Catalyst-directed guidance of sulfur-substituted enediolates to stereoselective carbon-carbon bond formation with aldehydes

Article abstract of DOI:10.1021/jacs.7b12949

A highly chemo-, regio-, and stereoselective glycolate aldol reaction of sulfur-substituted enediolates with aldehydes was developed by employing a l-cyclohexylglycine-derived chiral iminophosphorane as a catalyst. The key for establishing this protocol i

Full text of DOI:10.1021/jacs.7b12949

Subscriber access provided by Kaohsiung Medical University
Catalyst-Directed Guidance of Sulfur-Substituted Enediolates to
Stereoselective Carbon-Carbon Bond Formation with Aldehydes
Daisuke Uraguchi, Kohei Yamada, Makoto Sato, and Takashi Ooi
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b12949 • Publication Date (Web): 04 Mar 2018
Downloaded from http://pubs.acs.org on March 4, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 9
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Catalyst‐Directed Guidance of Sulfur‐Substituted Enediolates to
Stereoselective Carbon‐Carbon Bond Formation with Aldehydes
Daisuke Uraguchi,† Kohei Yamada,† Makoto Sato,‡ and Takashi Ooi*†,‡
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
†Institute of Transformative Bio-Molecules (WPI-ITbM) and Department of Molecular and Macromolecular Chemistry,
Graduate School of Engineering, Nagoya University, Nagoya 464-8601, Japan.
‡CREST, Japan Science and Technology Agency (JST), Nagoya University, Nagoya 464-8601, Japan.
ABSTRACT: A highly chemo-, regio-, and stereoselective glycolate aldol reaction of sulfur-substituted enediolates with aldehydes
was developed by employing a L-cyclohexylglycine-derived chiral iminophosphorane as a catalyst. The key for establishing this
protocol is the distinct ability of the iminophosphorane catalyst to precisely direct the equilibrium mixture of the enediolates toward
the intermolecular carbon-carbon bond formation with simultaneous yet rigorous control of relative and absolute stereochemistry.
The critical importance of the cyclohexyl substituents on the catalyst backbone in dictating the reaction pathway and the
stereochemical outcome was elucidated through an extensive quantum analysis by density functional theory calculations.
mechanistically similar isomerization of O-acylated
hemithioacetals 2, readily accessible from the Pummerer
INTRODUCTION
Sulfur-substituted 1,2-enediolates are key intermediates in
the first step of the glyoxalase pathway, an essential two-step
detoxification of methylglyoxal and other reactive α-
oxoaldehydes produced in the metabolism of living organisms.
In this step, these glyoxals spontaneously react with the
coenzyme glutathione to form the requisite hemithioacetals that
undergo enantioselective isomerization to optically active α-
hydroxy thioesters, wherein lactoylglutathione lyase
(glyoxalase I) facilitates the facial- and regioselective
protonation of the enediolate by transferring an internal proton
from C1 to C2 of the hemithioacetal (Figure 1a).1 Largely due
to the synthetic importance of the resulting α-hydroxy thioesters
as precursors of various chiral α-hydroxy carbonyl compounds,
several small-molecule catalysts that mimic this enzymatic
process have been introduced.2 On the other hand, a
reaction of β-keto sulfoxides, is a well-documented strategy for
straightforward access to O-acylated α-hydroxy thioesters 3
(Figure 1b).3–5 This transformation also involves a sulfur-
substituted enediolate as a crucial intermediate and it exists as
an equilibrium mixture of three types of enediolates. Upon
considering their mutual interconversion via intramolecular
acyl migration, the base-mediated isomerization can be
understood as a consequence of the site-selective protonation of
the enediolate. This outcome probably stems from the
considerable pKa difference between the α-protons of 2 and 3,
which makes the protonation leading to 3 irreversible, and the
use of a chiral tertiary amine base as a catalyst has proven to be
effective for controlling its absolute stereochemistry.5c
Ironically, however, the inherent reactivity profile of the sulfur-
substituted enediolates as ambident carbon nucleophiles
Figure 1. Isomerization of hemithioacetals to α-hydroxy acids via enediolate intermediates and potential intermolecular bond formations of
the enediolate with an electrophile (El).
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 9
1
2
3
4
5
6
7
8
9
remains obscure within the protonation manifold. Moreover,
the potential utility of the hemithioacetal isomerization as a
means for catalytic generation and functionalization of highly
basic enolates of carboxylic acid oxidation state has never been
explored. In fact, despite significant mechanistic and synthetic
relevance, the union of an enediolate and an external carbon or
heteroatom electrophile into an α-tetrasubstituted α-hydroxy
carbonyl entity is entirely unknown, and so are catalyst-
controlled stereoselective systems. This deficiency is tightly
associated with not only the ambiguity of the regioselectivity
issue but also the stringent difficulty in promoting the
intermolecular bond formation predominantly over facile
protonation, while rigorously controlling the stereoselectivity.
Herein, we disclose a catalyst-directed approach to address this
challenging problem; that is, chiral iminophosphorane catalysts
of type 1,6-12 upon deprotonation of 2, precisely direct the
corresponding sulfur-substituted enediolate toward aldolization
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
with aldehydes with
a high level of diastereo- and
Scheme 1. Initial investigation of intrinsic reactivity of 2a.
enantiocontrol, thereby revealing the intrinsic reactivity
preference of the enediolates through the development of a
novel protocol for the rapid assembly of chiral α,β-
dihydroxycarboxylic acids bearing a fully substituted α-carbon
Table 1. Optimization of Reaction Conditionsa
stereocenter.
The origin of the chemo-, regio-, and
stereoselectivity observed in the present glycolate aldol reaction
is elucidated based on the free-energy profile derived from the
density functional theory (DFT) calculation.
entry
yield (%)b
5a/3ac
12:1
drc
ee (%)d
6
1
1
2
3
4
5
6
7
90
51
76
95
87
99
90
13:1
15:1
12:1
20:1
18:1
>20:1
>20:1
1a
1b
1c
1d
1e
1f
1:1
55
4:1
70
>20:1
>20:1
>20:1
>20:1
87
92
91
Figure 2. P-Spiro chiral iminophosphoranes.
92
1g
a Reactions were performed with 0.1 mmol of 2a and 0.11 mmol of
RESULTS AND DISCUSSIONS
b
4a in Et2O (1.0 mL) in the presence of 1 (10 mol%) at −40 °C.
c
Isolated yield of a diastereomeric mixture of 5a. Ratios of
At the outset of our research, we selected racemic α-acetoxy-
β-ketosulfide 2a as a representative substrate and assessed the
inherent reactivity of the enediolate generated from 2a in the
reaction with benzaldehyde (4a) by using a catalytic amount of
common bases (10 mol%) in diethyl ether at −40 °C. With
DBU as a catalyst, 2a underwent quantitative isomerization to
O-acetyl lactic acid thioester 3a, indicating that C2-protonation
of the intermediary enediolate occurred exclusively over the
aldol coupling with 4a. The use of a stronger base, KOtBu,
allowed for the intervention of the aldol reaction followed by
acyl migration to preferably afford a diastereomeric mixture of
5a (dr = 2.1:1, 58% yield) likely due to the reluctant protonation
by HOtBu; however, 3a was also produced in 32% yield
(Scheme 1). Notably, the regioisomeric aldol adduct 6 was not
detected under these conditions, which would reflect the
intrinsic difference in reactivity between the two nucleophilic
sites of the enediolate.
products and diastereomers were determined by 1H NMR (400
MHz) analysis of crude aliquot. d Absolute configuration of major
diastereomer of 5a was determined by X-ray crystallographic
analysis (Figure 3).16 Enantiomeric excesses of major diastereomer
of 5a were indicated, which were determined by using chiral
stationary phase HPLC with Daicel CHIRALPAK AD-3
(hexane/2-propanol/EtOH = 45:4:1 as eluent).
conditions, smooth consumption of 2a was observed within 20
h, resulting in the predominant formation of 5a together with
the protonation product 3a (5a:3a = 12:1) (Table 1, entry 1).
This result primarily suggests the importance of not only the
pKa value but also the structure of the base catalyst in diverting
reactivity of the enediolate from the protonation to the aldol-
acyl migration sequence. Although 5a was obtained with
relatively high diastereoselectivity (dr = 13:1), the enantiomeric
excess of the major isomer was determined to be only 6%. We
therefore pursued modification of the iminophosphorane
structure to examine the effect on the selectivity profile; this
revealed that the structural features of the alkyl substituent (R)
constitute a critical element for modulating the ability of 1 to
dictate the reaction pathway and the stereochemical outcome.
Based on these observations, we set out to investigate the
reaction under the influence of chiral iminophosphorane 1 in
order to evaluate the possibility of simultaneously controlling
reactivity and selectivity of the enediolate. In the initial attempt
with L-valine-derived 1a13 as a catalyst under otherwise similar
ACS Paragon Plus Environment

Page 3 of 9
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Switching the isopropyl moiety in 1a to α-non-branched alkyl
groups, such as methyl (1b) and isobutyl (1c), delivered an
appreciable level of enantioselectivity, but the product
distribution showed substantial participation of the undesired
protonation (entries 2 and 3). In marked contrast, an increase
in the steric bulkiness of the α-branched alkyl group from
isopropyl to (S)-sec-butyl (1d) or cyclohexyl (1e) turned out to
be beneficial for completely suppressing the formation of 3a,
while significantly improving the stereoselectivity (entries 4
and 5).14 Further tuning of the properties of the aromatic
appendage (Ar) enabled the identification of L-
cyclohexylglycine-derived iminophosphorane 1g bearing 4-
fluorophenyl groups as an optimal catalyst, and 5a was solely
isolated as a single diastereomer (dr = >20:1) with high
enantiomeric excess (92%) (entries 6 and 7).15 The three-
dimensional structure of 5a was determined by single-crystal
X-ray diffraction analysis to confirm the relative and absolute
stereochemistry and regiochemistry at the carbon-carbon bond-
forming event (Figure 3).16
Figure 3. ORTEP diagram of 5a (Ellipsoids displayed at 50%
probability. Calculated hydrogen atoms except for them attached to
the stereogenic carbon are omitted for clarity. Gray: carbon, red:
oxygen, yellow: sulfur.).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
employable regardless of their steric and electronic attributes
(entries 1–10).
However, when a strongly electron-
withdrawing group such as methoxycarbonyl group was
introduced to the para-position of the aromatic ring, lowering
the reaction temperature was necessary to attain a synthetically
satisfactory level of enantioselectivity, while virtually complete
control of the reaction pathway and relative stereochemistry
was consistently achievable (entry 4). 2-Naphthaldehyde and
heteroaromatic aldehydes were also amenable to this catalytic
system (entries 11–14). Although (E)-cinnamaldehyde could
be accommodated by conducting the aldolization at lower
With the optimized iminophosphorane catalyst in hand,
further experiments were conducted to explore the scope and
limitation of this new glycolate aldol protocol (Table 2). As the
electrophilic component, a series of aromatic aldehydes 4 were
Table 2. Substrate Scope and Limitationa
entry
1
R1 (2)
Me (2a)
R2 (4)
4-MeC6H4 (4b)
4-ClC6H4 (4c)
4-BrC6H4 (4d)
4-(MeOCO)C6H4 (4e)
3-MeC6H4 (4f)
3-MeOC6H4 (4g)
3-FC6H4 (4h)
2-MeC6H4 (4i)
2-MeOC6H4 (4j)
3,4-(CH2O2)C6H3 (4k)
2-Naphthyl (4l)
3-furyl (4m)
yield (%)b
90
drc
ee (%)d
92
prod.
5b
5c
>20:1
>20:1
>20:1
>20:1
>20:1
>20:1
>20:1
>20:1
>20:1
12:1
2
Me (2a)
94
92
3
4e,f
Me (2a)
98
92
5d
5e
Me (2a)
83
85
5
Me (2a)
95
92
5f
6
Me (2a)
94
94
5g
5h
5i
7
Me (2a)
94
84
8
Me (2a)
74
93
9
Me (2a)
88
96
5j
10
11
12
13
14
15e,g
16h
17h
18
19e,f
20
Me (2a)
90
95
5k
5l
Me (2a)
98
>20:1
>20:1
>20:1
>20:1
20:1
93
Me (2a)
88
93
5m
5n
5o
5p
5q
5r
Me (2a)
3-thienyl (4n)
2-thienyl (4o)
(E)-PhCH=CH (4p)
Ph(CH2)2 (4q)
Ph (4a)
95
91
Me (2a)
96
86
Me (2a)
95
83
Me (2a)
66
10:1
37
Et (2b)
80
14:1
97
MeOCH2 (2c)
Me(CH2)3 (2d)
(E)-MeCH=CH (2e)
Ph (4a)
73
17:1
87
5s
Ph (4a)
68
6.7:1
1.5:1
94
92i
5t
Ph (4a)
79
5u
a Unless otherwise noted, reactions were conducted with 0.1 mmol of 2 and 0.11 mmol of 4 in Et2O (1.0 mL) in the presence of 1g (10 mol%)
at −40 °C for 20 h. b Isolated yield of a mixture of diastereomers was indicated. c Diastereomeric ratios were determined by 1H NMR (400
d
MHz) analysis of crude aliquot. Absolute configuration of 5 was assigned by analogy of 5a. Enantiomeric excesses of the major
diastereomer were indicated, which were analyzed by chiral stationary phase HPLC. e Reaction was conducted at −60 °C. Reaction time
f
was 96 h. g Reaction time was 72 h. h Reaction time was 48 h. i Enantiomeric excess of the minor diastereomer was 25%.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 9
1
2
3
4
5
6
7
8
9
temperature, considerable decrease in stereoselectivity was
Fukuyama’s procedure into the α-hydroxy aldehyde 8 without
inevitable in the coupling with an aliphatic aldehyde (entries 15
and 16). With respect to the nucleophile 2, variation in the acyl
substituent appeared feasible, albeit with a slight decrease in the
efficiency and stereocontrol (entries 17–19). It should be added
that the loss in diastereoselectivity was significant in the
reaction of α-acetoxy-β-ketosulfide having an enone moiety
(entry 20).
loss of the stereochemical integrity.17
To gain insight into the origin of the catalyst-directed
multiple selectivity control, particularly the precise dictation of
the reaction pathway, we performed DFT calculations
(Gaussian 09)18 for analyzing the reaction of the realistic
substrate set, α-acetoxy-β-ketosulfide 2a and benzaldehyde (4a),
with iminophosphorane 1e as a catalyst (entry 5 in Table 1).7k,19
Geometries of the intermediates and transition states were fully
optimized and characterized by frequency calculations at the
PCM(Et2O)-ωB97XD/6-31G(d,p) level. Single-point energies
of the optimized structures at the PCM(Et2O)-ωB97XD/6-
311++G(d,p) level were calculated and combined with the
thermal corrections estimated by PCM(Et2O)-ωB97XD/6-
31G(d,p) to obtain Gibbs free energies (ΔG). The calculated
Gibbs free energy profile (233.15 K, 1 atm) is displayed in
Figure 4.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
As illustrated in Scheme 2, treatment of the product 5a with
First, deprotonation of racemic 2a with 1e generates the
corresponding aminophosphonium enediolate 1eꞏHꞏ[2a–H]
(Enol1-E), where 1eꞏH featuring two N-H protons captures
[2a–H]− in a way that one N-H proton makes a hydrogen bond
with the anionic enolate oxygen and the other with the acetate
carbonyl. This deprotonation proceeds via the relatively high-
energy transition state (TS) TS-PT1-RE (from (R)-2a) and TS-
PT1-SE (from (S)-2a) (+6.9 and +9.1 kcalꞏmol−1, respectively).
Scheme 2. Conversion of 5a to α,β-dihydroxy ester 7 and α-
hydroxy aldehyde 8.
sodium methoxide at
0 °C facilitated smooth double
The Enol1-E thus generated could potentially participate in
the three different reactions, namely (i) re-protonation at C1 to
liberate the parent 2a and the catalyst 1e (TS-PT1-RE), (ii)
aldol reaction with 4a via TS-CC1-SS,20 and (iii)
intramolecular acyl migration to give the putative cyclic
enediolate INT-Cyc1 via TS-Cyc1. Optimization of the
structures and calculation of Gibbs free energies of appropriate
TS models for these reaction courses confirmed that the acyl
migration was the most favorable pathway as TS-Cyc1 has the
lowest free energy (+5.9 kcalꞏmol−1). This outcome reflects the
higher energy estimated for the TS for the C-C bond formation,
transesterification of the thioester moiety and the acetyl unit to
afford α,β-dihydroxy ester 7 in good yield. This simple
derivatization demonstrates the immediate utility of the present
method as a reliable tool for straightforward access to
stereochemically homogeneous α,β-dihydroxycarboxylic acids
and their derivatives possessing a fully substituted stereocenter
at the α-carbon atom, which are often encountered as partial
structures of biologically active natural products and also serve
as valuable chiral synthons. By further taking advantage of the
thioester moiety that is known to be easily converted into other
carbonyl functionalities, 5a could be effectively reduced by the
Figure 4. Gibbs free energy profile of the reaction at PCM(Et2O)-ωB97XD/6-311++G(d,p)//PCM(Et2O)-ωB97XD/6-31G(d,p). Gibbs free
energies relative to RT are in kcalꞏmol−1 (233.15 K, 1 atm). RT = reactant, RC = reactant complex, PC = product complex, PD = product.
ACS Paragon Plus Environment

Page 5 of 9
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
TS-CC1-SS, which is in accord with the experimental
substituents of 1eꞏH (highlighted in green dotted circle), and the
steric repulsion between the acetate moiety and the other
cyclohexyl substituent is relieved. Moreover, the SPh moiety of
[3a−H]− is disposed outside the chiral cavity of 1eꞏH and is
liberated from any steric constraints. The interplay of the
attractive CH-π interaction and the favorable steric effects
compensates the destabilization caused by the loss of otherwise
preferable hydrogen-bonding interaction to stabilize the enolate
anion, which contributes to the small energy barrier calculated
for the aldolization via TS-CC2-SR.
observation that the aldol reaction from Enol1-E did not take
place regardless of the base catalyst employed likely due to the
innate instability of the product ion pair.21
The resulting INT-Cyc1 either leads to generate another
aminophosphonium enediolate 1eꞏHꞏ[3a–H] (Enol2-E) via TS-
Cyc2 or goes back to Enol1-E. Judging from the relatively low
energy barriers in the acyl migration processes, the two
aminophosphonium enediolates, Enol1-E and Enol2-E, would
equilibrate slowly through the intermediacy of INT-Cyc1,
although the equilibrium is largely inclined toward Enol1-E.
The transient INT-Cyc1 lies higher in energy and would have
low probability of being present. In fact, a TS model for the C-
C bond formation from INT-Cyc1 could not be located.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In addition, the high energy barrier calculated for the
protonation of [3a−H]− by the pairing 1eꞏH is crucial for 1eꞏH
to guide [3a−H]− toward intermolecular aldol coupling.23 The
instability of the TSs for the protonation at the C-2 position of
[3a−H]− (TS-PT2-SE and TS-PT2-RE) could be attributed to
the structural feature of 1eꞏH bearing cyclohexyl substituents.
The steric hindrance of the cyclohexyl groups allows 1eꞏH to
exploit a significant repulsive nonbonding interaction with
[3a−H]− in restraining the formation of TS-PT2-SE and TS-
PT2-RE; this raises the energy barrier for [3a−H]− to undergo
protonation to yield 3a (+5.6 and +7.6 kcalꞏmol−1 for S- and R-
3a, respectively), exceeding the free energy of TS-CC2-SR for
the C-C bond formation process.
Interestingly, the TS for the aminophosphonium enediolate
Enol2-E to engage in the aldolization with 4a (TS-CC2-SR)
has a lower free energy (+4.4 kcalꞏmol−1) than that of the TS for
the reverse acyl migration (TS-Cyc2), thus rendering this C-C
bond formation pathway predominant. It is worthy of note that,
upon forming TS-CC2-SR, the aminophosphonium ion 1eꞏH
releases the anionic enolate oxygen of [3a−H]−, rather than the
carbonyl oxygen, through the disruption of the hydrogen bond
and incorporates 4a using the resultant free N-H proton as
visualized by the three-dimensional representation in Figure
5a.22 Compared to another possible TS that retains the
hydrogen bonding from the N-H proton to the anionic oxygen
(TS-CC2-SR’, Figure 5b), 4a is aligned as to enjoy CH-π
interaction between the phenyl group and one of the cyclohexyl
The critical importance of the cyclohexyl substituents of
1eꞏH in overturning the inherent reactivity preference of the
enediolate [3a−H]− was also corroborated by delineating the
free energy profile of the reaction with the L-alanine-derived
iminophosphorane 1b possessing methyl substituents as a
catalyst. The calculated Gibbs free energy of the analogous TS
for the 1b-catalyzed aldolization giving SR-5a (TSala-CC2-SR)
was higher than that of TS-CC2-SR mainly due to the lack of
the CH-π interaction (Figure 5a vs 5c). Rather, another TS,
TSala-CC2-SR’ (Figure 5d), where one N-H proton of 1bꞏH
interacts with the anionic oxygen of [3a−H]−, lies lower in
energy (+4.8 kcalꞏmol−1), suggesting that no particular steric
repulsion between the acetate moiety of [3a−H]− and one
methyl substituent of 1bꞏH allows to appreciate the stabilization
of the enolate anion by the hydrogen-bonding interaction
(Figure 5b vs 5d). Furthermore, the sterically less demanding
1bꞏH is also able to form a slightly tighter ion pair with [3a−H]−
without being interfered with notable collisions, lowering the
free energy of the TS for protonation (TSala-PT2-RE: +5.1
kcalꞏmol−1);21 this makes the free energy difference between
TSala-CC2-SR’ and TSala-PT2-RE small enough for both
processes to occur in a comparable rate as experimentally
observed (see Table 1, entry 2).
Having grasped
a
detailed picture of how the
iminophosphorane catalyst guides the sulfur-substituted
enediolate toward the aldol reaction, we next compared the
Gibbs free energies of the possible diastereomeric TS models
for the 1e-catalyzed aldolization of 2a with 4a to understand the
stereoselectivity profile (Figure 6). While the energy values
calculated
at
the
PCM(Et2O)-ωB97XD/6-
311++G(d,p)//PCM(Et2O)-ωB97XD/6-31G(d,p) level (233.15
K, 1 atm) successfully predicted that TS-CC2-SR giving the
experimentally obtained major stereoisomer was the most
stable TS, the free energy differences (ΔΔG) between other
diastereomeric TSs were not well correlated with the observed
diastereo- and enantioselectivity. Therefore, in order to
improve the quantitative reproducibility in stereoselectivity, we
conducted single point calculations on these TSs using the
M06-2X functional known to be suitable for analyzing aldol
Figure 5. Three dimensional structures and Gibbs free energies
relative to RT (233.15 K, 1 atm) in TSs for C-C-bond formation in
1e- and 1b-catalyzed reactions. (a) TS-CC2-SR, (b) TS-CC2-SR’,
(c) TSala-CC2-SR, (d) TSala-CC2-SR’. Interatomic distances are
in Å. Unimportant hydrogen atoms are omitted for clarity.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 9
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(e)
ΔEdis (kcalꞏmol−1)
ΔEint
ΔEsol
ΔEtot (ΔEdis+ΔEint+ΔEsol
(kcalꞏmol−1)
)
sub
cat
(kcalꞏmol−1)
(kcalꞏmol−1)
SR
RS
SS
RR
0.0
0.0
0.0
−10.7
−2.8
0.0
0.0
+7.5
+5.9
+9.4
+3.6
−0.2
+3.6
+0.5
−0.6
+0.5
+0.9
+2.3
+1.9
−11.7
Figure 6. (a)~(d) TS-models leading possible four diastereomers at the aldolization. Gibbs free energies relative to TS-CC2-SR are in
kcalꞏmol−1 (233.15 K, 1 atm) at PCM(Et2O)-ωB97XD/6-311++G(d,p)//PCM(Et2O)-ωB97XD/6-31G(d,p) [PCM(Et2O)-M06-2X/6-
311++G(d,p)//PCM(Et2O)-ωB97XD/6-31G(d,p)]. Unimportant hydrogen atoms are omitted for clarity. (e) Distortion/interaction analysis on
diastereomeric TS-CC2 structures at M06-2X/6-311++G(d,p)//PCM(Et2O)-ωB97XD/6-31G(d,p).
reactions (indicated in square brackets).24 As a result, the
derived energy difference between TS-CC2-SR and TS-CC2-
RS (ΔΔG = +1.5 kcalꞏmol−1) was in good agreement with the
experimentally observed enantioselectivity, but the difference
aldehyde (Table 2, entry 16). The large negative ΔEint values in
TS-CC2-RR and TS-CC2-RS clearly indicate that the
hydrogen bond between the N-H proton of 1eꞏH and the anionic
oxygen of [3a−H]− is strongly favored. However, the catalyst
must be distorted to incorporate the substrates for the
intermolecular aldolization in a manner that the acetate moiety
of [3a−H]− is placed close to one of the cyclohexyl substituents
of 1eꞏH (+3.6 kcalꞏmol−1). Consequently, distortion energies
slightly surpass the stabilization effect provided by the
hydrogen-bonding interactions, resulting in the destabilization
of TS-CC2-RR and TS-CC2-RS. In TS-CC2-SS, the mode of
the hydrogen-bonding interactions was the same with that in
TS-CC2-SR and they were even stronger as evident from the
shorter HꞏꞏꞏO distance and more linear N-HꞏꞏꞏO angle of the
hydrogen bonding to the acetate carbonyl of [3a−H]−.
Nevertheless, the interactions beneficial for gaining
stabilization appear to be too small to override the influence of
the distortion energy mainly arising from the proximity between
the SPh moiety of [3a−H]− and one of the geminal phenyl
substituents of 1eꞏH, rendering TS-CC2-SS unstable.
between TS-CC2-SR and TS-CC2-SS (ΔΔG
= +0.8
kcalꞏmol−1) was not sufficiently large to rationalize the
observed diastereoselectivity.
Since the calculations at M06-2X/6-311++G(d,p)
qualitatively accounted for the stereoselectivity of the C-C bond
formation, we investigated the origin of the obtained energy
differences by applying the distortion/interaction analysis to the
diastereomeric TSs. The relative distortion (ΔEdis) and
interaction (ΔEint) energies with respect to the values in TS-
CC2-SR are summarized in Figure 6e. In TS-CC2-RR and TS-
CC2-RS, a large positive ΔEdis (+13.0 and +11.1 kcalꞏmol−1)
and negative ΔEint ( − 11.7 and − 10.7 kcalꞏmol−1) were
recognized, where ΔEdis of the substrate was much larger than
that of the catalyst. On the other hand, TS-CC2-SS showed a
moderate positive ΔEdis of the substrate (+5.9 kcalꞏmol−1) and a
small negative ΔEint ( − 2.8 kcalꞏmol−1). The significant
differences of ΔEdis of the substrate in these three TSs from that
in TS-CC2-SR could originate from the electrostatic repulsion
between the oxygen atoms in the approaching substrates (see
Newman projections in Figure 6). Specifically, the oxygen
atom of 4a, on which a negative charge is developed along with
the formation of a new C-C bond, is oriented gauche to both the
acetate and the thioester enolate components of the enediolate
[3a−H]−. Only TS-CC2-SR is free from this issue given that
the oxygen atom of 4a adopts an anti relation to the negatively
charged thioester enolate oxygen. In addition, the CH-π
interaction between the phenyl group of 4a and the cyclohexyl
substituent of 1eꞏH contributes to the stabilization of TS-CC2-
SR, which is in accordance with the diminished
stereoselectivity observed in the reaction with the aliphatic
The C-C bond-forming event accompanied by the
asynchronous proton transfer from 1eꞏH afforded a neutral
complex 1eꞏ5a’ (INT-CC2N). Subsequently, INT-CC2N was
converted into a stable complex of 1e with 5a (PC) through
multiple processes, including intramolecular acyl migration.19
PC would liberate the aldol product with the concomitant
regeneration of the iminophosphorane 1e, closing the entire
catalytic cycle.
CONCLUSIONS
We demonstrated that sulfur-substituted enediolates, generated
in situ from α-acylated hemithioacetals 2, selectively engaged
in the glycolate aldol reaction with aldehydes 4 with an
excellent level of diastereo- and enantiocontrol under the
ACS Paragon Plus Environment

Page 7 of 9
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
(2) (a) Schmitt, E.; Schiffers, I.; Bolm, C. Tetrahedron Lett. 2009,
50, 3185–3188. (b) Park, S. Y.; Hwang, I.-S.; Lee, H.-J.; Song, C. E.
Nat. Commun. 2017, 8, 14877.
catalysis
iminophosphorane 1g.
of
the
L-cyclohexylglycine-derived
chiral
While the pKa balance in the
intermediary formed aminophosphonium enediolate 1ꞏHꞏ[3−H]
would urge a pseudo-intramolecular proton transfer to liberate
1 and α-hydroxy thioester 3, the intermolecular carbon-carbon
bond formation proceeded predominantly over the protonation
in a highly stereoselective manner when 1g was employed as a
catalyst. The unique and relevant features of 1g for precisely
dictating the reaction pathway and the stereochemical outcome
were dissected by extensive quantum analysis, revealing the
critical importance of cyclohexyl substituents in the
simultaneous control of the multiple selectivities. These results
shed light on the previously unexplored possibility of utilizing
the base-catalyzed hemithioacetal isomerization as a tool for the
generation of highly basic enolates by the judicious choice of a
catalyst capable of kinetically suppressing the rapid protonation
and directing the subsequent stereoselective bond-forming
reactions. This significant implication will stimulate the
development of new catalyst-controlled strategies for achieving
otherwise difficult transformations under simple acid-base
catalysis.
(3) (a) de Lucchi, O.; Miotti, U.; Modena, G. Org. React. 1991, 40,
157–405. (b) Padwa, A.; Gunn, Jr., D. E.; Osterhout, M. H. Synthesis
1997, 1353–1377. (c) Padwa, A.; Bur, S. K.; Danca, D. M.; Ginn, J. D.;
Lynch, S. M. Synlett 2002, 851–862. (d) Bur, S. K.; Padwa, A. Chem.
Rev. 2004, 104, 2401–2432.
(4) (a) Iriuchijima, S.; Maniwa, K.; Tsuchihashi, G.-I. J. Am. Chem.
Soc. 1975, 97, 596–599. (b) Russell, G. A.; Mikol, G. J. J. Am. Chem.
Soc. 1966, 88, 5498–5504. (c) Capitta, F.; Melis, N.; Secci, F.;
Romanazzi, G.; Frongia, A. J. Sulfur Chem. 2014, 35, 649–660.
(5) (a) Mimoto, T.; Terashima, K.; Nojima, S.; Takaku, H.;
Nakayama, M.; Shintani, M.; Yamaoka, T.; Hayashi, H. Bioorg. Med.
Chem. 2004, 12, 281–293. (b) Suzuki, T.; Honda, Y.; Izawa, K.;
Williams, R. M. J. Org. Chem. 2005, 70, 7317–7323. (c) Capitta, F.;
Frongia, A.; Piras, P. P.; Pitzanti, P. Adv. Synth. Catal. 2010, 352,
2955–2960.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(6) (a) Corbett, M. T.; Uraguchi, D.; Ooi, T.; Johnson, J. S. Angew.
Chem. Int. Ed. 2012, 51, 4685–4689. (b) Horwitz, M. A.; Tanaka, N.;
Yokosaka, T.; Uraguchi, D.; Johnson, J. S.; Ooi, T. Chem. Sci. 2015, 6,
6086–6090.
(7) (a) Uraguchi, D.; Ooi, T. J. Synth. Org. Chem. Jpn. 2010, 68,
1185–1194. (b) Uraguchi, D.; Nakamura, S.; Ooi, T. Angew. Chem. Int.
Ed. 2010, 49, 7562–7565. (c) Uraguchi, D.; Yoshioka, K.; Ueki, Y.;
Ooi, T. J. Am. Chem. Soc. 2012, 134, 19370–19373. (d) Uraguchi, D.;
Tsutsumi, R.; Ooi, T. J. Am. Chem. Soc. 2013, 135, 8161–8164. (e)
Uraguchi, D.; Yamada, K.; Ooi, T. Angew. Chem. Int. Ed. 2015, 54,
9954–9957. (f) Uraguchi, D.; Yoshioka, K.; Ooi, T. Nat. Commun.
2017, 8, 14793. (g) Uraguchi, D.; Ito, T.; Kimura, Y.; Nobori, Y.; Sato,
M.; Ooi, T. Bull. Chem. Soc. Jpn. 2017, 90, 546–555. (h) Yoshioka, K.;
Yamada, K.; Uraguchi, D.; Ooi, T. Chem. Commun. 2017, 53, 5495–
5498. (i) Tanaka, N.; Tsutsumi, R.; Uraguchi, D.; Ooi, T. Chem.
Commun. 2017, 53, 6999–7002. (j) Uraguchi, D.; Kawai, Y.; Sasaki,
H.; Yamada, K.; Ooi, T. Chem. Lett. 2018, 47, doi: 10.1246/cl.180031.
(k) Uraguchi, D.; Shibazaki, R.; Tanaka, N.; Yamada, K.; Yoshioka,
K.; Ooi, T. Angew. Chem. Int. Ed. 2018, 57, doi:
10.1002/anie.201800057.
(8) For a review on organic base catalysis, see: Palomo, C.; Oiarbide,
M.; López, R. Chem. Soc. Rev. 2009, 38, 632–653.
(9) For reviews on phosphonium salt catalyses, see: (a) Werner, T.
Adv. Synth. Catal. 2009, 351, 1469–1481. (b) Enders, D.; Nguyen, T.
V. Org. Biomol. Chem. 2012, 10, 5327–5331.
(10) Triaminoiminophosphoranes are known as a P1-phosphazenes,
which was developed by Schwesinger, see: (a) Schwesinger, R.;
Schlemper, H. Angew. Chem. Int. Ed. Engl. 1987, 26, 1167–1169. For
a review, see: (b) Ishikawa, T. Superbases for Organic Synthesis:
Guanidines, Amidines, Phosphazenes and Related Organocatalysts;
John Wiley & Sons: West Sussex, U.K., 2009.
(11) For a review on chiral iminophosphorane catalysis, see:
Krawczyk, H.; Dzięgielewski, M.; Deredas, D.; Albrecht, A.; Albrecht,
Ł. Chem. Eur. J. 2015, 21, 10268–10277.
(12) (a) Takeda, T.; Terada, M. J. Am. Chem. Soc. 2013, 135, 15306–
15309. (b) Núñez, M. G.; Farley, A. J. M.; Dixon, D. J. J. Am. Chem.
Soc. 2013, 135, 16348–16351. (c) Gao, X.; Han, J.; Wang, L. Org. Lett.
2015, 17, 4596–4599. (d) Kondoh, A.; Oishi, M.; Takeda, T.; Terada,
M. Angew. Chem. Int. Ed. 2015, 54, 15836–15839. (e) Farley, A. J. M.;
Sandford, C.; Dixon, D. J. J. Am. Chem. Soc. 2015, 137, 15992–15995.
(f) Robertson, G. P.; Farley, A. J. M.; Dixon, D. J. Synlett 2016, 27,
21–24. (g) Horwitz, M. A.; Zavesky, B. P.; Martinez-Alvarado, J. I.;
Johnson, J. S. Org. Lett. 2016, 18, 36–39. (h) Gao, X.; Han, J.; Wang,
L. Org. Chem. Front. 2016, 3, 656–660. (i) Takeda, T.; Kondoh, A.;
Terada, M. Angew. Chem. Int. Ed. 2016, 55, 4734–4737. (j) Yang, J.;
Farley, A. J. M.; Dixon, D. J. Chem. Sci. 2017, 8, 606–610.
(13) The hydrochloric acid salt of 1a (P-VIPꞏCl) is commercially
available from Wako pure chemical industries (Cat. No. 226-02381).
(14) Upon treatment of 3a with 10 mol% of 1d in the presence of 4a
in diethyl ether, neither formation of 5a nor consumption of 3a was
detected even at 0 °C. This indicated that the observed rigorous
chemoselectivity was kinetically established.
ASSOCIATED CONTENT
*Supporting Information
The Supporting Informations are available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.0000000.
Experiment details, spectral data, copies of 1H and 13C
NMR spectra, and copies of HPLC traces (PDF)
Computational details and discussion (PDF).
Cartesian coordinates and energies (PDF).
X-ray crystallographic data for 1eꞏHCl and 5a (CIF)
AUTHOR INFORMATION
Corresponding Author
*tooi@chembio.nagoya-u.ac.jp
ORCID
Takashi Ooi: 0000-0003-2332-0472
Daisuke Uraguchi: 0000-0002-6584-797X
Makoto Sato: 0000-0001-8903-6226
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
Financial support was provided by CREST-JST (JPMJCR13L2:
13418441), Program for Leading Graduate Schools "Integrative
Graduate Education and Research Program in Green Natural
Sciences" in Nagoya University, and Grants of JSPS for Scientific
Research. KY acknowledges JSPS for financial support. The
computations were performed using Research Center for
Computational Science, Okazaki, Japan.
REFERENCES
(1) (a) Hall, S. S.; Doweyko, A. M.; Jordan, F. J. Am. Chem. Soc.
1976, 98, 7460–7461. (b) Hall, S. S.; Doweyko, A. M.; Jordan, F. J.
Am. Chem. Soc. 1978, 100, 5934–5939. (c) Shinkai, S.; Yamashita, T.;
Kusano, Y.; Manabe, O. Chem. Lett. 1979, 1323–1326. (d) Shinkai, S.;
Yamashita, T.; Kusano, Y.; Manabe, O. J. Am. Chem. Soc. 1981, 103,
2070–2074. (e) Landro, J. A.; Brush, E. J.; Kozarich, J. W.
Biochemistry 1992, 31, 6069–6077. (f) Ridderström, M.; Cameron, A.
D.; Jones, T. A.; Mannervik, B. J. Biol. Chem. 1998, 273, 21623–21628.
(g) Thornalley, P. J. Biochem. Soc. Trans. 2003, 31, 1343–1348.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 9
1
2
3
4
5
6
7
8
9
(15) When the 1g-catalyzed reaction was quenched after 2 h (37%
T. J. Org. Chem. 2017, 82, 541–548. (c) Simón, L.; Paton, R. S. J. Org.
Chem. 2017, 82, 3855–3863.
(20) See the Computational Supporting Information (CSI) for other
TSs having higher energy.
(21) See CSI for details.
(22) CYLview, 1.0b, Legault, C. Y. Université de Sherbrooke, 2009.
(23) When the reaction was conducted in the presence of MS4A,
essentially the same result was obtained, indicating that protonation of
[3a−H]− with water molecule is unlikely. While proton delivery from
either 2a or 3a is another possible scenario, intervention of this process
would also be negligible because the protonation/aldolization ratio was
found to be insensitive to the reaction concentration and optimized TSs
for such protonation of [3a−H]− have higher energy than that of TS-
PT2-SE (See S34 in CSI).
conversion), enantiomerically enriched 2a (15% ee) was recovered.
(16) Crystallographic data (excluding structure factors) for the
structure reported in this paper have been deposited with the Cambridge
Crystallographic Data Centre as supplementary publication no. CCDC
1588205 (5a) and CCDC 1588206 (1gꞏHCl). Copies of the data can be
obtained free of charge on application to CCDC, 12 Union Road,
Cambridge CB2ꢀ1EZ, UK (fax: (+44)ꢀ1223-336-033; e-mail:
deposit@ccdc.cam.ac.uk).
(17) (a) Fukuyama, T.; Lin, S.-C.; Li, L. J. Am. Chem. Soc. 1990,
112, 7050–7051. (b) Tokuyama, H.; Yokoshima, S.; Lin, S.-C.; Li, L.;
Fukuyama, T. Synthesis 2002, 1121–1123.
(18) Gaussian 09, Revision D.01., Frisch, M. J. et al. Gaussian, Inc.:
Wallingford, CT, 2013.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(19) Computational analysis on P-spiro chiral iminophosphorane
catalysis, see: (a) Simón, L.; Paton, R. S. J. Org. Chem. 2015, 80, 2756–
2766. (b) Yamanaka, M.; Sakata, K.; Yoshioka, K.; Uraguchi, D.; Ooi,
(24) (a) Um, J. M.; Gutierrez, O.; Schoenebeck, F.; Houk, K. N.;
MacMillan, D. W. C. J. Am. Chem. Soc. 2010, 132, 6001–6005. (b)
Grayson, M. N.; Houk, K. N. J. Am. Chem. Soc. 2016, 138, 9041–9044.
ACS Paragon Plus Environment

Page 9 of 9
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9
ACS Paragon Plus Environment