, 647–666; b) B.
Sundararaju, M. Achard, C. Bruneau, Chem. Soc. Rev.
2012, 41, 4467–4483; c) R. Kumar, E. V. Van der
Eycken, Chem. Soc. Rev. 2013, 42, 1121–1146; d) N. A.
Butt, W. Zhang, Chem. Soc. Rev. 2015, 44, 7929–7967;
e) M. Dryzhakov, E. Richmond, J. Moran, Synthesis
2016, 48, 935–959; f) K. Spielmann, G. Niel, R. M. de
Figueiredo, J.-M. Campagne, Chem. Soc. Rev. 2018, 47,
1159–1173.
[7] Satyanarayana et al. succeeded in reducing the amount
of a base in some cases, but the reaction still required
as much as 50 mol% of KOtBu. See ref 5b.
[8] a) M. Sai, Adv. Synth. Catal. 2018, 360, 3482–3487; b)
M. Sai, Adv. Synth. Catal. 2018, 360, 4330–4335; c) M.
Sai, S. Matsubara, Adv. Synth. Catal. 2019, 361, 39–43;
d) M. Sai, Eur. J. Org. Chem. 2019, 1102–1106.
[2] For reviews, see: a) R. C. van der Drift, E. Bouwman,
E. Drent, J. Organomet. Chem. 2002, 650, 1–24; b) R.
Uma, C. Crévisy, R. Grée, Chem. Rev. 2003, 103, 27–
51; c) H. Suzuki, T. Takao in Ruthenium in Organic
Synthesis (Ed.: S.-I. Murahashi), Wiley-VCH,
Weinheim, 2004, pp. 309–331; d) G. C. Fu in Modern
Rhodium-Catalyzed Organic Reactions (Ed.: P. A.
Evans), Wiley-VCH, Weinheim, 2005, pp. 79–91; e) V.
Cadierno, P. Crochet, J. Gimeno, Synlett 2008, 1105–
1124; f) L. Mantilli, C. Mazet, Chem. Lett. 2011, 40,
341–344; g) N. Ahlsten, A. Bartoszewicz, B. Martín-
Matute, Dalton Trans. 2012, 41, 1660–1670; h) P.
Lorenzo-Luis, A. Romerosa, M. Serrano-Ruiz, ACS
Catal. 2012, 2, 1079–1086; i) D. Cahard, S. Gaillard,
J.-L. Renaud, Tetrahedron Lett. 2015, 56, 6159–6169;
j) H. Li, C. Mazet, Acc. Chem. Res. 2016, 49, 1232–
1241; k) D. Fiorito, S. Scaringi, C. Mazet, Chem. Soc.
Rev. 2021, 50, 1391–1406.
[9]For examples of the product-base mechanism, see: a) Y.
Yamashita, H. Suzuki, S. Kobayashi, Org. Biomol.
Chem. 2012, 10, 5750–5752; b) H. Suzuki, I. Sato, Y.
Yamashita, S. Kobayashi, J. Am. Chem. Soc. 2015, 137,
4336–4339.
[10] According to the reaction mechanism, the high
efficiency of potassium bases toward allylic
isomerization is likely due to the strong basicity, which
facilitates the 1,2-hydride shift.
[11] For a recent report on the electrophilicities of ,-
unsaturated carbonyl compounds, see: D. S. Allgäuer,
H. Jangra, H. Asahara, Z. Li, Q. Chen, H. Zipse, A. R.
Ofial, H. Mayr, J. Am. Chem. Soc. 2017, 139, 13318–
13329.ꢀ
[3] For selected examples of base-promoted allylic
isomerization, see: a) M. Tiffeneau, Bull. Soc. Chim. Fr.
1907, 1, 1205 (footnote on p. 1209); b) H. Burton, C. K.
Ingold, J. Chem. Soc. 1928, 904–921; c) D. R. Dimmel,
W. Y. Fu, S. B. Gharpure, J. Org. Chem. 1976, 41,
3092–3096; d) H. M. R. Hoffman, A. Köver, D.
Pauluth, J. Chem. Soc., Chem. Commun. 1985, 812–
814; e) G. A. Schmid, H.-J. Borschberg, Helv. Chim.
Acta 2001, 84, 401–415; f) X. Wang, D. Z. Wang,
Tetrahedron 2011, 67, 3406–3411.
[12] We also tested other potassium and cesium bases.
The use of KOtBu gave a somewhat lower yield (65%)
of 10aa. The use of CsOH·H2O provided a mixture of
oxa-Michael adduct (32%) and 10aa (22%) probably
due to the enhanced nucleophilicity of a cesium
alkoxide.
[13] The structure of 10ma was unambiguously identified
by spectroscopic and X-ray crystallographic analysis.
CCDC 1944310 contains the supplementary
crystallographic data for 10ma. These data can be
obtained free of charge from The Cambridge
[4] For examples of base-catalyzed allylic isomerization,
see: a) H.-X. Zheng, Z.-F. Xiao, C.-Z. Yao, Q.-Q. Li,
X.-S. Ning, Y.-B. Kang, Y. Tang, Org. Lett. 2015, 17,
6102–6105; b) K. Mondal, B. Mondal, S. C. Pan, J.
Org. Chem. 2016, 81, 4835–4840; c) S. Martinez-Erro,
A. Sanz-Marco, A. B. Gómez, A. Vázquez-Romero, M.
S. G. Ahlquist, B. Martín-Matute, J. Am. Chem. Soc.
2016, 138, 13408–13414; d) H.-X. Zheng, C.-Z. Yao,
J.-P. Qu, Y.-B. Kang, Org. Chem. Front. 2018, 5,
1213–1216; e) H.-X. Zheng, J.-P. Qu, Y.-B. Kang, Org.
Chem. Front. 2018, 5, 2349–2352; f) N. Molleti, S.
Martinez-Erro, A. C. Cerdán, A. Sanz-Marco, E.
Gomez-Bengoa, B. Martín-Matute, ACS Catal. 2019, 9,
9134–9139.
Crystallographic
Data
Centre
via
www.ccdc.cam.ac.uk/data_request/cif.
[14] In this case, oxa-Michael adduct was detected in 53%
NMR yield. The poor reactivity of 1q for the allylic
isomerization is probably due to the absence of an
anion-stabilizing substituent at the 3-position.
[15] The reaction of (E)-1-phenylhept-1-en-3-ol with
amide 9a under the standard reaction conditions
resulted in a 57% yield of the corresponding oxa-
Michael adduct.ꢀ
[16] Yates et al. suggested that a direct 1,2-proton shift
was a symmetry forbidden process and calculated to
have a barrier of 39.1 kcal/mol in a model stetter
reaction. a) K. J. Hawkes, B. F. Yates, Eur. J. Org.
[5] For examples of base-promoted tandem allylic
isomerization/electrophilic trapping sequence, see: a) A.
6
This article is protected by copyright. All rights reserved.

10.1002/adsc.202100272
Advanced Synthesis & Catalysis
Chem. 2008, 5563–5570. For DFT calculations on a
[22] See the Supporting Information for details of the
kinetics studies.
direct 1,2-proton shift, also see: b) B. Goldfuss, M.
Schumacher, J. Mol. Model. 2006, 12, 591–595; c) M.
Schumacher, B. Goldfuss, Tetrahedron 2008, 64,
1648–1653.
[23] For reviews on tunneling in organic chemistry, see: a)
R. S. Sheridan in Reviews of Reactive Intermediate
Chemistry (Eds.: M. S. Platz, R. A. Moss, M. Jones Jr.),
John Wiley & Sons, Hoboken, New Jersey, 2007, pp.
415–463; b) J. T. Hynes, J. P. Klinman, H.-H. Limbach,
R. L. Schowen in Hydrogen-Transfer Reactions,
Wiley-VCH, Weinheim, 2007; c) D. Ley, D. Gerbig, P.
R. Schreiner, Org. Biomol. Chem. 2012, 10, 3781–
3790; d) P. R. Schreiner, J. Am. Chem. Soc. 2017, 139,
15276–15283.
[17] We also conducted DFT calculations of a direct 1,2-
proton shift pathway, which had a high energy of 44.9
kcal/mol (see the Supporting Information).
[18] The mechanism of base-catalyzed D/H exchange at
the C2 position in 5-d2 is as follows:
O
O
OK
ROK
ROD
ROH
ROK
Ph
Ph
Ph
Ph
[24] J. Zheng, J. L. Bao, S. Zhang, J. C. Corchado, R.
Meana-Pañeda, Y.-Y. Chuang, E. L. Coitiño, B. A.
Ellingson, D. G. Truhlar, GAUSSRATE, version 2017-
B, University of Minnesota, Minneapolis, MN 2017.
Ph
Ph
D
H
5-d2
5
OH
ROH =
Ph
Ph
D
[25] J. Zheng, J. L. Bao, R. Meana-Pañeda, S. Zhang, B. J.
Lynch, J. C. Corchado, Y.-Y. Chuang, P. L. Fast, W.-P.
Hu, Y.-P. Liu, G. C. Lynch, K. A. Nguyen, C. F.
Jackels, A. Fernandez Ramos, B. A. Ellingson, V. S.
Melissas, J. Villà, I. Rossi, E. L. Coitiño, J. Pu, T. V.
Albu, A. Ratkiewicz, R. Steckler, B. C. Garrett, A. D.
Isaacson, D. G. Truhlar, POLYRATE, version 2017-B,
University of Minnesota, Minneapolis, MN 2017.
[19] Incorporation of a small amount of deuterium at the
C3 position is likely owing to the reaction of
homoenolate B with deuterated substrate ROD, which
is generated as C2–D/H exchange of 5-d2 (see ref 17).
OH
Ph
Ph
D
O
H
(ROH, excess)
2
Ph
Ph
C3-protonation
[26] M. J. Frisch, et al., Gaussian 16, Revision B.01,
Gaussian, Inc., Wallingford, CT, 2016. For the full
citation as well as the full computational details, see the
Supporting Information.
K+
3
O
O–K+
–A
D
major product
2
OD
2
1
Ph
Ph
Ph
3
1,2-hydride
shift
Ph
D
Ph
D
Ph
alkoxide A
O
D
3
D
homoenolate B
(ROD, small amount)
2
Ph
Ph
C3-deuteration
[27] A. V. Marenich, C. J. Cramer, D. G. Truhlar, J. Phys.
Chem. B. 2009, 113, 6378–6396.ꢀ
–A
D
minor product
[20] This mechanism is consistent with the results
reported in Table 1 where benzylation of 1a occurs
exclusively at the C2 position (3) and not at the C3
position (4) even in the presence of a stoichiometric
amount of a base.
[21] The abstraction of C1–D of alkoxide
A by
homoenolate B and dianion C would be much faster
than a 1,2-deuteride shift. This explains the negligible
amount of deuteration at the C2 position in 5-d3.
7
This article is protected by copyright. All rights reserved.

10.1002/adsc.202100272
Advanced Synthesis & Catalysis
FULL PAPER
Potassium Base-Catalyzed Michael Additions of
Allylic Alcohols to ,-Unsaturated Amides:
Scope and Mechanistic Insights
O–K+
1
K+
3
OH
O
cat. KHMDS
R1
H
R2
tunneling-assisted
1,2-hydride shift
R1
R2
R1
R2
2
homoenolate
allylic alcohol 1
C3-protonation/
C2-deprotonation
1
Adv. Synth. Catal. Year, Volume, Page – Page
O
O
O
O
O–K+
NR3R4
O–K+
NR3R4
Michael addition
R1
NR3R4 R1
product
R1
R2
Masahiro Sai* and Hiroaki Kurouchi
enolate
R2
R2
8
This article is protected by copyright. All rights reserved.