Transition state analysis of enantioselective br?nsted base catalysis by chiral cyclopropenimines

Article abstract of DOI:10.1021/ja504532d

Experimental ¹³C kinetic isotope effects have been used to interrogate the rate-limiting step of the Michael addition of glycinate imines to benzyl acrylate catalyzed by a chiral 2,3-bis(dicyclohexylamino) cyclopropenimine catalyst. The reaction is found to proceed via rate-limiting carbon-carbon bond formation. The origins of enantioselectivity and a key noncovalent CH?O interaction responsible for transition state organization are identified on the basis of density functional theory calculations and probed using experimental labeling studies. The resulting high-resolution experimental picture of the enantioselectivity-determining transition state is expected to guide new catalyst design and reaction development.

Full text of DOI:10.1021/ja504532d

Subscriber access provided by MEMORIAL UNIV
Article
Transition State Analysis of Enantioselective
Brønsted Base Catalysis by Chiral Cyclopropenimines
Jeffrey S Bandar, Gregory S. Sauer, William D. Wulff, Tristan H. Lambert, and Mathew J. Vetticatt
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/ja504532d • Publication Date (Web): 07 Jul 2014
Downloaded from http://pubs.acs.org on July 8, 2014
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 8
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Transition State Analysis of Enantioselective Brønsted Base
Catalysis by Chiral Cyclopropenimines
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
Jeffrey S. Bandar,‡£ Gregory S. Sauer,‡§ William D. Wulff,∆ Tristan H. Lambert,£* and Mathew J. Vetti-
catt§*
£Department of Chemistry, Columbia University, 3000 Broadway, New York, NY 10027. ∆Department of Chemistry, Michi-
gan State University, East Lansing, MI 48823. §Department of Chemistry, Binghamton University, Binghamton, NY 13902.
ABSTRACT: Experimental 13C kinetic isotope effects have been used to interrogate the rate-limiting step of the Michael addition
of glycinate imines to benzyl acrylate catalyzed by a chiral 2,3-bis(dicyclohexylamino) cyclopropenimine catalyst. The reaction is
found to proceed via rate-limiting carbon-carbon bond formation. The origins of enantioselectivity and a key non-covalent CH…O
interaction responsible for transition state organization are identified based on density functional theory calculations and probed
using experimental labeling studies. The resulting high-resolution experimental picture of the enantioselectivity-determining transi-
tion state is expected to guide new catalyst design and reaction development.
n INTRODUCTION
n RESULTS AND DISCUSSION
Chiral cyclopropenimines developed in one of our labs are
emerging as a powerful class of enantioselective Brønsted base
Determination of Experimental KIEs. We chose the reac-
tion of the glycine imine (2) and benzyl acrylate (3b) catalyzed by
1 for the measurement of 13C KIEs using NMR methodology at
natural abundance.7 Complementary approaches were used for the
determination of 13C KIEs for the two reaction components 2 and
3b. The KIEs for 2 were determined by analysis of product sam-
ples.8 Thus, the isotopic composition of 4b was measured from
two independent experiments taken to 21±2% and 22±2% conver-
sion in 2 and compared to samples of 4b isolated from reactions
taken to 100% conversion. The KIEs for 3b were determined by
analysis of recovered starting material:9 samples of 3b were re-
isolated from two independent experiments taken to 72±2% and
74±2% conversion (with respect to 3b) and were compared to
samples of unreacted 3b. The experimental KIEs calculated from
the change in 13C isotopic composition and the fractional conver-
sion are shown in Figure 1.10
catalysts.1,2 3 The prototypical reaction of glycine imine (2) and
,
methyl acrylate (3a) catalyzed by 2,3- bis(dicyclohexylamino)
cyclopropenimine (1) gives the Michael adduct 4a in 99% yield
with 98% ee. The rate of this transformation is approximately 2-3
orders of magnitude faster than that achieved with an analogous
guanidine based catalyst.4 This example of the superior perfor-
mance of cyclopropenimines hints at the tremendous potential of
this new class of Brønsted base catalysts.5
Bn
OH
N
10 mol%
1
Cy2N
NCy2
Ph
N
CO2tBu
CO2R
Ph
N
CO2tBu
+
CO2R
1h, EtOAc, r.t.
Ph
Ph
3a R = Me
3b R = Bn
2
4a - 99% yield, 98% ee
4b - 97% yield, 98% ee
Scheme 1. Chiral Brønsted base-catalyzed Michael addition of
glycine imine to acrylates.
An in-depth understanding of this catalyst system would be
invaluable to our efforts to pursue new methodological applica-
tions and further catalyst development studies. With this goal in
mind, we undertook a mechanistic study of the title reaction using
experimental 13C kinetic isotope effects (KIEs) and density func-
tional theory (DFT) calculations. 13C KIEs are uniquely sensitive
probes of the rate-limiting transition state geometry of a reaction
and have been successfully used to probe the mechanism of sev-
eral fundamental organic reactions.6 Herein we report identifica-
tion of the rate-limiting step, elucidation of stabilizing transition
state interactions, and the origin of enantioselectivity of this novel
organocatalytic reaction.
Figure 1. Experimental 13C KIEs for reaction of 2 and 3b cata-
lyzed by 1. The two sets of KIEs for each carbon represent two
independent experiments and the numbers in parentheses repre-
sent the standard deviation in the last digit as determined from six
measurements. KIEs for the bond-forming carbon atoms are
shown in red.
The experimental 13C KIEs provide qualitative information
about the rate-limiting step in a catalytic cycle. In our initial pub-
lication,1 we proposed a mechanism for this reaction that involves
initial deprotonation of 2 by 1 to generate H-bonded cycloprope-
nium ion-enolate complex 5 (Scheme 2). Subsequent Michael
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 8
addition of 5 to 3 affords adduct 7 via transition state 6. Finally,
tants using the semi-empirical method (AM1). The division of
1
2
3
4
5
6
7
proton transfer from the cyclopropenium ion to the Michael ad-
duct results in release of 4 and regeneration of 1. Substantial 13C
KIEs for the two bond-forming carbon atoms (shown in red in
Figure 1) suggest carbon-carbon bond formation to be the rate-
limiting step in the catalytic cycle. All other KIEs are close to
unity, consistent with minor rehybridization occurring at this iso-
tope-sensitive step.
layers for the ONIOM calculations is shown in Figure 2. The time
efficiency of the ONIOM method allowed us to explore a range of
transition structures, including those involving higher energy
catalyst conformations, those involving different H-bonding sce-
narios, and those lacking multiple H-bonding interactions. The
summary of the results from the ONIOM study is presented in the
Supporting Information.
8
9
O
Me
O
Me
Bn
Me
Me
Me
Me
Ph
N
Ph
N
OH
O
O
N
Ph
4a
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
Ph
2
Cy2N
NCy2
O
OMe
1
dep
roto
natio
n
proton transfer
O–
Ph
Ph
MeO
N
1•H+
N
Ph
1•H+
O
O–
Ph
O
O
5
7
Me
Me
Me
Me
Me
Me
O
Ph
Ph
MeO
Figure 2. Division of layers for the ONIOM method used for
initial exploration of transition structures.
O
C-C bond formation
N
MeO
1•H+
Me
3a
O
O–
Me
Me
6
The exploratory ONIOM study led to the identification of a
subset of viable transition structures arising from four distinct
binding modes (5a–d, Figure 3) of the catalyst–enolate complex
5. A simple template that can be used to describe these ‘most
likely’ transition state assemblies is shown in Figure 3. After ini-
tial deprotonation of 2 by 1, the resulting catalyst–bound enolate 5
can adopt either the E or Z geometry. In mono-coordinated bind-
ing mode 5a, the enolate is held by a single H-bonding interaction
between the hydroxyl group on the catalyst and the enolate oxy-
gen. The NH moiety of protonated 1 presumably directs 3a for
conjugate attack by H-bonding to the oxygen atom of 3a at the
transition state. Four possible orientations of binding mode 5a that
allow for this combination of H-bonding interactions are shown in
Figure 3. They are labeled 5aRE, 5aRZ, 5aSE, and 5aSZ based
on the binding mode (5a), the enantiomer of product formed (R or
S), and the enolate geometry (E or Z). A similar (and comple-
mentary) situation arises when the enolate oxygen is H-bonded to
the NH moiety of catalyst 1 – binding mode 5b. In this mono-
coordinated binding mode, the enolate can once again adopt either
a E or Z conformation and the hydroxyl group directs 3a for con-
jugate attack via H-bonding to the oxygen atom. Four additional
conformations – 5bRE, 5bRZ, 5bSE, and 5bSZ – can be envi-
sioned from binding mode 5b.
In binding modes 5c and 5d, the enolate is bound to the cata-
lyst via two H-bonding interactions – between the oxygen and
nitrogen acceptor atoms (of the enolate) and the two H-bond do-
nors in the protonated catalyst. Two important features distinguish
these di-coordinated binding modes from the mono-coordinated
counterparts: (1) the enolate is forced to adopt only the E confor-
mation to accommodate the two H-bonding interactions, and (2)
in the transition structures for conjugate attack based on binding
modes 5c and 5d, the oxygen atom of 3a is not involved in H-
bonding with the catalyst. For obvious reasons, transition struc-
tures wherein neither the NH nor the OH moiety of 1 are involved
in H-bonding were not considered in detail.
Scheme 2. Proposed mechanism based on experimental KIEs.
Theoretical Studies. We next sought to gain detailed in-
sight into the organization of this enantioselectivity-determining
carbon-carbon bond-forming transition state through theoretical
studies. Preliminary results reported in our initial publication1
suggest a key role of the dicyclohexylamino substituents in medi-
ating reaction efficiency and enantioselectivity, while the NH
proton and the pendant hydroxyl group in catalyst 1 are vital ele-
ments in assembling this transition state via multiple H-bonding
interactions. A comprehensive theoretical study of this reaction is
complicated by a number of factors, namely: (1) the size of the
system – the cyclopropenimine-catalyzed reaction of 2 and 3a
involves 150 atoms, (2) the possibility of several conformations
for the catalyst, (3) the possibility of either E or Z geometry of the
intermediate enolate (5) and of s-cis or s-trans geometry of the
acrylate (3a), and (4) several competing H-bonding scenarios in
the assembly of the transition state.
A detailed investigation of the catalyst geometry revealed a
preference for a conformation wherein the cyclohexyl rings are
geared in the same direction and the hydrogen atom at the chiral
center is oriented anti to the NH proton (the H-C-N-H dihedral
angle is -146°; B3LYP/6-31+G**).11 The next step in the study
was the identification of carbon-carbon bond forming transition
structures leading to the major and minor enantiomers of product
4a and the theoretical prediction of enantioselectivity. Due to the
large size of the system, initial explorations of transition struc-
tures were performed using the hybrid ONIOM12 (B3LYP/6-
31+G**:AM1) method as implemented in Gaussian '09.13 The
ONIOM method treats the key bond-forming and H-bonding por-
tions of the transition state using the high-level DFT method
(B3LYP/6-31+G**) and the steric bulk of the catalyst and reac-
ACS Paragon Plus Environment

Page 3 of 8
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
Figure 3. Possible binding modes of the catalyst-enolate complex 5.
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
Transition structures based on the 12 distinct binding modes
coordinated binding modes 5c and 5d (Supporting Information)
were found to be over 12 kcal/mol (at least) higher in energy than
TS5bSE, illustrating the importance of H-bonding between the
catalyst and both reacting partners at the transition state.
of 5 were re-calculated using the B3LYP/6-31G* method.14 The
DFT treatment of the whole system is expected to yield a better
description of the energetics of non-covalent interactions (such as
15
CH-π interactions) compared to the ONIOM calculations. All
Six of the eight transition structures in Figures 4 and 5 are
within 5 kcal/mol of the lowest energy transition structure
TS5bSE. Interestingly, the two transition structures that are sig-
nificantly higher in energy – TS5bSZ (Erel=11.5 kcal/mol) and
TS5bRE (Erel=10.4 kcal/mol) – are almost identical except with
respect to the geometry of the enolate. In our attempt to under-
stand this result, we took a closer look at the key difference be-
tween these two transition structures and the six lower energy
transition structures. All transition structures in Figures 4 and 5
have similar carbon-carbon bond forming distances (2.0–2.2 Å),
and are stabilized by the two short H-bonding interactions (NH…O
and OH…O in the range of 1.7–1.9 Å) discussed earlier. An addi-
tional (and rather unusual) CH…O interaction, in the range of 2.1–
2.3 Å between the cyclohexyl–CH and the hydroxyl oxygen atom,
appears to be contributing to transition state stabilization of the
six lower energy transition structures in Figures 4 and 5. Intri-
guingly, this CH…O distance is significantly larger in the two
high-energy transition structures TS5bRE (3.39 Å, Figure 5) and
TS5bSZ (3.52 Å, Figure 4). This finding suggests that the contri-
bution of the CH…O interaction to transition state stabilization is,
crudely, around 5 kcal/mol. This is based on the observation that
there is no obvious deleterious interaction (apart from the elongat-
ed CH…O distance) in these two transition structures that is not
present in one of the six transition structures that are within 5
kcal/mol of the lowest energy transition structure. A more detailed
evaluation of this interaction will be presented in a later section.
reported distances are in angstroms and all reported energies are
E+zpe energy from the B3LYP/6-31G* calculations. Frequency
calculations performed on these transition structures revealed one
imaginary frequency corresponding to carbon-carbon bond for-
mation.
Shown in Figure 4 (S transition structures) and Figure 5 (R
transition structures) are the eight transition structures correspond-
ing to the eight geometries shown in Figure 3 for mono-
coordinated binding modes 5a and 5b. Two features are common
to all eight transition structures namely (1) strong H-bonding in-
teractions between both reactants (2 and 3a) and the two H-bond
donors in the catalyst, and (2) an s-cis conformation of 3a.16 The
lowest energy transition structures leading to each enantiomer –
TS5bSE (Erel= 0.0 kcal/mol, Figure 4) and TS5aRZ (Erel= 1.7
kcal/mol, Figure 5) – are highlighted using green and red boxes,
respectively. This energy difference (1.7 kcal/mol) corresponds to
a predicted 89% ee. Consideration of a contributing second transi-
tion structure leading to the major enantiomer (TS5aSE, Erel = 0.9
kcal/mol; which is still lower in energy than TS5aRZ by 0.8
kcal/mol) gives an altered prediction of 92% ee – this is in good
agreement with the experimental 98% ee.17
The exact origin of the favorability of TS5bSE over
TS5aRZ is probably attributable to a complex interplay of several
stabilizing/destabilizing interactions. A detailed examination of
some of these stabilizing interactions will be discussed in a later
section. Finally, all transition structures resulting from the di-
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 8
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 4. Transition structures leading to major (S) enantiomer of product 4a that utilize mono-coordinated binding modes 5a and 5b.
Most hydrogen atoms have been removed for clarity. All transition structures are oriented with the acrylate in the foreground and the eno-
late in the background.
ACS Paragon Plus Environment

Page 5 of 8
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
Figure 5. Transition structures leading to minor (R) enantiomer of product 4a that utilize mono-coordinated binding modes 5a and 5b.
Most hydrogen atoms have been removed for clarity. All transition structures are oriented with the acrylate in the foreground and the eno-
late in the background.
the experimental values obtained for each carbon atom, are shown
in Figure 6. The excellent agreement of experiment and theory
validates the transition state model and supports carbon-carbon
bond-formation to be the rate-limiting step of the reaction.
0.999
1.000 (assumed)
1.006 (3)
1.011 (7)
0.996 (3)
0.991 (3)
1.001
1.008
O
O
Ph
N
1.002 (2)
1.010 (7)
O
On the origin of enantioselectivity. In order to obtain a
qualitative breakdown of the energy difference between the key
transition structures contributing to the formation of either enanti-
omer of product, an analysis of H-bonding and other stabilizing
interactions was performed (Table 1). Also listed is the carbon-
carbon bond-forming distance of each transition structure. The
strength of an H-bonding interaction depends on both the H-
bonding distance and the donor-hydrogen-acceptor angle – a line-
ar arrangement of the three atoms forming the angle results in the
strongest H-bonding interaction.21 Each H-bonding distance is
listed in Table 1 along with the donor-hydrogen-acceptor angle.
Finally, the structures have been analyzed for stabilizing CH-π
interactions – the criterion being that the relevant hydrogen is
appropriately oriented with respect to the π system and that the
1.007
1.034
O
Ph
Ph
1.036 (4)
1.035 (3)
1.033
1.000 (assumed)
1.000
1.032 (2)
1.032 (4)
Figure 6. Comparison of experimental (black) and predicted (red)
13C KIEs
Predicted KIEs. In order to interpret the experimental KIE
results described above, 13C KIEs were computed from the scaled
vibrational frequencies of the two lowest energy transition struc-
tures leading to the major enantiomer (TS5bSE and TS5aSE, as a
weighted average based on their energies) using the program
ISOEFF98.18 19 A one-dimensional tunneling correction20 was
,
applied to the predicted 13C KIEs. The predicted KIEs, along with
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 8
distance between the hydrogen and the closest aromatic carbon
atom is < 3.0 Å.
ther supported by the energies of the transition structures resulting
1
2
3
4
5
6
7
8
from binding modes 5c and 5d – even though all four transition
structures TS5cRE, TS5cSE, TS5dRE and TS5dSE had Cy-
CH…O distances between 2.1 and 2.4 Å (not shown, See Support-
ing Information), they were found to be 12.0, 13.2, 19.5, and 16.5
kcal/mol (respectively) higher in energy than TS5bSE due to lack
of transition state stabilization of the acrylate oxygen atom.
While the contribution of the CH…O interaction to transition
state stabilization is significant, its role in determining enantiose-
lectivity is expected to be minimal since both TS5bSE and
TS5aRZ benefit from the stabilization afforded by this interac-
tion. Nevertheless, we envisioned that slight perturbation of enan-
tioselectivity could be used as an experimental probe of this inter-
action – since the CH…O distance is slightly different in the key
transition structures. With this in mind, we synthesized the deu-
terated version of the catalyst – 1a with all four Cy-CHs substitut-
ed by deuterium. In order to observe even a small KIE on the
enantioselectivity, a catalyst mixture of 50% (S)-d4-1 and 50%
(R)-1 was employed for the addition of glycine imine 2 to 3a
(Figure 7). A KIE would be manifested in the reaction in the
form of a non-racemic product. The product from this reaction
was essentially racemic (<5% ee), however, indicating a KIE of
~1.0. The lack of an observed KIE does not undermine the im-
portance of a transition state CH…O interaction. Indeed, because
the relevant hydrogen is not participating in any bond-forming or
bond-breaking events, and because both the major and minor
enantiomer transition states likely benefit from this interaction to
similar extents, it is not unreasonable that any KIE would be too
small to measure.22
Table 1. Analysis of stabilizing interactions for TS5bSE,
TS5aSE, and TS5aRZ
TS5bSE
TS5aSE
TS5aRZ
1. C-C distance
1.99 Å
2.20 Å
2.22 Å
+NH (1.93 Å)
OH (1.78 Å)
OH (1.73 Å)
2. enolate bound
to
9
∠NHO = 154°
∠NHO = 159°
∠OHO = 171°
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
OH (1.69 Å)
+NH (1.82 Å)
+NH (1.83 Å)
3. acrylate bound
to
∠OHO = 173°
∠OHO = 158°
∠NHO = 162°
2.21 Å
2.16 Å
2.15 Å
4. Cy-CH···O
distance
∠CHO = 171°
∠CHO = 163°
∠CHO = 168°
1
2
0
5. CH-π interac-
tions
The two main features that distinguish TS5bSE from the
other two structures are: (1) the shorter C-C bond distance (1.99
Å) and (2) a stronger interaction between the acrylate oxygen
atom and the H-bond donor atom (1.69 Å and ∠OHO = 173.2°).
Since C-C bond formation is more advanced in TS5bSE, there is
a greater build-up of negative charge at the acrylate oxygen atom
resulting in the stronger H-bonding interaction. Favorable CH-π
interactions were also identified in both TS5bSE and TS5aSE
that were absent in TS5aRZ. (See Supporting Information for
pdbs of these structures).
A full understanding of the exact origin of enantioselection
in this system is complicated by the accessibility of both E and Z
enolates and the possibility of multiple binding modes for the
catalyst-enolate complex 5 (Figure 3). In contrast to the vast ma-
jority of asymmetric transformations, enantioselectivity in this
reaction is not a result of selective access of one face versus the
other of a single complex; rather it appears to result from the best
network of H-bonding interactions, geometry of the enolate, and
other stabilizing interactions. The observation that the major enan-
tiomer (S) is formed from two very geometrically distinct transi-
tion structures – TS5bSE (~83%) and TS5aSE (~17%) – suggests
that prediction of enantioselectivity for other reactions catalyzed
by 1 (or even the same reaction with a different Michael acceptor)
will require a full consideration of all possible transition state
assemblies. The main contribution of this work is the develop-
ment of a template (Figure 3) that provides a systematic basis for
the prediction of enantioselectivity in reactions catalyzed by 1.
We are currently adopting this template for the optimization of
reactions that are in the early stages of development.
Analysis of the CH…O interaction. We next turned our at-
tention to understanding why the Cy-CH…O distance is elongated
in the two high energy transition structures TS5bRE and
TS5bSZ. Among all the binding modes shown in Scheme 4, the
bond-forming enolate carbon atom is the farthest away from the
pendant hydroxyl group in binding modes 5bRE and 5bSZ. As a
result, the hydroxyl group has to adopt a conformation that is
different (from the six other lower energy transition structures in
Figures 4 and 5) in order to stabilize the oxygen atom of 3a via H-
bonding at the transition state. This different conformation of the
hydroxyl group does not allow close proximity between the Cy-
CH and the hydroxyl oxygen. In other words, maintaining the
CH…O interaction appears to be less important than stabilization
(via H-bonding) of the developing negative charge of the oxygen
atom of 3a in these transition structures. This observation is fur-
10 mol%
catalyst mixture
Ph
N
CO2tBu
Ph
N
CO2tBu
+
CO2Me
PhMe, 23 °C, 3 h
Ph
2, 1.0 eq
Ph
4a
3a, 3.0 eq
CO2Me
< 3% ee
catalyst mixture:
Bn
N
Bn
% ee
2.4 (S)
1.4 (S)
1.9 (S)
Trial
OH
OH
1
2
3
N
+
Cy2N
NCy2
d2Cy2N
NCy2d2
50%
50%
Figure 7. Labeling studies as an experimental probe of CH…O
interaction.
Finally, examination of this interaction in the crystal struc-
ture (~2.51 Å)3 and computed structures (2.35 Å, B3LYP/6-31G*)
of the protonated catalyst revealed that it is present, albeit to a
lesser extent, in the ground state. The discrepancy in these num-
bers suggests that our calculations might be slightly over-
estimating this interaction.23 However, the compression of this
distance on going from the protonated catalyst to the transition
state, by 0.15 – 0.2 Å suggests that the role of this interaction in
transition state stabilization is non-trivial. We recognize that gas
phase B3LYP/6-31G* calculations might be inadequate in accu-
rately describing the energetic contribution of this interaction but
we expect higher-level calculations to exhibit similar trends on
going from the ground state to the transition state. The prelimi-
nary investigations using ONIOM methods, where this interaction
was calculated using a 6-31+G** basis set, showed similar trends
in the CH…O distance at the transition state (2.10 – 2.3 Å, See
Supporting Information for complete details and analyses of these
calculations).
n CONCLUSIONS
ACS Paragon Plus Environment

Page 7 of 8
Journal of the American Chemical Society
In conclusion, experimental 13C KIEs and theoretical studies
provide high-resolution picture of the enantioselectivity-
determining transition state of the Michael addition of glycinate
imines to acrylate catalyzed by chiral 2,3-
Corresponding Author
1
2
3
4
5
6
7
8
a
vetticatt@binghamton.edu
tl2240@columbia.edu
a
Notes
bis(dicyclohexylamino) cyclopropenimine catalyst. On the basis
of these studies, we have developed a template for predicting
enantioselectivity in reactions catalyzed by 1. An unusual intra-
molecular CH…O interaction has been identified as a key element
in transition state organization. Transition structures where H-
bond donors on the catalyst stabilize both reactants via H-bonding
are favored. Ultimately, enantioselection results from the best
network of H-bonding interactions, geometry of the enolate and
other stabilizing interactions. The use of this information for the
rational design of novel cyclopropenimine-based Brønsted base
catalysts will be reported in due course.
The authors declare no competing financial interest.
Author Contributions
JSB and GSS contributed equally to this work.
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
n ACKNOWLEDGMENTS
MJV acknowledges support from startup funding at Bing-
hamton University from the SUNY Research Foundation and
the National Science Foundation through XSEDE resources
provided by the XSEDE Science Gateways program. THL
acknowledges financial support from NIH NIGMS
(R01 GM102611) and an Ely Lilly Grantee Award. WDW
acknowledges support from NIH NIGMS (R01 GM094478). JSB
is grateful for NSF and NDSEG doctoral fellowships. We thank
the Leighton group (Columbia University) for use of their in-
strumentation.
n ASSOCIATED CONTENT
Supporting Information. Experimental and computational
details and discussions, NMR spectra and integrations, rele-
vant pdb files, absolute energies and the coordinates of the
atoms in all the molecules whose geometries were optimized..
This material is available free of charge at http://pubs.acs.org.
n AUTHOR INFORMATION
n REFERENCES
(1) Bandar, J.; Lambert, T. H. J. Am. Chem. Soc. 2012, 134, 5552- 5555.
(2) Bandar, J. S.; Lambert, T. H. J. Am. Chem. Soc. 2013, 135, 11799-11802.
(3) Bandar, J. S.; Lambert, T. H. J. Am. Chem. Soc. 2014, Submitted (see Supporting Information for Review Only).
(4) Ishikawa, T.; Araki, Y.; Kumamoto, T.; Seki, H.; Fukuda, K.; Isobe, T. Chem. Comm. 2001, 245.
(5) Several catalytic enantioselective Brønsted base catalyzed reactions are currently being developed using 1 and related catalysts. 1 is available
as a salt from Sigma-Aldrich.
(6) Selected examples include (a) Beno, B. R.; Houk, K. N.; Singleton, D. A. J. Am. Chem. Soc. 1996, 118, 9984-9985. (b) Singleton, D. A.;
Merrigan, S. R.; Liu, J.; Houk, K. N.; J. Am. Chem. Soc. 1997, 119, 3385-3386. (c) DelMonte, A. J.; Haller, J.; Houk, K. N.; Sharpless, K. B.;
Singleton, D. A.; Strassner, T.; Thomas, A. A. J. Am. Chem. Soc. 1997, 119, 9907-9908. (d) Keating, A. E.; Merrigan, S. R.; Singleton, D. A.;
Houk, K. N. J. Am. Chem. Soc. 1999, 121, 3933-3938. (e) Frantz, D. E.; Singleton, D. A. J. Am. Chem. Soc. 2000, 122, 3288-3295. (f) Zhu, H.;
Clemente, F. R.; Houk, K. N.; Meyer, M. P. J. Am. Chem. Soc. 2009, 131, 1632-1633. (g) Vetticatt, M. J.; Desai, A. A.; Wulff, W. D. J. Org.
Chem. 2013, 78, 5142-5152.
(7) Singleton, D. A.; Thomas, A. A. J. Am. Chem. Soc. 1995, 117, 9357-9358.
(8) (a) Frantz, D. E.; Singleton, D. A.; Snyder, J. P. J. Am. Chem. Soc. 1997, 119, 3385–3386. (b) Singleton, D. A.; Schulmeier, B. E. J. Am.
Chem. Soc. 1999, 121, 9313–9317.
(9) We chose this approach because re-isolation of 3b from the product mixture could be accomplished relatively easily unlike 2, which is prone
to hydrolysis during chromatography.
(10) Full experimental details of the KIE measurements are provided in the Supporting Information.
(11) This computed geometry of the protonated catalyst is consistent with the crystal structure of the protonated catalyst complexed to a chloride
ion and a water molecule. See Ref. 1 for published crystal structure.
(12) (a) Svensson, M.; Humbel, S.; Morokuma, K. J. Chem. Phys. 1996, 105, 3654-3661. (b) Vreven, T.; Morokuma, K. J. Comput. Chem.
2000, 21, 1419-1432. (c) Dapprich, S.; Koma´romi, I.; Byun, K. S.; Morokuma, K.; Frisch, M. J. J. Mol. Struct. 1999, 461, 1-21.
(13) Gaussian 09, Revision A.1, Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.;
Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnen-
berg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.;
Montgomery, Jr., J. A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Nor-
mand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross,
J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.;
Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö.; Foresman,
J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian, Inc., Wallingford CT, 2009.
(14) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(15) Usually, detailed computational investigations of such large systems are carried out using hybrid(DFT:semi-empirical) methods such as
ONIOM followed by single-point energy calculations to obtain the best-possible energy. These methods are significantly more error-prone than a
full-DFT treatment.
(16) This observation has been validated experimentally – 2-cyclopentenone (s-trans) is found to be a poor Michael acceptor while 2-
methylenecyclopentanone (s-cis) is an excellent acceptor.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 8
1
2
3
4
5
6
7
8
9
(17) In order to confirm that the trend in the relative energies, of transition structures discussed in the manuscript, is not an artifact of the meth-
od used, we computed TS5bSE, TS5aSE, TS5aRE and TS5aRZ using mPW1K/6-31G* calculations followed by single-point energy calculations
using mPW1K/6-31+G**. Additionally, we performed full transition state optimizations using B3LYP/6-31G* PCM (ethylacetate) calculations
and B3LYP/gen calculations (with the ‘high layer’ of the ONIOM scheme (Figure 2) computed using B3LYP/6-31+G** and the ‘low layer’ using
B3LYP/6-31G*) for these four transition structures. All these calculations revealed similar energy trends to those reported in the manuscript, with
TS5bSE being the lowest energy transition structure in all methods employed. See Supporting Information for full details of these calculations. We
chose B3LYP/6-31G* because the energies were in closest agreement to experimental enantioselectivity, and because we had performed an ex-
haustive exploration of transition state assemblies using this method.
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
(18) Anisimov, V.; Paneth, P. J. Math. Chem. 1999, 26, 75.
(19) Frequencies were scaled by 0.9614; Scott, A. P.; Radom, L. J. Phys. Chem. 1996, 100, 16502.
(20) Bell, R. P. The tunnel effect in chemistry; Chapman & Hall: London, 1980.
(21) Taylor, M. S.; Jacobsen, E. N. Angew. Chem., Int. Ed. 2006, 45, 1520–1543.
(22) See Supporting Information for a detailed interpretation of this KIE.
(23) For the protonated catalyst calculations, addition of a PCM solvent model for ethyl acetate to the B3LYP/6-31G* calculations increased the
CH…O distance to 2.43 Å. Adding polarization and diffuse functions (6-31+G**, gas phase) further increased this distance to 2.46 Å. These values
are more in line with the crystal structure – however, employing a PCM model with a higher basis set is prohibitive for transition state calculations.
Table of Contents Graphic
8
ACS Paragon Plus Environment