Chiral Phosphoric Acid Catalyzed Enantioselective Ring Expansion Reaction of 1,3-Dithiane Derivatives: Case Study of the Nature of Ion-Pairing Interaction

Article abstract of DOI:10.1021/jacs.7b13274

Chiral counterion controlled asymmetric catalysis via an ion-pairing interaction has attracted immense attention in recent years. Despite a number of successful studies, the mechanistic elucidation of the stereocontrolling element in the ion-pairing interaction is rarely conducted and hence its nature is still far from being well understood. Herein we report an in-depth mechanistic case study of a newly developed enantioselective ring expansion reaction of 1,3-dithiane derivatives catalyzed by chiral phosphoric acid (CPA). An unprecedented enantioselective 1,2-sulfur rearrangement/stereospecific nucleophilic addition sequence was proven to be the stereoselective pathway. More importantly, by thorough investigation of the intrinsic nature of the stereospecific nucleophilic addition to the cationic thionium intermediate, we discovered that the key interaction in this process is the nonclassical C-H···O hydrogen bonds formed between the conjugate base of the CPA catalyst and the cationic intermediate. These C-H···O hydrogen bonds not only bind the catalyst to the substrates to form energetically favored states throughout the overall processes but also firmly maintain the relative positions of these fragments as the fixed contact ion pair to sustain the chiral information generated at the initial sulfur rearrangement step. This mechanistic case study provides a very clear understanding of the nature of the ion-pairing interaction in organocatalysis. The conclusion encourages the further development of the research field with the focus to design new organocatalysts and cultivate novel organocatalytic transformations.

Full text of DOI:10.1021/jacs.7b13274

Subscriber access provided by MT ROYAL COLLEGE
Article
Chiral Phosphoric Acid-Catalyzed Enantioselective Ring Expansion Reaction
of 1,3-Dithiane Derivatives: Case Study of the Nature of Ion-Pairing Interaction
Feng Li, Toshinobu Korenaga, Taishi Nakanishi, Jun Kikuchi, and Masahiro Terada
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b13274 • Publication Date (Web): 29 Jan 2018
Downloaded from http://pubs.acs.org on January 29, 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 free 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 accessible to all
readers and 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.
Journal of the American Chemical Society 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 16
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Chiral Phosphoric Acid-Catalyzed Enantioselective Ring Expansion
Reaction of 1,3-Dithiane Derivatives: Case Study of the Nature of
Ion-Pairing Interaction
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
Feng Li,† Toshinobu Korenaga,§ Taishi Nakanishi,† Jun Kikuchi,† Masahiro Terada*,†
† Department of Chemistry, Graduate School of Science, Tohoku University, Aoba-ku, Sendai 980-8578, Japan
§ Department of Chemistry and Biological Sciences, Faculty of Science and Engineering, Iwate University, 4-3-5 Ueda, Morioka 020-
8551, Japan
ABSTRACT: Chiral counter ion controlled asymmetric catalysis via an ion-pairing interaction has attracted immense attention in
recent years. Despite a number of successful studies, the mechanistic elucidation of the stereocontrolling element in the ion-pairing
interaction is rarely conducted and hence its nature is still far from being well understood. Herein we report an in-depth mechanistic
case study of a newly developed enantioselective ring expansion reaction of 1,3-dithiane derivatives catalyzed by chiral phosphoric
acid (CPA). An unprecedented enantioselective 1,2-sulfur rearrangement/stereospecific nucleophilic addition sequence was proven
to be the stereoselective pathway. More importantly, by thorough investigation of the intrinsic nature of the stereospecific nucleo-
philic addition to the cationic thionium intermediate, we discovered that the key interaction in this process is the non-classical C-
H···O hydrogen bonds formed between the conjugate base of the CPA catalyst and the cationic intermediate. These C-H···O hy-
drogen bonds not only bind the catalyst to the substrates to form energetically favored states throughout the overall processes, but
also firmly maintain the relative positions of these fragments as the “fixed” contact ion pair to sustain the chiral information gener-
ated at the initial sulfur rearrangement step. This mechanistic case study provides a very clear understanding of the nature of the
ion-pairing interaction in organocatalysis. The conclusion encourages the further development of the research field with an eye to
designing new organocatalysts and cultivating novel organocatalytic transformations.
INTRODUCTION
The chiral anion controlled asymmetric catalysis via an ion-
pairing interaction in organic synthesis has drawn much atten-
tion in recent years.1 Chiral phosphate anion, particularly the
one derived from the BINOL backbone (BINOL = 1,1’-bi-2-
naphthol), is one of the most widely used chiral anions in en-
antioselective transformations. It is the conjugate base of chi-
ral phosphoric acid (CPA), a powerful and versatile chiral
Brønsted acid catalyst used in numerous enantioselective reac-
tions (Figure 1).2 The activation mode in the CPA-catalyzed
reactions of electronically neutral substrates is well recognized
through the double hydrogen-bonding interaction: the
Brønsted acidic site (P-OH) interacts with an electrophile (e.g.,
imine, aldehyde, etc.) to form a hydrogen bond of O···H···X
(X = NR, O, etc.) and the Brønsted basic site of phosphoryl
oxygen (P=O) simultaneously forms a hydrogen bond with a
nucleophile (Figure 1a).2 In sharp contrast, a similar hydrogen
Figure 1. Activation mode of CPA-catalyzed reaction.
bond cannot be clearly identified in the ion-pairing interaction
due to the lack of an H-X moiety, an active proton attached to
a heteroatom, in cationic species (Figure 1b). In spite of the
large number of successful examples, mechanistic studies of
this type of reaction are rare. Our research group has disclosed
that the non-classical C-H···O hydrogen bonds between an
oxocarbenium ion and a chiral phosphate anion are crucial to
achieving a high level of stereocontrol in a CPA-catalyzed
Petasis-Ferrier-type rearrangement.3 To better understand the
nature of the ion-pairing interaction in asymmetric catalysis,
further studies of different types of stereoselective transfor-
mations involving the cationic species as a reactive intermedi-
ate are keenly required.
The thionium ion is a representative sulfur-stabilized carbo-
cation and the sulfur analog of oxocarbenium ion. It is widely
found as a reactive intermediate in a variety of reactions, such
as the Pummerer rearrangement and related reactions.4 How-
ever, the catalytic enantioselective reactions of thionium to
furnish enantio-enriched sulfur-containing compounds remain
largely unexplored.5 In this context, we envisioned that chiral
phosphate would be applicable as a counter anion of the thio-
nium intermediate to control the stereochemical outcome in an
enantioselective fashion. The method provides a new entry
into the catalytic asymmetric synthesis of enantio-enriched
sulfur-containing compounds.6 More importantly, the thionium
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 16
functionality possesses no H-X (X = sulfur) moiety to form states between the substrate and the catalyst throughout the
1
2
3
4
5
6
7
8
hydrogen bonds and hence, a thorough mechanistic study
would offer some insights into stereocontrolling elements in
the ion-pairing interaction.
overall processes, but also firmly maintain their relative posi-
tions to form a “fixed” contact ion pair. Hence, the chiral in-
formation generated at the initial sulfur rearrangement step
remains unchanged during the course of the addition step.
Herein we report our journey toward the discovery of this
novel enantioselective transformation and the results of our
comprehensive mechanistic case study.
To accomplish the proposed enantioselective transformation
involving the thionium intermediate, we focused our attention
on the 1,2-sulfur rearrangement of dithio-acetal or -ketal de-
rivatives with a leaving group at the -position (Scheme 1a).7
The conventionally assumed mechanism is as follows. At first,
a nucleophilic attack by sulfur atom initiates a substitution
reaction at the -carbon with the elimination of the leaving
group, affording a meta-stable episulfonium ion. The episul-
fonium ion undergoes ring opening immediately to generate a
thionium ion via the 1,2-sulfur rearrangement. Finally, the
thionium ion is trapped by a nucleophile to afford an addition
product or undergoes deprotonation to yield an alkene (elimi-
nation) product.
RESULTS AND DISCUSSION
Substituent effect of 1,3-dithiane derivative
After an extensive screening process, the reaction conditions
were optimized as follows:8 1,3-dithiane derivative 2 with a
trichloroacetimidate group at the -position was used as the
substrate.9 The reaction was performed with 1.5 equivalents of
pyrrole (3a) and 10 mol % of (R)-1a10 as CPA in dichloro-
methane at -40 ºC for 12 h.
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
Scheme 1. Reaction of thionium ion generated by 1,2-
sulfur rearrangement.
With the optimized reaction conditions in hand, the substit-
uent effect of the substrate was tested (Table 1). As expected,
the yield of product 4 was dependent on the electronic proper-
ty of substituent R1 at the 2-position of 1,3-dithiane 2: the sub-
stitution by an electron-donating group (EDG) resulted in a
higher yield than that by an electron-withdrawing group
(EWG) because of the stabilization of the cationic thionium
intermediate by EDG (Table 1, 4a-g).11 In contrast, it is note-
worthy that the enantioselectivities were comparable even
when the electronic property of aromatic substituent R1 was
changed, affording 4 with around 80% ee in all cases (4a-g).
The absolute configuration of the product was determined to
be (R) by the single-crystal X-ray diffraction analysis of 4f.12
Switching substituent R1 to an aliphatic n-propyl group gave
product 4h in moderate yield with a slight reduction of enanti-
oselectivity (74% ee). The introduction of geminal substituents
at the 5-position of 1,3-dithiane led to a significant increase in
enantioselectivity (4i and 4j). It should be underscored that the
substituents (R2 = R3 ≠ H) at the 5-position markedly influence
enantioselectivity compared to substituent R1 at the 2-position,
despite the fact that substituents R2 and R3 at the 5-position are
located far from the reaction site.
Table 1. Substituent effect of 1,3-dithiane derivatives 2.a
On the basis of conventional assumption,7 1,3-dithiane de-
rivative undergoes the 1,2-sulfur rearrangement via the activa-
tion of the leaving group by CPA catalyst to generate an ion
pair intermediate consisting of chiral phosphate and achiral
thionium ion (Scheme 1b). It is assumed that under the influ-
ence of chiral phosphate, the following nucleophilic addition
would take place in an enantioselective manner to afford an
enantio-enriched 1,4-dithiepane product. However, our mech-
anistic case study revealed that the reaction proceeds through
an unprecedented enantioselective 1,2-sulfur rearrange-
ment/stereospecific nucleophilic addition sequence (Scheme
1c). Notably, the present stereospecific addition is enabled by
the non-classical C-H···O hydrogen bonding interaction be-
tween the catalyst and the substrate. The C-H···O hydrogen
bonds not only allow the formation of energetically favored
ACS Paragon Plus Environment

Page 3 of 16
Journal of the American Chemical Society
diastereoselectivities for both 4k and 4l without marked loss
1
of enantioselectivity (Schemes 2a and 2b, condition 2).
2
3
Scheme 2. Reactions of 1,3-dithiane derivatives having
mono-substituent at 5-position.
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
Mechanistic study-1: Is nucleophilic addition the enan-
tio-determining step (EDS)?
In the course of our investigation of the substituent effect in
this reaction, we uncovered two intriguing features: 1) enanti-
oselectivity is markedly dependent on the substituent at the 5-
position cis/trans to the leaving group; 2) the reactions of 2k
and 2l yield different diastereomeric products with excellent dr
values. In other words, the reaction is stereospecific. These
results encouraged us to further investigate the reaction mech-
anism. The most important issue is whether an enantioselec-
tive addition to thionium takes place under the influence of a
chiral phosphate counter ion, as shown in Scheme 1b. To veri-
fy this issue, we conducted two control experiments.
a All reactions were carried out using 0.01 mmol of (R)-1a (10
mol %), 0.1 mmol of 1,3-dithiane derivative 2, and 0.15 mmol of
pyrrole 3a (1.5 equiv) in dichloromethane (1.0 mL, 0.1 M) at -
40 °C for 12 h.
In the first control experiment, we used a substrate having a
different leaving group. We hypothesized that if the addition
to thionium ion paired with chiral phosphate were the enantio-
determining step (EDS), the enantioselectivity of the product
would be the same irrespective of the leaving group. That is
because the leaving group is not involved in the nucleophilic
To further investigate the substituent effect at the 5-position
of substrate 2, cis/trans isomers 2k and 2l having a mono-
benzyl substituent were tested under the optimized reaction
conditions (Scheme 2).13 From geometrically single 2k, prod-
uct 4k was obtained in good yield with perfect diastereoselec-
tivity and excellent enantioselectivity (Scheme 2a, condition
1). The absolute configuration of 4k was unambiguously de-
termined to be (2R, 6S) by single-crystal X-ray diffraction
analysis.12 From geometrically single 2l, diastereomer 4l was
obtained as the sole product.14 However, the reactivity of 2l
was much lower than that of 2k; the conversion of 2l was only
56% after 12 h (Scheme 2b, condition 1). Moreover, the enan-
tiomeric excess of 4l (79% ee) was much lower than that of 4k
(97% ee) and was identical to that of 4a (79% ee) obtained
from 2a that had no substituent at the 5-position of 1,3-
dithiane (Table 1). These results clearly indicate that the sub-
stituent at the 5-position cis to the leaving group significantly
enhances the enantioselectivity, whereas the substituent trans
to the leaving group has nearly no effect on the enantioselec-
tivity. Further optimization of the reaction conditions by using
molecular sieve (MS) 4A resulted in nearly perfect yields and
addition step. With this hypothesis in mind, N-phenyl tri-
5b
fluoroacetimidate
was selected as the leaving group. The
reaction of 5d12,15 having an N-phenyl trifluoroacetimidate
moiety was conducted at room temperature for 4 h.16 As a
result, (S)-4d was obtained in 17% yield with 11% ee along
with alkene byproduct 6d (Scheme 3a). In contrast, substrate
2d was completely consumed within 1 h under the same con-
ditions and (R)-4d was obtained in 88% yield with 44% ee
(Scheme 3b). 4d enantiomers with different enantioselectivi-
ties were the major products obtained from 2d and 5d. These
results strongly suggest that nucleophilic addition to thionium
is not the EDS of this reaction.
Scheme 3. Reactions of 1,3-dithiane derivatives having
different leaving groups.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 16
1
2
3
4
5
6
Considering the stereochemical relationship between 2k and
4k as well as 2l and 4l, it is clear that the nucleophilic addition
of pyrrole 3a to thionium proceeds exclusively on the same
side of the cyclic framework with the leaving group in a stere-
ospecific manner. Consequently, the direction of the addition
reaction is unambiguously determined by the location of the
leaving group. The above consideration coupled with the re-
sults obtained by changing the leaving group (Scheme 3) leads
to a solid outline of the overall process: the addition step is not
EDS and hence the initial step, 1,2-sulfur rearrangement,
should be EDS (Scheme 6). Indeed, the estimated outline is
consistent with the experimental evidence that enantioselectiv-
ity changes dramatically when the substrate leaving group is
switched, because the leaving group is involved in the initial
step. 1,3-Dithiane substrates 2 are achiral molecules; however,
the two sulfur atoms are enantiotopic (shown in red and blue).
Therefore, it can be proposed that CPA differentiates these
two enantiotopic sulfur atoms and the rearrangement of one
sulfur atom (shown in red) is more favorable than that of the
other (shown in blue). After the selective sulfur rearrangement,
two reactive and “well-ordered” intermediates are generated
with one being the major intermediate. Subsequent addition to
these intermediates proceeds in a stereospecific manner, and
both major and minor intermediates are transformed into the
corresponding major and minor products. The ratio of the two
intermediates should be equal to the enantiomeric ratio of final
product 4.
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
In the second control experiment, we performed the reac-
tions using 2k and 2l in the presence of achiral acid catalyst 1b
without a substituent at the ortho-position of the phenol group.
We hypothesized that if the ion pair depicted in Scheme 1b
were the plausible intermediate, an identical ion pair would be
generated from 2k and 2l using 1b and this would result in the
formation of products with the same dr value (Scheme 4, Path
I). Otherwise, as observed in the reaction using 1a, 4k and 4l
would be formed from 2k and 2l, respectively, in a stereospe-
cific manner, regardless of the steric congestion around the
phosphoric acid functionality (Scheme 4, Path II).
Scheme 6. Proposed revised mechanism with CPA.
Scheme 4. Plausible mechanisms with achiral catalyst.
The reactions of 2k and 2l were carried out with 10 mol %
of achiral phosphoric acid 1b and 1.5 equivalents of pyrrole 3a
in dichloromethane at 0 ºC for 2 h. As shown in Scheme 5,
racemic 4k and 4l were formed from 2k and 2l, respectively,
in high yields with excellent dr values. These results clearly
indicate that the substituents at the 3,3’-positions of BINOL-
derived CPA are not essential for the specific formation of 4k
and 4l in excellent diastereoselectivity. More importantly, the
ion pair depicted in Scheme 1b (see also Scheme 4, Path I),
which was assumed from a number of previous reports,7 is not
suited as the plausible intermediate in this reaction. The reac-
tions of 2k and 2l should proceed via “well-ordered” interme-
diates because of the stereospecific process (Scheme 4, Path
II).
Mechanistic study-2: Reaction with different nucleo-
philes.
In the previous section, it was proposed that the 1,2-sulfur
rearrangement step should be EDS. It can be considered that in
principle, enantioselectivity should be determined irrespective
of the nucleophile employed, because only the substrate and
CPA are involved in this step (Scheme 6). On the basis of this
working hypothesis of EDS, coupled with the subsequent ste-
reospecific addition, products with the same ee value should
be obtained even when different nucleophiles are used
(Scheme 7).
Scheme 7. Hypothesis: different nucleophiles give products
with the same ee.
Scheme 5. Reactions using achiral catalyst.
ACS Paragon Plus Environment

Page 5 of 16
Journal of the American Chemical Society
of 4k (Scheme 2a: 96% ee), whereas 7l (Scheme 8b: 73% ee)
1
2
3
4
5
6
7
8
had an obviously lower ee than 4l (Scheme 2b: 78% ee).
Scheme 8. Reactions of 2k and 2l with indole.
To validate our hypothesis, indole (3b) was employed as the
nucleophile because of the similarity in reactivity of indole
(3b) and pyrrole (3a). As shown in Table 2, the reaction of
indole (3b) (1.5 equiv.) was performed with five 1,3-dithiane
derivatives 2 using 5 mol % of optimized CPA catalyst (R)-1a
at -40 ºC for 6 h. Analogous products 7 were obtained in ex-
cellent yields. However, it is noteworthy that the enantiomeric
excess of all the five products 7 was obviously lower than that
of the corresponding products 4 derived from pyrrole (3a)
(Table 2 vs. Table 1). This result is not consistent with our
hypothesis that products with the same enantioselectivity
would be formed even if the nucleophile was changed, as
shown in Scheme 7.
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
From the results of these diastereomeric products as well as
those shown in Table 1 [pyrrole (3a)] and Table 2 [indole
(3b)], it can be concluded that enantioselectivity is dependent
on the nucleophile employed even though the reaction pro-
ceeds in a stereospecific manner. It implies that further
amendment of the mechanism proposed in Scheme 6 is neces-
sary to rationalize the stereochemical outcome governed by
nucleophiles. A plausible mechanism is that the nucleophile is
involved in EDS, because the nucleophile is the only differing
factor in these reactions (Scheme 9). Based on the dual activa-
tion mode of CPA (Figure 1a), it is considered that the
Brønsted acidic site of CPA activates the leaving group, tri-
chloroacetimidate, in the initial step of this reaction. Mean-
while, a hydrogen bond is formed between the Brønsted basic
site of CPA, the phosphoryl oxygen, and N-H moiety of pyr-
role or indole nucleophile (Scheme 9), which would stabilize
the phosphate anion generated by the protonation of the leav-
ing group. Subsequently, the 1,2-sulfur rearrangement occurs
under the chiral environment created by CPA that bears the
nucleophile through the hydrogen bond.17 This would rational-
ize why pyrrole and indole gave products with different enan-
tioselectivities.
Table 2. Reactions of 1,3-dithiane derivatives with indole.a
Scheme 9. Further amended mechanism and proposed
EDS.
a All reactions were carried out using 0.005 mmol of (R)-1a (5
mol %), 0.1 mmol of 1,3-dithiane derivative 2, and 0.15 mmol of
indole (3b) (1.5 equiv) in dichloromethane (1.0 mL, 0.1 M) at -
40 °C for 6 h. Ee values of corresponding 4 are shown in paren-
theses.
Before discussing this contradiction, to ascertain that the
nucleophilic attack by indole (3b) is stereospecific, we con-
ducted the reactions using substrates 2k and 2l under the opti-
mized reaction conditions. As shown in Scheme 8, products
7k and 7l were formed from 2k and 2l, respectively, in excel-
lent yields with perfect diastereoselectivity and hence the reac-
tion of indole (3b) is also stereospecific.14 In addition, the ee
value of 7k was much higher than that of 7l, similar to that
observed in the reaction of pyrrole (3a) (Scheme 2). Further
analysis of these diastereomeric products showed that the ee
value of 7k (Scheme 8a: 95% ee) was slightly lower than that
Mechanistic study-3: reaction without nucleophile or
with N-Me-pyrrole and N-Me-indole as nucleophile
As proposed in the previous section, pyrrole and indole are
involved in EDS by forming a hydrogen bond with phosphoryl
oxygen. In order to verify this interaction,17 we excluded the
hydrogen bond from the reaction system in which reactions
ACS Paragon Plus Environment

Journal of the American Chemical Society
were performed without a nucleophile or by using an N-
Page 6 of 16
1
methyl substituted pyrrole and indole as the nucleophile.
Based on our examination of previous reports,7 we expect
that a ring expansion/deprotonation product would be formed
in the absence of a nucleophile (Scheme 1a). The reaction of
2k or 2l in the absence of nucleophile affords the correspond-
ing elimination product in an enantioselective manner because
of the generation of the stereogenic center at the 6-position,
which reflects the chiral information of EDS. The enantiose-
lectivity of these products may offer some clues to help us
determine to what extent pyrrole and indole affect the enanti-
oselectivity of EDS.
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
The reaction of 2k and 2l in the absence of nucleophile was
carried out using 10 mol % of optimized catalyst (R)-1a and
MS 4A at -40 ºC for 12 h. As shown in Scheme 10, ring ex-
pansion/deprotonation product 6k was obtained from 2k and
2l with 96% ee (S) and 79% ee (R), respectively.14 In the reac-
tion of 2k, the ee value of (S)-6k (96% ee) is the same as that
of 4k (96% ee) and slightly higher than that of 7k (95% ee)
(Schemes 10a, 2a, and 8a). Likewise in the reaction of 2l, up-
on comparing the enantioselectivities of (R)-6k (79% ee), 4l
(78% ee), and 7l (73% ee) (Schemes 10b, 2b, and 8b), it be-
came clear that the addition of indole (3b) resulted in an obvi-
ous decrease in enantioselectivity. It can be concluded that
indole (3b) influences the ee values, whereas pyrrole (3a) has
little effect.
The yields of ring expansion/addition products N-Me-4a
and N-Me-7a were low in the reactions with N-methylpyrrole
(3c) and N-methylindole (3d), respectively (Scheme 11). In
sharp contrast, pyrrole (3a) and indole (3b) underwent the ring
expansion/addition reaction in excellent yields without the
formation of ring expansion/deprotonation product 6 (Tables 1
and 2). These observations contradict the idea that, in principle,
a more reactive nucleophile should give a ring expan-
sion/addition product in higher yield, because 3c and 3d are
more nucleophilic than 3a and 3b, respectively.20 The results
obtained using N-methyl derivatives 3c and 3d clearly reveal
that the N-H moiety of 3a and 3b is the key to achieving high
chemoselectivity. It is considered that the N-H moiety of these
nucleophiles stabilizes the chiral phosphate anion generated
throughout the ring expansion/addition process. The hydrogen
bonding interaction between the N-H moiety and the oxygen
atom of the phosphate anion is probably essential to circum-
vent the formation of ring expansion/deprotonation product 6
because of a decrease in basicity of the phosphate anion.
Scheme 10. Reactions of 2k and 2l without nucleophile (H-
Nu).
Possible structures of “well-ordered” reactive interme-
diate
From the experimental studies demonstrated above, we con-
clude that this reaction proceeds through an enantioselective
1,2-sulfur rearrangement/stereospecific nucleophilic addition
sequence and a nucleophile, pyrrole or indole, is involved in
the initial 1,2-sulfur rearrangement step.21 However, the struc-
ture of the “well-ordered” reactive intermediate and the intrin-
sic nature of the stereospecific nucleophilic addition have re-
mained unclear.
The reactions with N-methylpyrrole (3c) and N-
methylindole (3d) were also carried out (Scheme 11). In these
cases, ring expansion/addition products N-Me-4a18 and N-Me-
7a were obtained in 23% and 32% yields, respectively and
ring expansion/deprotonation product 6a was the major prod-
uct. The ee values of N-Me-4a and N-Me-7a are not suitable
for mechanistic discussion, because the existence of two com-
peting reaction pathways would probably lead to the partial
resolution of the enantio-enriched (“well-ordered”) reactive
intermediate under the influence of the chiral catalyst.19
Three pathways are theoretically conceivable to fulfill the
requirements of the stereospecific addition (Scheme 12).
[1] Path A: As the 1,2-sulfur rearrangement with the elimi-
nation of a leaving group is well recognized as an SN2 process,
episulfonium intermediate INT-A should be firstly generated.
The product is formed from INT-A through SN2-like transition
state TS-A.
[2] Path B: Episulfonium intermediate INT-A undergoes
ring opening to form INT-B immediately.22 Meanwhile, thio-
nium ion and phosphate anion maintain their relative positions
in INT-B, namely, the formation of a “fixed” contact ion pair.
Phosphate stays on one side of the thionium basal plane and
does not change its original position to the other side. It is
entirely different from the ion-pair structure presumed initially
because the relative positions of thionium and phosphate are
not restricted in the conventional assumption (Scheme 1b).
Then, phosphate anion directs a nucleophilic attack from the
same side on the thionium basal plane, affording the product
through TS-B in a stereospecific manner.
Scheme 11. Reactions with N-methylpyrrole and N-
methylindole.
ACS Paragon Plus Environment

Page 7 of 16
Journal of the American Chemical Society
[3] Path C: By directly transforming from INT A or through
tained in the reaction of 2k even by changing the acidity of the
catalyst and the polarity of the solvent employed, whereas in
the reaction of 2l the stereospecificity is highly dependent on
these parameters. The reaction of 2l using trifluoroacetic acid
(TFA) (pKa 3.37 in DMSO),25 of which acidity is slightly
higher than that of phosphoric acid 1b (pKa 3.72 in DMSO),25
resulted in a decrease in stereospecificity to afford a diastere-
omeric mixture of 4k and 4l in 7:93 (Scheme 13a). The reduc-
tion of stereospecificity was also observed in the reaction of 2l
using 1b in polar acetonitrile, instead of less polar dichloro-
methane, affording 4k and 4l in 10:90 (Scheme 13b). En-
hancement of the catalyst acidity and the solvent polarity re-
sulted in the reduction of stereospecificity in the reaction of 2l.
1
2
3
4
5
6
7
8
INT-B, the generation of covalently bonded phosphate inter-
mediate INT-C10c,23 is also considered possible by the addition
of phosphate to episulfonium or thionium. Subsequent syn
displacement of phosphate by the nucleophile from INT-C
affords the product through concerted asynchronous transition
state TS-C, namely, the SNi reaction pathway.24
Scheme 12. Possible intermediates and reaction pathways.
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
Scheme 13. Effects of acids and solvents.
It should be pointed out that the reaction of 2k proceeded in
a stereospecific manner even by changing the acid catalyst and
the solvent used. These results suggest that these parameters
do not significantly affect the reaction mechanism when using
2k as the substrate. It would be postulated that the reaction
mechanism is maintained even by using 2l, instead of 2k, un-
der these reaction conditions. On the basis of this postulation,
Path A is unlikely because SN2-like transition state TS-A is the
origin of the stereospecificity in Path A, which is initiated
from stereochemically well-defined episulfonium intermediate
INT-A. Even if the catalyst acidity and the solvent polarity
were changed, the stereospecificity originated from this as-
sumed mechanism should be preserved in Path A. However,
this is not consistent with the experimental observations in the
reactions of 2l. Furthermore, the experimental results, coupled
with the NMR monitoring data, imply that Path C is also un-
likely on the basis of the above postulation. In Path C, in prin-
ciple, the stereospecificity is presumably preserved even by
changing the reaction conditions because it would originate
from the SNi reaction of stereochemically defined intermediate
INT-C via transition state TS-C. INT-C would be generated
from phosphate and episulfonium or thionium through the
formation of a configurationally stable covalent bond that
should aid the stereospecific process. However, the decrease in
stereospecificity was observed in the reactions of 2l. In addi-
tion, intermediate INT-C could not be observed clearly
throughout the NMR monitoring process. Therefore, Path C is
As mentioned in the previous section, the hydrogen bonding
interaction between phosphate and nucleophile (H-Nu) shown
in all the proposed intermediates not only decreases the basici-
ty of phosphate but also increases the nucleophilicity of N-
heterocyclic nucleophiles. Thus, the elimination is decelerated
whereas the nucleophilic addition is accelerated.
To identify the structure of the “well-ordered” reactive in-
termediate, we performed NMR analyses and experimental
studies by using achiral acid catalyst 1b or other acid catalysts
and changing the reaction solvents. At the outset, the reaction
of 2a with or without 3a in the presence of 1b (20 mol %) was
1
monitored by H and 31P NMR spectroscopy (Figures S1 and
S2). In the presence of 3a, the gradual consumption of both
reactants and the generation of ring expansion/addition prod-
uct 4a were observed and no other side products or intermedi-
ates, such as phosphate ester derivative INT-C, were detected
1
by H NMR measurement (Figure S1). Similarly, in the ab-
sence of 3a, the 31P NMR spectra displayed one slightly broad
peak presumed to be phosphoric acid 1b (Figure S2). No other
phosphorus peaks were observed during the measurement,
despite the fact that the phosphate anion would be one of the
most nucleophilic species in the reaction system. We also con-
ducted the reaction of 2k and 2l with 3a using achiral acid
catalysts with different acidities and in different solvents
(Scheme 13: also see Tables S4–S6). The results indicate that
high stereospecificity of the nucleophilic addition step is main-
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 16
probably not the plausible pathway, although the concerted SNi
sulfur atom than the equatorial one, although these conforma-
tional changes eventually lead to the destabilization of these
transition states.
reaction is mechanistically intriguing.24
1
2
3
4
5
6
7
8
Based on the above considerations, the “fixed” contact ion
pair of ring-expanded thionium ion and phosphate anion is the
best candidate for the “well-ordered” reactive intermediate
(Scheme 12, INT-B) and hence Path B is the plausible path-
way for the stereospecific addition process. However, experi-
mental evidence is not sufficient to completely rule out other
possible pathways. Therefore, we further conducted DFT cal-
culation to understand the reaction mechanism and identify the
“well-ordered” reactive intermediate and the stereospecific
reaction pathway.
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
DFT Calculation: model study
To simplify the calculation, phosphoric acid 1c having a bi-
phenol backbone as well as standard substrate 2a and pyrrole
(3a) were used in the model study (Scheme 14).17,21,26 All cal-
culations were performed with the Gaussian 09 package (Re-
vision C.01).27 Geometries were optimized and characterized
using frequency calculations at the B97D/6-31G(d,p) level,
unless otherwise noted.28,29 Energies in the solution phase were
calculated using single-point energy calculations at the same
level according to the CPCM solvation model for the opti-
mized structures.30
As 1c has an axially chiral structure [(R)-isomer], the gener-
ation of major enantiomer (R)-4a was calculated as observed
in the experimental results. From the mechanistic studies
shown above, we added pyrrole in the calculation from the
initial 1,2-sulfur rearrangement step even for the model studies.
Scheme 14. Model system for calculation.
The reaction is known to be initiated by a nucleophilic at-
tack by one of the sulfur atoms of 1,3-dithiane and the elimi-
nation of the leaving group via the SN2 mechanism.7,31 In cal-
culating the transition states of this initial step, the orientation
of the leaving group and the location of phosphoric acid cata-
lyst were carefully considered. The trichloroacetimidate group
would adopt an axial or equatorial orientation. In the mean-
time, phosphoric acid would locate in the proximity of 1,3-
dithiane or far from it. Considering various combinations of
these orientations, four possible transition states were calculat-
ed.32 As shown in Figure 2a, TS-1-axial-in, in which the leav-
ing group is located at the axial position and phosphoric acid
is close to 1,3-dithiane, turned out to be the most favorable
transition state. In contrast, TS-1-axial-out is energetically
disfavored even though the steric congestion between phos-
phate and 1,3-dithiane is avoided. More interestingly, as ob-
served in TS-1-equatorial-in and TS-1-equatorial-out, the
optimization of transient structures, which was performed
from the initial structures having the leaving group at the
equatorial position, enforced a conformational change in 1,3-
dithiane from the chair form to the boat-like one with the leav-
ing group occupying the axial position (Figure 2a). These re-
sults clearly indicate that the axial orientation of the leaving
group is much more favorable for the nucleophilic attack by
Figure 2. Transition states for sulfur rearrangement. Geometries
were optimized and characterized using frequency calculations at
the B97D/6-31G(d,p) level. Relative free energies (kcal/mol)
obtained by single-point energy calculations at the same level for
the optimized structures with the SCRF method based on CPCM
(ϵ = 8.93 for CH2Cl2) are shown.
The 3D structure of energetically most favorable TS-1-
axial-in is shown in Figure 2b. Phosphoric acid 1c protonates
the nitrogen atom of the leaving group to form a hydrogen
bond and pyrrole also forms a hydrogen bond with phosphoryl
oxygen. More importantly, four non-classical C-H···O hydro-
gen bonds are also clearly observed between 1,3-dithiane and
ACS Paragon Plus Environment

Page 9 of 16
Journal of the American Chemical Society
imate intermediate. However, subsequent analysis of the in-
phosphate anion. These C-H···O hydrogen bonds, namely,
electrostatic interaction,3,33,34 stabilize the transition structures
of partially generated phosphate anion, the location of which is
regulated in proximity to 1,3-dithiane by these interactions.
This orientation results in the efficient reduction of the degree
of charge separation in this transition state, albeit causing ste-
ric congestion between these partially charged moieties.
trinsic reaction coordinate (IRC) from TS-1-axial-in revealed
the formation of ion-pair intermediate INT-1 consisting of
ring-expanded thionium ion, phosphate anion, and trichloroa-
cetamide as the next meta-stable intermediate. The IRC analy-
sis clearly indicates that episulfonium undergoes ring opening
immediately, even if episulfonium would form with a small
energy barrier, and thus it should not be the intermediate for
the nucleophilic addition (Scheme 12: Path A is ruled out). A
clue for the immediate ring expansion is also obtained from
the bond length data of TS-1-axial-in (Figure 2b). The S1-C2
bond elongates to 1.93 Å, thus becoming weaker, whereas the
S2-C2 bond measures 1.82 Å, which is in the normal range of a
S-C single bond.35
1
2
3
4
5
6
Based on the calculations shown in Figure 2, the difference
in reactivity between substrates 2k and 2l is well rationalized
(Scheme 15). The experimental results revealed that 2k was
more reactive than 2l (Scheme 2, condition 1: 100% conver-
sion of 2k, 56% conversion of 2l). As shown in Figure 2a, the
sulfur rearrangement is more favorable for the leaving group
located at the axial position than for that located at the equato-
rial position. 2k-ax and 2l-ax should be the reactive conform-
ers for 2k and 2l, respectively (Scheme 15). However, 2k-ax
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
From INT-1, after trichloroacetamide is cleaved off, more
stable contact ion-pair intermediate INT-2 is generated and is
consistent with INT-B (Scheme 12). Followed by a small en-
ergy barrier, 4.3 kcal/mol, nucleophilic addition takes place
smoothly to afford C-C bond formation intermediate INT-4
via TS-2. After a proton shift to aromatize, addition product
4a complexed with 1c is formed. The formation of a C-O bond
from INT-2 affords covalent bonded phosphate INT-3, con-
sistent with INT-C (Scheme 12), and INT-3 is energetically
disfavored compared to INT-2. The energy level of INT-3 is
comparable to that of TS-2. Even when pyrrole is removed,
adduct INT-3’ is still less stable than INT-2. This indicates
that phosphate adducts INT-3’ and INT-3 should not be the
reactive intermediate in this reaction (Scheme 12: Path C is
ruled out).
1
and 2l-eq were the favored conformers on the basis of the H
NMR analysis of 2k and 2l, respectively, involving NOE ex-
periments (based on coupling constants and NOE, see NMR
spectra of 2k and 2l in SI for details). A conformational
change is required of 2l from thermodynamically stable 2l-eq
to less stable 2l-ax, because the axial orientation of the leaving
group is necessary for the sulfur rearrangement. In contrast, no
conformational change is necessary for thermodynamically
favorable 2k-eq and hence 2k exhibits higher reactivity than
2l. It is therefore considered that the sulfur rearrangement
would be the rate-determining step (RDS) of the overall reac-
tion. The results of NMR monitoring showed that no other
species were observed except the starting material, the catalyst,
and the product (Figures S1 and S2) also suggests that the
initial sulfur rearrangement is RDS.17 In order to confirm
whether the sulfur rearrangement is RDS or not and to further
elucidate the “well-ordered” reactive intermediate, we clarified
the energy profile of this reaction.
The energy profile of the overall processes clearly reveals
that the ion pair of ring-expanded thionium and phosphate
INT-2 (= INT-B) can be regarded as the “well-ordered” reac-
tive intermediate. In order to understand how and why the
stereospecific addition proceeded via this intermediate, we
carefully analyzed the 3D structures of INT-2 and the transi-
tion state of nucleophilic addition, TS-2. As shown in Figure
4a, in INT-2, five C-H···O hydrogen bonds are formed be-
tween thionium and phosphate, similar to that observed in TS-
1-axial-in (Figure 2b). In addition, phosphate directs pyrrole
to locate in close proximity through the O···H···N hydrogen
bond and hence phosphate as well as pyrrole is localized on
the same side of the thionium framework.
Scheme 15. Possible reason for the difference in reactivity
between 2k and 2l.
The structure of TS-2 (Figure 4b) is apparently the same as
that of INT-2 (Figure 4a) except for the slight variation of the
relative positions between these ternary fragments. Most of the
key interactions in INT-2 are observed in TS-2 with a minor
change of the interaction pattern of the multiple C-H···O hy-
drogen bonds between thionium and phosphate. Besides these
hydrogen bonding interactions, -hyperconjugation, *(S1-C)-
(C-C)formation, should also be taken into consideration. TS-2 is
stereo-electronically preferred for the nucleophilic attack on
thionium, because the forming C-C bond is stabilized by the
antiperiplanar effect of the S1-C bond (Figure 4b).36
The energy profile is shown in Figure 3. The largest energy
barrier is the initial step, namely, the 1,2-sulfur rearrangement
with elimination of the leaving group (TS-1-axial-in), and
hence this step is indeed RDS. After this transition state, in
principle, the episulfonium ion should be formed as the prox-
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 16
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
Figure 3. Energy profile of model reaction. The potential energy for the sum of 1c, 2a, and 3a was set to zero. Geometries were optimized
and characterized using frequency calculations at the B97D/6-31G(d,p) level. Energies (kcal/mol) in solution phase were calculated using
single-point energy calculations at the same level for the optimized structures according to the SCRF method based on CPCM (CH2Cl2).
Figure 4. 3D models of INT-2 and TS-2.
Three factors: (i) C-H···O hydrogen bonds between thioni-
um and phosphate, (ii) phosphate anion directing nucleophile
through a hydrogen bond, and (iii) -hyperconjugation, partic-
ipate synergistically to establish this highly stereospecific ad-
dition process. As regards the origin of this stereospecific ad-
dition, it should result from the location of the phosphoric acid
catalyst in the 1,2-sulfur rearrangement. This is because C-
H···O hydrogen bonds are formed between phosphoric acid 1
and 1,3-dithiane substrate 2 in TS-1-axial-in, the transition
state of the sulfur rearrangement, in which phosphoric acid is
in proximity to 1,3-dithiane on the same side of the axially
oriented leaving group (Figure 2b). Phosphoric acid/phosphate
and 1,3-dithiane/thionium are tethered by the C-H···O hydro-
gen bonds from the initial sulfur rearrangement step and hence
the relative positions of these fragments do not change from
the 1,2-sulfur rearrangement, ring expansion, and elimination
ACS Paragon Plus Environment

Page 11 of 16
Journal of the American Chemical Society
of the leaving group steps to the nucleophilic addition step. It
can be concluded that the ternary association of phosphoric
1
acid, substrate, and nucleophile through the C-H···O and
N···H···O hydrogen bonds is the origin not only of the “well-
ordered” reactive intermediate but also of the stereospecific
addition.37
2
3
4
5
As the ternary association in the initial sulfur rearrangement
step is the origin of the stereospecific addition, it is surmised
that the use of (E)-N-phenyltrifluoroacetimidate as the leaving
group would also allow the reaction to proceed in a stereospe-
cific manner (Figure 5). That is because the (E)-geometry of
the N-phenyl group should form a transient structure similar to
that observed in the substrate having the trichloroacetimidate
group, albeit resulting in low reactivity and enantioselectivity
as expected from Scheme 3.
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
DFT calculation of enantio-determining step (EDS): Re-
al system
The reaction pathway and the reactive intermediate were
clarified on the basis of experimental studies and model calcu-
lations. We next turned our attention to elucidation of the
mechanism underlying EDS, in particular, the intriguing sub-
stituent effect at the 5-position of 1,3-dithiane. The substituent
at the 5-position cis to the leaving group of 1,3-dithiane sub-
strate 2k enhances the enantioselectivity markedly, as com-
pared with 2l having a substituent trans to the leaving group
(Table 1 and Scheme 2). In order to gain an insight into the
origin of the stereochemical outcome as well as the substituent
effect at the 5-position of 1,3-dithiane, we calculated EDS,
namely, the 1,2-sulfur rearrangement step, using substrate 2k
and catalyst (R)-1a instead of simplified acid catalyst 1c. To
expedite our calculation, no nucleophile is added because
based on our experimental result with or without using pyrrole
as the nucleophile, the enantioselectivity of ring-
expansion/addition product 4k is identical to that of ring-
expansion/deprotonation product (S)-6k (Schemes 2a and 10a).
This implies that the structures of these transition states are
fundamentally similar in EDS, irrespective of the presence or
absence of pyrrole.
Figure 5. Proposed transition states of 1,2-sulfur rearrangement
of substrates with different leaving groups.
With this hypothesis in mind, the reaction of dithiane 5l
having (E)-N-phenyltrifluoroacetimidate with pyrrole 3a (1.5
equiv.) was conducted using 10 mol % of optimized catalyst
(R)-1a and MS 4A in dichloromethane at room temperature
for 4 h (Scheme 16a).16 Substrate conversion was only 34%
and desired product (2S,6S)-4l was obtained in 25% NMR
yield with 98:2 dr and 15% ee. In comparison with 5l, the re-
action of 2l under the same conditions proceeded smoothly
and complete conversion was achieved within 1 h (Scheme
16b). Product (2R,6R)-4l was obtained in 81% NMR yield
with 99:1 dr and 51% ee. In both cases, ring expan-
sion/deprotonation product (R)-6k was obtained as the by-
product. These results verify our hypothesis that the reaction
using (E)-N-phenyltrifluoroacetimidate as the leaving group is
also stereospecific, although the reactivity was much lower
than that of the substrate having the trichloroacetimidate group.
Once again, the different enantioselectivities of these two re-
actions clearly reveal that EDS should be the 1,2-sulfur rear-
rangement instead of the following nucleophilic addition
(Scheme 9).
The computed transition states for both major and minor re-
action pathways of the reaction using 2k are shown in Figure 6.
The energy difference between these two transition states is
2.6 kcal/mol with the one resulting in (S)-6k being the ener-
getically favored one. The result of the calculations is con-
sistent with the experimental finding that (S)-6k is the major
enantiomer in the reaction of 2k catalyzed by (R)-1a (Scheme
10), although the expected enantiomeric excess calculated
from the energy difference does not perfectly match the exper-
imental result.38
Scheme 16. Stereospecificity of reactions using substrates
with different leaving groups.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 16
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
Figure 6. Transition states of EDS of 2k with catalyst (R)-1a. Geometries were optimized and characterized using frequency calculations
at the B97D/6-31G(d,p) level. Relative free energies (kcal/mol) obtained by single-point energy calculations at the same level for the opti-
mized structures with the SCRF method based on CPCM (CH2Cl2) are shown.
By carefully analyzing the key interactions in these two
transition states, three non-classical C-H···O hydrogen bonds
were observed in not only TS-major-2k but also TS-minor-
2k. This indicates that these C-H···O hydrogen bonds formed
between the substrate and the catalyst are vital to determine
the relative positions accurately and hence these interactions
are the key to differentiating the major and minor pathways. In
fact, the relative arrangements of 2k and (R)-1a are absolutely
different in these two transition states. In TS-minor-2k, 2k is
oriented from side to side as shown in Figure 6b, where the
phenyl group is located close to one of the bulky triphenylsilyl
groups, resulting in steric repulsion. In addition, the steric
congestion between the benzyl group introduced at the 5-
position and the other triphenylsilyl group further destabilizes
this transition state. In contrast, in TS-major-2k (Figure 6a),
substrate 2k is placed back and forth to avoid steric congestion
between the two triphenylsilyl groups and the phenyl group
substituted at the 2-position. Furthermore, the benzyl group
introduced at the 5-position is oriented in such a way as to
avoid steric repulsion with the triphenylsilyl group and further
stabilize this transition state through the T-shape C-H··· in-
teraction39,40 with the phenyl ring of the triphenylsilyl group.
The substituent effect at the 5-position of 1,3-dithiane is also
well rationalized based on these transition structures. In the
substrates having the substituent cis to the leaving group,
namely, 2k as well as 2i and 2j, the substituent occupies the
equatorial position of 1,3-dithiane to induce repulsive interac-
tion with the triphenylsilyl group in the minor transition state,
as likely shown in Figure 6b, resulting in a marked increase in
enantioselectivity (Table 1). In contrast, in the reaction of 2l
having a substituent trans to the leaving group, steric conges-
tion does not occur between the triphenylsilyl group and the
substituent in the minor transition state, because the substitu-
ent is located in the axial position and exhibits little effect on
the enantioselectivity.
CONCLUSION
We have reported herein our in-depth mechanistic case
study of a newly developed CPA-catalyzed enantioselective
ring expansion reaction of 1,3-dithiane derivatives. Experi-
mental studies were well-designed to elucidate that an unprec-
edented
enantioselective
1,2-sulfur
rearrange-
ment/stereospecific nucleophilic addition sequence is the con-
crete mechanism. This discovery is completely different from
conventionally assumed two-step reactions. Moreover, a nu-
cleophile is involved not only in the nucleophilic addition step,
but also in the initial 1,2-sulfur rearrangement step, even
though it does not directly react with the substrate in the initial
step.
The structure of the “well-ordered” reactive intermediate
and the intrinsic nature of the stereospecific nucleophilic addi-
tion were identified by DFT calculation using a simplified
model, where non-classical C-H···O hydrogen bonds formed
between the phosphoric acid catalyst and the substrate were
found to be the key interaction in this reaction. These hydro-
gen bonds not only linked the catalyst with the substrate to
form energetically favored states, but also firmly maintained
their relative positions and stored chiral information as the
“fixed” contact ion pair in the reactive intermediates during
the overall processes.
The origin of the enantio-control by CPA was also elucidat-
ed by DFT calculation of the real reaction system. In EDS,
non-classical C-H···O hydrogen bonds were involved in the
key interaction. Moreover, it was confirmed that the intriguing
substituent effect at the 5-position of 1,3-dithiane originated
from the interaction between the substituent at the 5-position
ACS Paragon Plus Environment

Page 13 of 16
Journal of the American Chemical Society
(4) Selected recent reviews of thionium mediated reactions, see: (a)
cis to the leaving group and the triphenylsilyl groups at the
Bur, S. K.; Padwa, A. Chem. Rev. 2004, 104, 2401-2432. (b) Feldman,
K. S. Tetrahedron 2006, 62, 5003-5034. (c) Smith, L. H. S.; Coote, S.
C.; Sneddon, H. F.; Procter, D. J. Angew. Chem., Int. Ed. 2010, 49,
5832-5844.
(5) Examples of chiral anion controlled asymmetric catalysis using
another sulfur-stabilized carbocations, see: (a) Hamilton, G. L.; Kanai,
T.; Toste, F. D. J. Am. Chem. Soc. 2008, 130, 14984-14986. (b) Lin,
S.; Jacobsen, E. N. Nat. Chem. 2012, 4, 817-824.
(6) Recent comprehensive review of asymmetric catalytic synthesis
of optically active sulfur compounds having quaternary stereogenic
center, see: Yu, J.-S.; Huang, H.-M.; Ding, P.-G.; Hu, X.-S.; Zhou, F.;
Zhou, J. ACS Catal. 2016, 6, 5319-5344.
(7) For selected reviews of 1,2-sulfur rearrangement, see: (a) Fox,
D. J.; House, D.; Warren, S. Angew. Chem., Int. Ed. 2002, 41, 2462-
2482. (b) Sromek, A. W.; Gevorgyan, V. Top. Curr. Chem. 2007, 274,
77-124. For selective papers of 1,2-sulfur rearrangement, see: (c)
Marshall, J. A.; Roebke, H. J. Org. Chem. 1969, 34, 4188-4191. (d)
Gateau-Olesker, A.; Sepulchre, A. M.; Vass, G.; Géro, S. D. Tetrahe-
dron 1977, 33, 393-397. (e) Tarnchompoo, B.; Thebtaranonth, Y.
Tetrahedron Lett. 1984, 25, 5567-5570. (f) Nickon, A.; Rodriguez, A.
D.; Shirhatti, V.; Ganguly, R. J. Org. Chem. 1985, 50, 4218-4226. (g)
Saigo, K.; Hashimoto, Y.; Fang, L.; Hasegawa, M. Heterocycles 1989,
29, 2079-2082. (h) Afonso, C. A. M.; Barros, M. T.; Godinho, L. S.;
Maycock, C. D. Synthesis 1991, 575-580. (i) Barros, M. T.; Maycock,
C. D.; Santos, L. S. Heterocycles 1998, 48, 1121-1138. (j) Afonso, C.
A. M.; Barros, M. T.; Maycock, C. D. Tetrahedron 1999, 55, 801-814.
(k) Michel, K.; Glotzbach, C.; Fröhlich, R.; Würthwein, E.-U. Eur. J.
Org. Chem. 2012, 960-966.
(8) See Supporting Information (Table S1 and S2) for details.
(9) 1,3-dithiolane (five-membered ring) and 1,3-dithiepane deriva-
tives (seven-membered ring) were also employed as the substrate
under the optimized reaction conditions [(R)-1a (10 mol%) in CH2Cl2
at -40 ºC for 12 h]. The reactions of these cyclicdithioacetals having
the phenyl substituent with pyrrole (3a) afforded the corresponding
1,4-dithiane product 4m and 1,4-dithiocane product 4n in 75% yield
with 32% ee and in 79% yield with 64% ee, respectively. See SI for
details.
1
2
3
4
5
6
7
8
3,3’-positions of CPA.
This mechanistic case study provided a very clear under-
standing of the nature of the ion-pairing interaction in CPA-
catalyzed reactions. It will be very helpful for the further de-
velopment of novel organocatalytic transformations using a
CPA catalyst. The results of this in-depth mechanistic case
study are expected to hasten the design of new organocatalysts
in the future.
9
ASSOCIATED CONTENT
Supporting Information.
The exploratory investigation, experimental procedures and char-
acterization data. This material is available free of charge via the
Internet at http://pubs.acs.org.
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
AUTHOR INFORMATION
Corresponding Author
mterada@m.tohoku.ac.jp
Notes
The authors declare no competing financial interest.
Present Address
†F.L.: Department of Chemical and Environmental Engineering,
Yale University, New Haven, CT 06511, USA.
Author Contributions
All authors have given approval to the final version of the manu-
script.
ACKNOWLEDGMENT
This work was partially supported by a Grant-in-Aid for Scientific
Research on Innovative Areas “Advanced Molecular Transfor-
mations by Organocatalysts” from The MEXT, Japan (No.
23105002). We thank JSPS for a research fellowship for Young
Scientists (F.L.).
(10) For examples of using BINOL-derived chiral phosphoric acid
with electron-withdrawing group at 6,6’-position, see: (a) Akiyama,
T.; Honma, Y.; Itoh, J.; Fuchibe, K. Adv. Synth. Catal. 2008, 350,
399-402. (b) Harada, S.; Kuwano, S.; Yamaoka, Y.; Yamada, K.;
Takasu, K. Angew. Chem., Int. Ed. 2013, 52, 10227-10230. (c)
Kuroda, Y.; Harada, S.; Oonishi, A.; Kiyama, H.; Yamaoka, Y.;
Yamada, K.; Takasu, K. Angew. Chem., Int. Ed. 2016, 55, 13137-
13141.
ABBREVIATIONS
CPA, chiral phosphoric acid; EDS, enantio-determining step; IRC,
intrinsic reaction coordinate; NOE, nuclear Overhauser effect;
RDS, rate-determining step.
(11) The attempt to synthesize a substrate bearing an electron-
donating 4-methoxyphenyl group failed.
(12) Supplementary crystallographic data of 4f, 4k, 2l and 5d are
deposited at the Cambridge Crystallographic Data Center. The deposi-
tion numbers of them are CCDC 1437351 (4f), CCDC 1437411 (4k),
CCDC 1520496 (2l), and CCDC 1520496 (5d), respectively.
(13) The geometries of 2k and 2l were determined by NMR analy-
sis in combination with NOE spectra and coupling constants JHH (See
SI for details). The geometry of 2l was also determined by single-
crystal X-ray crystallography.
(14) The absolute stereochemistries of 4l, 6k, 7k, and 7l were un-
ambiguously determined by chemical derivatizations on the basis of
the stereochemically assigned (2R,6S)-4k whose stereochemistry has
been determined by X-ray crystallography. See SI for details.
(15) The geometry of the C=N double bond of 5d was determined
to be E by X-ray crystallography. However, N-aryl imidates were
reported to undergo facile E/Z isomerization based on the line broad-
ening phenomenon of the N-aryl moiety in the NMR spectrum, see:
Fischer, D. F.; Barakat, A.; Xin, Z.-Q.; Weiss, M. E.; Peters, R. Chem.
Eur. J. 2009, 15, 8722-8741. We did not observe such line broadening
in the NMR spectra of 5d and 5l.
REFERENCES
(1) For selected reviews about chiral anion controlled asymmetric
catalysis, see: (a) Lacour, J.; Moraleda, D. Chem. Commun. 2009,
7073-7089. (b) Phipps, R. J.; Hamilton, G. L.; Toste, F. D. Nat. Chem.
2012, 4, 603-614. (c) Parra, A.; Reboredo, S.; Castro, A. M. M.;
Alemán, J. Org. Biomol. Chem. 2012, 10, 5001-5020. (d) Mahlau, M.;
List, B. Angew. Chem., Int. Ed. 2013, 52, 518-533. (e) Brak, K.;
Jacobsen, E. N. Angew. Chem., Int. Ed. 2013, 52, 534-561. Also see:
(f) Parmar, D.; Sugiono, E.; Raja, S.; Rueping, M. Chem. Rev. 2014,
114, 9047-9153.
(2) Seminal works of chiral phosphoric acid, see: (a) Akiyama, T.;
Itoh, J.; Yokota, K.; Fuchibe, K. Angew. Chem., Int. Ed. 2004, 43,
1566-1568. (b) Uraguchi, D.; Terada, M. J. Am. Chem. Soc. 2004, 126,
5356-5357. For selected reviews of chiral phosphoric acid, see: (c)
Akiyama, T. Chem. Rev. 2007, 107, 5744-5758. (d) Adair, G.;
Mukherjee, S.; List, B. Aldrichimica Acta 2008, 41, 31-39. (e) Terada,
M. Chem. Commun. 2008, 4097-4112. (f) Terada, M. Synthesis 2010,
1929-1982. Also see: Ref. 1f.
(16) When the reaction of 5 was conducted under the optimized
conditions (-40 ºC, 12 h), no conversion of 5 was observed.
(17) Kinetic studies were carried out by changing the concentration
of pyrrole (3a) (0.10, 0.15, 0.30, and 0.60 M) under the optimized
(3) Kanomata, K.; Toda, Y.; Shibata, Y.; Yamanaka, M.; Tsuzuki,
S.; Gridnev, I. D.; Terada, M. Chem. Sci. 2014, 5, 3515-3523.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 16
reaction conditions (using 10 mol% (R)-1a and 0.10 M solution of 2a
(25) Christ, P.; Lindsay, A. G.; Vormittag, Sonja S.; Neudcrfl, J.-
in CH2Cl2 at -40 °C). As the result, the rate of the reaction was de-
pendent on the concentration of pyrrole (3a), however it was reduced
when increasing the concentration of 3a. Thus, 3a was found to func-
tion as an inhibitor of the present reaction and hence it is difficult to
conclude whether 3a is involved into the rate-determining step or not
on the basis of the present kinetic studies. See SI for details.
(18) The absolute configuration of the major enantiomer of N-Me-
4a was determined to be R by derivatization (See SI for details). This
selectivity is probably explained by using the following model:
M.; Berkessel, A.; O’Donoghue, A. C. Chem. Eur. J. 2011, 17, 8524-
8528.
1
2
3
4
5
6
7
8
(26) During the course of the reaction, white precipitate was
formed at the beginning of the reaction. After collecting the white
precipitate by filtration, the precipitate was assigned to be trichloroa-
cetamide by NMR and mass spectroscopy. The control experiment by
adding extra trichloroacetamide (1 equiv.) to the reaction of 2a with
3a catalyzed by (R)-1a resulted in little difference in yield and % ee
between the control and the original data. This result suggests that the
generated trichloroacetamide does not involve into the catalytic reac-
tion. Therefore involvement of trichloroacetamide in the reaction was
not considered in DFT studies.
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
(27) Frisch, M. J.; et al. Gaussian 09, Revision C.01; Gaussian,
Inc.: Wallingford, CT, 2010. See SI for the full citation.
(28) For the B97D method, see: Grimme, S. J. Comput. Chem.
2006, 27, 1787-1799.
(29) For Gaussian basis sets, see: (a) Frisch, M. J.; Pople, J. A.;
Binkley, J. S. J. Chem. Phys. 1984, 80, 3265-3269. (b) Hehre, W. J.;
Radom, L.; Schleyer, P. v. R.; Pople, J. A. Ab initio Molecular Or-
bital Theory; John Wiley: New York, USA, 1986, and references
cited therein.
(30) (a) Barone, V.; Cossi, M. J. Phys. Chem. A 1998, 102, 1995-
2001. (b) Barone, B.; Cossi, M.; Tomasi, J. J. Comput. Chem. 1998,
19, 404-417.
This CNu-H···O hydrogen bond has been reported in the CPA-
catalyzed enantioselective Friedel-Crafts reaction of 2-methoxyfuran
with ketimine, see: Kondoh, A.; Ota, Y.; Komuro, T.; Egawa, F.;
Kanomata, K.; Terada, M. Chem. Sci. 2016, 7, 1057-1062.
(19) The reactions of 2l with N-methylpyrrole (3c) and N-
methylindole (3d) were also carried out under the optimized reaction
conditions. In both of the reactions, the ring expansion/deprotonation
product (R)-6k was obtained as the major product in 88% yield with
82% ee for 3c and in 81% yield with 81% ee for 3d. The ring expan-
sion/addition products N-Me-4l (derived from 3c) and N-Me-7l (de-
rived from 3d) were obtained in a nearly diastereomerically pure form
(N-Me-4l: 11% yield, 98.5:1.5 dr, 76% ee, N-Me-7l: 19% yield,
>99:1 dr, 72% ee) and hence the reaction proceeded in a stereospecif-
ic manner, regardless in the absence/presence of the hydrogen bond
donor nucleophile. In addition, the observed ee values of N-Me-4l
and N-Me-7l were slightly lower than those of (R)-6k. These results
suggest that parallel kinetic resolution of the enantio-enriched (“well-
ordered”) reactive intermediate occurred under the influence of the
chiral phosphate anion.
(20) (a) Mayr, H.; Bug, T.; Gotta, M. F.; Hering, N.; Irrgang, B.;
Janker, B.; Kempf, B.; Loos, R.; Ofial, A. R.; Remennikov, G.;
Schimmel, H. J. Am. Chem. Soc. 2001, 123, 9500-9512. (b) Kempf,
B.; Hampel, N.; Ofial, A. R.; Mayr, H. Chem. Eur. J. 2003, 9, 2209-
2218. (c) Lakhdar, S.; Westermaier, M.; Terrier, F.; Goumont, R.;
Boubaker, T.; Ofial, A. R.; Mayr, H. J. Org. Chem. 2006, 71, 9088-
9095.
(31) The direct displacement of trichloroacetimidate by phosphoric
acid catalyst via SN2 reaction, affording phosphate ester as the reac-
tive intermediate, was also considered as the plausible pathway. How-
ever, in general, the SN2 reaction is significantly influenced by the
steric congestion around the reaction site. In fact, substrate 2a has the
sterically hindered tetrasubstituted carbon next to the reaction site.
This steric congestion, namely the neopentyl effect on the SN2 reac-
tion, prevents the proposed SN2 reaction. Therefore it is considered
that the proposed reaction pathway is unfavorable.
(32) Structural optimization of chair-equatorial conformation mod-
els was initially conducted using B3LYP/6-31G(d,p), because error
occurred when optimization of these equatorial models using the
B97D/6-31G(d,p) method. The preliminarily calculations of these
equatorial models at the B3LYP/6-31G(d,p) resulted in a conforma-
tional change in 1,3-dithiane from the chair forms to the boat-like
ones. Further optimization of these boat-axial conformation models
was carried out at the B97D/6-31G(d,p) method to afford the corre-
sponding transition states, TS-1-equatrial-in and TS-1-equatrial-out.
Regarding the B3LYP method, see: (a) Becke, A. D. J. Chem. Phys.
1993, 98, 5648-5652. (b) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B
1988, 37, 785-789.
(33) (a) Desiraju, G. R.; Steiner, T. The Weak Hydrogen Bond; Ox-
ford University Press: Oxford, 1999. (b) Castellano, R. K. Curr. Org.
Chem. 2004, 8, 845-865. (c) Knowles, R. R.; Jacobsen, E. N. Proc.
Natl. Acad. Sci. U. S. A. 2010, 107, 20678−20685. (d) Johnston, R.
C.; Cheong, P. H.-Y. Org. Biomol. Chem. 2013, 11, 5057-5064.
(34) For representative and recent examples for C-H···O hydrogen
bonds as the key interaction in asymmetric catalysis, see: (a) Corey, E.
J.; Rohde, J. J. Tetrahedron Lett. 1997, 38, 37–40. (b) Grayson. M.
N.; Goodman, J. M. J. Am. Chem. Soc., 2013, 135, 6142-6148. (c)
Maity, P.; Pemberton, R. P.; Tantillo, D. J.; Tambar, U. K. J. Am.
Chem. Soc. 2013, 135, 16380-16383. (d) Wang, H.; Jain, P.; Antilla. J.
C.; Houk, K. N. J. Org. Chem., 2013, 78, 1208-1215. (e) Ishii, T.;
Watanabe, R.; Moriya, T.; Ohmiya, H.; Mori, S.; Sawamura, M.
Chem. Eur. J. 2013, 19, 13547-13553. (f) Grimblat, N.; Sugiura, M.;
Pellegrinet, S. C. J. Org. Chem. 2014, 79, 6754-6758. (g) Xie, Y.;
Cheng, G.-J.; Lee, S.; Kaib, P. S. J.; Thiel, W.; List, B. J. Am. Chem.
Soc. 2016, 138, 14538-14541. (h)Lovie-Toon, J. P.; Tram, C. M.;
Flynn, B. L.; Krenske, E. H. ACS Catal. 2017, 7, 3466-3476. (i)
Grayson, M. N.; Yang, Z.; Houk, K. N. J. Am. Chem. Soc. 2017, 139,
7717-7720. (j) Takayama, Y.; Ishii, T.; Ohmiya, H.; Iwai, T.;
Schwarzer, M. C.; Mori, S.; Taniguchi, T.; Monde, K.; Sawamura, M.
Chem. Eur. J. 2017, 23, 8400-8404.
(21) The kinetic studies were performed to determine the reaction
order and it was confirmed that the reaction is first order in the cata-
lyst. Therefore the monomeric catalyst was considered to be involved
in the system. See SI for details.
(22) For a review of the reactive intermediate, episulfonium ions
and oxocarbenium ions, see: Beaver, M. G.; Billings, S. B.; Woerpel,
K. A. Eur. J. Org. Chem. 2008, 771-781.
(23) Covalent bonded phosphate intermediate has been proposed in
some chiral phosphoric acid catalyzed reactions, see: (a) Sun, Z.;
Winschel, G. A.; Zimmerman, P. M.; Nagorny, P. Angew. Chem., Int.
Ed. 2014, 53, 11194-11198. (b) Liu, L.; Leutzsch, M.; Zheng, Y.;
Wasim Alachraf, M.; Thiel, W.; List, B. J. Am. Chem. Soc. 2015, 137,
13268-13271. (c) Khomutnyk, Y. Y.; Argüelles, A. J.; Winschel, G.
A.; Sun, Z.; Zimmerman, P. M.; Nagorny, P. J. Am. Chem. Soc. 2016,
138, 444-456. (d) Tay, J.-H.; Argüelles, A. J.; DeMars II, M. D.;
Zimmerman, P. M.; Sherman, D. H.; Nagorny, P. J. Am. Chem. Soc.
2017, 139, 8570-8578.
(24) For selected examples of mechanistic studies of SNi reaction,
see: (a) Sinnott, M. L.; Jencks, W. P. J. Am. Chem. Soc. 1980, 102,
2026-2032. (b) Guthrie, R. D.; Jencks, W. P. Acc. Chem. Res. 1989,
22, 343-349. (c) Lee, S. S.; Hong, S. Y.; Errey, J. C.; Izumi, A.; Da-
vies, G. J.; Davis, B. G. Nature Chem. Biol. 2011, 7, 631-638. (d)
Chan, J.; Tang, A.; Bennet, A. J. J. Am. Chem. Soc. 2012, 134, 1212-
1220. (e) Blanco Capurro, J. I.; Hopkins, C. W.; Pierdominici Sottile,
G.; Gonzꢀlez Lebrero, M. C.; Roitberg, A. E.; Marti, M. A. J. Phys.
Chem. B 2017, 121, 471-478.
(35) Organic Chemistry of Sulfur; Oae, S., Eds.; Springer US, 1977.
(36) Seminal works and review about Felkin-Anh model, see: (a)
Cherest, M.; Felkin, H.; Prudent, N. Tetrahedron Lett. 1968, 9, 2199-
ACS Paragon Plus Environment

Page 15 of 16
Journal of the American Chemical Society
2204. (b) Anh, N. T.; Eisenstein, O.; Lefour, J-M.; Dâu, M-E. T. H. J.
(39) For C-H··· interaction, see: (a) Krenske, E. H.; Houk, K. N.
Acc. Chem. Res. 2013, 46, 979-989. (b) Neel, A. J.; Hilton, M. J.;
Sigman, M. S.; Toste, F. D. Nature 2017, 543, 637-646.
Am. Chem. Soc. 1973, 95, 6146-6147. (c) Anh, N. T.; Eisenstein, O.
Tetrahedron Lett. 1976, 17, 155-158. (d) Anh, N. T.; Eisenstein, O.
Nouv. J. Chim. 1977, 1, 61-70. (e) Houk, K. N.; Paddon-Row, M. N.;
Rondan, N. G.; Wu, Y.-D.; Brown, F. K.; Spellmeyer, D. C.; Metz, J.
T.; Li, Y.; Loncharich, R. J. Science 1986, 231, 1108-1117. (f) Men-
gel, A.; Reiser, O. Chem. Rev. 1999, 99, 1191-1223.
1
2
3
4
5
6
7
8
(40) For representative and recent examples for C-H··· as the key
interaction in asymmetric catalysis, see: (a) Yamakawa, M.; Yamada,
I.; Noyori, R. Angew. Chem. Int. Ed. 2001, 40, 2818–2821. (b) Dang,
Y.; Qu, S.; Wang, Z.-X.; Wang, X. J. Am. Chem. Soc. 2014, 136, 986-
998. (c) Jindal, G.; Sunoj, R. B. Angew. Chem., Int. Ed. 2014, 53,
4432-4436. (d) Meng, S. S.; Liang, Y.; Cao, K. S.; Zou, L.; Lin, X.
B.; Yang, H.; Houk, K. N.; Zheng, W. H. J. Am. Chem. Soc. 2014,
136, 12249-12252. (e) Bhaskararao, B.; Sunoj, R. B. J. Am. Chem.
Soc. 2015, 137, 15712-15722. (f) Changotra, A.; Sunoj, R. B. Org.
Lett. 2016, 18, 3730-3733. (g) Doney, A. C.; Rooks, B. J.; Lu, T. X.;
Wheeler, S. E. ACS Catal. 2016, 6, 7948-7955. (h) Pareek, M.; Sunoj,
R. B. ACS Catal. 2016, 6, 3118-3126. (i) Korenaga，T.; Sasaki, R.;
Takemoto, T.; Yasuda, T.; Watanabe, M. Adv. Synth. Catal. 2018,
360, 322-333.
(37) The formation of the present association through the C-H···O
hydrogen bonds could be one of the plausible mechanisms of the SNi
reaction pathway via a stepwise process, although covalently bonded
phosphate intermediate INT-3 or INT-3’ was not detected during the
course of the reaction.
9
TS
∆∆G
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
)−1
(38) We calculated ee value as follows: ee = exp(
.
RT
TS
∆∆G
RT
exp(
)+1
where GTS is the energy gap between major and minor transition
states, R is gas constant, and T is absolute temperature. The 2.6
kcal/mol energy gap at -40 ºC corresponds to 99% ee. However, the
experimental result using 2k is 96% ee.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 16
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
16
ACS Paragon Plus Environment