Kinetic resolution of unsaturated amides in a chlorocyclization reaction: Concomitant enantiomer differentiation and face selective alkene chlorination by a single catalyst

Article abstract of DOI:10.1021/ja407241d

The first example of a kinetic resolution via chlorofunctionalization of olefins is reported. The enantiomers of racemic unsaturated amides were found to have different hydrogen-bonding affinities for chiral Lewis bases in numerous solvents. This interaction was exploited in developing a kinetic resolution of racemic unsaturated amides via halocyclization. The same catalyst serves to both sense chirality in the substrate as well as mediate a highly face-selective chlorine delivery onto the olefin functionality, resulting in stereotriad products in up to 99:1 dr and up to 98.5:1.5 er. The selectivity factors were typically greater than 50 to allow for the simultaneous synthesis of both the products and unreacted substrates in highly enantioenriched form at yields approaching 50%. The reaction employs catalytic amounts (≤0.50 mol %) of a commercially available and recyclable organocatalyst.

Full text of DOI:10.1021/ja407241d

Article
pubs.acs.org/JACS
Kinetic Resolution of Unsaturated Amides in a Chlorocyclization
Reaction: Concomitant Enantiomer Differentiation and Face
Selective Alkene Chlorination by a Single Catalyst
Arvind Jaganathan, Richard J. Staples, and Babak Borhan*
Department of Chemistry, Michigan State University, East Lansing, Michigan 48824, United States
S
* Supporting Information
ABSTRACT: The first example of a kinetic resolution via
chlorofunctionalization of olefins is reported. The enantiomers
of racemic unsaturated amides were found to have different
hydrogen-bonding affinities for chiral Lewis bases in numerous
solvents. This interaction was exploited in developing a kinetic
resolution of racemic unsaturated amides via halocyclization.
The same catalyst serves to both “sense chirality” in the
substrate as well as mediate a highly face-selective chlorine
delivery onto the olefin functionality, resulting in stereotriad
products in up to 99:1 dr and up to 98.5:1.5 er. The selectivity
factors were typically greater than 50 to allow for the
simultaneous synthesis of both the products and unreacted
substrates in highly enantioenriched form at yields approaching 50%. The reaction employs catalytic amounts (≤0.50 mol %) of a
commercially available and recyclable organocatalyst.
reducing the available reaction pathways for any given reaction
is a worthwhile endeavor.
INTRODUCTION
■
Nature relies on enzyme catalysis to effect formidable reactions
with astonishing efficiency. The remarkable chemo- and
stereospecificity of enzymatic reactions leads to rapid
generation of complex architectures from simple precursors.
Key to the success of these reactions is the extensive
preorganization of the substrate into reactive conformations,
as well as a multitude of covalent and/or noncovalent substrate-
enzyme interactions. The combined effect of differential
entropy and enthalpy modulations serve to distinguish between
numerous reaction pathways that would have all been equally
accessible in the absence of the enzyme.1 Efforts by synthetic
chemists to develop small-molecule catalysts that mimic
enzymes have led to a better appreciation for the power of
enzyme catalysis, as well as the development of novel chemistry
(examples include the polyene cascade reactions that generate
terpenes, the putative cascade cyclization reactions that yield
polyheterocyclic natural products, and products of the
polyketide biosynthesis, to name a few).2 But despite these
advances, biasing reactions exclusively into one of the multitude
of available reaction pathways remains a daunting challenge
within the realm of small-molecule catalysis; this is especially
true when many of these pathways are energetically accessible
under reaction conditions. Unlike enzymes, small molecule
catalysts do not enjoy large decreases in the differential entropy
and enthalpy of activation for mediating reactions.3 Nonethe-
less, small-molecule catalysts are generally more promiscuous in
terms of substrate specificity and are easily manipulated by
means of modular syntheses. Consequently, the pursuit to
discover catalysts that are capable of mimicking enzymes by
Advances in synthetic chemistry have led to robust
“chemical” equivalents for numerous reactions that are
traditionally perceived as the domain of enzyme catalysis.
Arguably, the most mimicked of these transformations is the
kinetic resolution of racemic mixtures, whereby an enzyme can
selectively function on one of two enantiomers.4 The interest in
developing kinetic resolutions goes beyond the academic
interest of enabling proposals or validations of theoretically
predicted “molecular-recognition” phenomena; they often serve
as a practical means for the synthesis of enantioenriched
molecules. Jacobsen and co-workers have elegantly summarized
the criteria for an “ideal kinetic resolution”.5 The numerous
requirements include the ready availability of the catalyst and
substrates, paucity of methods to access the desired chiral
molecules using alternate methods, the cost of the catalyst,
substrate scope, and practical ease in recovery of products and
catalysts. Our lab recently had the occasion to examine the
possibility of a double stereoselection by an organocatalyst in
the context of an asymmetric alkene halogenation reaction.6 In
our prior studies, we had demonstrated that unsaturated amides
can be subjected to a highly enantioselective chlorocyclization
reaction to afford corresponding dihydro-oxazine heterocycles
with excellent stereoselectivities (Figure 1a, typically greater
than 97.5:2.5 er).7,8 These findings indicated that in the
presence of the chiral catalyst the two diastereomeric transition
Received: July 15, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja407241d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 1. Enantioselective chlorocyclization of unsaturated amides. (a) Previous work. (b) Proposed stereoselective chlorocyclization of racemic
unsaturated amides: Reagent or substrate control may predominate. (c) Mechanistic possibilities for the transformation: two distinct stereoselective
steps can lead to a kinetic resolution of substrate leading to the diastereo- and enantioselective synthesis of stereotriad products.
states are sufficiently well differentiated energetically to allow
for high levels of enantioinduction.
intrigued by the possibility of using a single chiral catalyst
that could promote the formation of predominantly one
diastereomer of the product with high enantioselectivity. In
order to meet this challenge, the catalyst must serve to “sense
chirality” of the substrate in the first stage and enable highly
face selective chlorenium delivery to the “matched” catalyst-
bound substrate in the second stage. If the catalyst can
selectively accelerate the reaction of one of the two
enantiomers (i.e., promote a kinetic resolution of the
racemate), then only two of the four pathways will be
accessible. The two remaining pathways are defined by the
face selectivity in the chlorenium delivery to the olefin moiety.
If the same catalyst could further enable a highly face selective
chlorenium delivery to the fast reacting enantiomer of the
substrate, then one would expect a single stereoisomer of the
cyclized product. It is evident that high levels of stereoselection
at two stages must operate in concert to enable a
diastereoselective kinetic resolution. This enzyme-like catalysis
will be manifested as a kinetic resolution of the racemate that
also leads to stereotriad products with high diastereoselectivity.
Kinetic resolution via alkene halofunctionalization is
captivating for many reasons. First, the products are stereo-
triads (with two new stereocenters being created in the
process) adorned with numerous functional handles for further
elaboration. Resolutions that create additional stereocenters in
the products are relatively rare and are challenging due to the
additional requirement of controlling the product diastereose-
lectivity.11 Second, the kinetic resolution would necessitate the
discovery of an intermolecular catalyst-substrate interaction that
will serve to “sense” chirality of racemic amides. Finally, a
kinetic resolution via chlorofunctionalization of olefins was an
unrealized transformation prior to this work.8e,x,12 It must be
emphasized that high face selectivity in the delivery of a
halenium ion on to the olefin moiety is neither a necessary nor
Nonetheless, in the presence of pre-existing chirality in the
substrate (Figure 1b), the challenge of controlling the absolute
stereochemistry of products is compounded by the additional
requirement of controlling the relative stereochemistry of the
two newly created stereocenters. Furthermore, the propensity
of chloronium ions to readily isomerize to the corresponding
carbocations9 followed subsequently by a “syn” or “anti”
capture of the carbocation by the pendant amide nucleophile
results in two additional diastereomeric transition states. The
capacity to achieve the required level of selectivity with the
catalyst system was in doubt, until recent mechanistic studies
with an analogous reaction system, which demonstrated the
ability of the catalytic complex to not only dictate face
selectivity of chlorenium transfer, but also exquisitely control
the stereochemistry of cyclization.10 The capability to control
the stereochemical outcome of two independent events by the
same catalytic system piqued our interest in challenging this
system to further stereodifferentiate a racemic substrate as a
result of preferential binding, much like that in enzymatic
systems.
The chlorocyclization of a generic chiral amide, such as the
one depicted in Figure 1c, can yield eight distinct stereo-
isomeric products E1-aa to E2-sa in the presence of a chiral
catalyst. Preliminary studies revealed that there was no evidence
of the “syn” opening pathway for both the noncatalyzed or the
catalyzed chlorocyclization of racemic substrates (i.e., E1-sa to
E2-sa, which would arise from a “syn” opening were not
detected in the reaction). Also, there was no inherent bias for
the pathways that lead to the anti-syn (E1-as + E2-as) and the
anti-anti (E1-aa + E2-aa) diastereomers at ambient temperature
(noncatalyzed reactions had an anti-syn:anti-anti ratio of
∼60:40; see the Supporting Information, SI). We were
B
dx.doi.org/10.1021/ja407241d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Table 1. Catalyst Screen for Kinetic Resolution in the Chlorocyclization of Unsaturated Amides
a
b
c
c
c
entry
catalyst (equiv)
temp/time(°C/h)
Cl+ source
conv. (%)
dr (2a:3a)
er (2a)
er (3a)
er (S)-1a
1
2
3
4
5
6
7
8
9
none
4/4.5
4/4.5
4/4.5
4/4.5
4/4.5
−30/1
−30/1
−30/1
−30/1
24/1
DCDMH
DCDMH
DCDMH
DCDMH
DCDMH
DCDMH
DCDMH
DCDMH
DCDMH
NCS
42
44
47
28
69
49
56
60
61
53
55
48
49
28
55
65:35
71:29
70:30
66:34
44:56
24:76
22:78
18:82
92:8
0.1 A
53:47 (ent-2a)
51:49
51:49
54:46
50:50
0.1 B
50:50
50:50
53:47
70:30 (ent-3a)
52:48
60:40
98:2
0.1 C
51:49
50:50
0.1 D
57:43
55:45
0.1 E
54:46 (ent-2a)
62:38 (ent-2a)
62:38
52:48
0.1 F
53:47
0.1 G
59:41
0.1 H
94:6
97:3
d
d
d
d
d
d
,
,
,
,
,
,
e
e
e
e
e
e
10
11
12
13
14
15
0.03 H
0.03 H
0.005 H
0.0025 H
0.0001 H
quasi-ent-H
96:4
97:3
90:10
92:8
97:3
24/1
NCP
95:5
97:3
98.5:1.5
95.5:4.5
93:7
24/0.17
0/1
NCP
94:6
96.5:3.5
98:2
74:26
72:28
58:42
NCP
94.5:5.5
94:6
0/1
NCP
98.5:1.5
68:32
24/1
NCP
94:6
97:3 (ent-2a)
91:9 (ent-3a)
98:2 (R)-1a
a
b
c
Based on calibrated GC yield of unreacted olefin using undecane as internal standard (see SI). Determined by GC analysis. Determined by chiral
d
e
HPLC. Reaction concentration was 0.10 M. 0.55 equiv of chlorenium source was used.
a sufficient condition for achieving an efficient resolution. The
crucial requirement is for the chiral catalyst to sufficiently
differentiate the reaction rates of the two enantiomers of the
racemate. The face selectivity merely dictates the diastereose-
lectivity of the cyclized products.
and (b) if competent catalysts for a kinetic resolution could be
identified.
Our studies commenced with the evaluation of numerous
hydrogen-bonding catalysts such as BINOL and its derivatives
in the hope that they may preferentially associate with one
enantiomer of the substrate.13 However, the results with
catalysts A and B were disappointing (entries 2 and 3, Table 1).
This was followed by the evaluation of other catalysts that have
shown promise in enantioselective alkene halogenation
reactions in recent years. The bis-amidine derived Bronsted
acid catalyst C pioneered by Johnston and co-workers8d and the
thiocarbamate catalyst D analogous to those used by Yeung’s
group8ab also proved ineffective in resolving the racemic
substrate (entries 4 and 5, Table 1). Reactions A−D did not
exhibit rate acceleration, little to no enantioselectivity, and
similar diastereoselectivity as the noncatalyzed reaction. Better
results were obtained when cinchona alkaloid derived chiral
Lewis bases were screened. Four catalysts (catalysts E-H) that
RESULTS AND DISCUSSION
■
Reaction Optimization. The development of the kinetic
resolution began with the evaluation of the inherent
diastereoselectivity of this transformation in the absence of
any catalyst (i.e., mapping the substrate control). Marginal
preference for the formation of 2a was observed (entry 1, Table
1; 2a:3a varied from 60:40 to 70:30 depending on the
chlorenium source employed; see SI for results with other
chlorenium sources). This was followed by the evaluation of
numerous organocatalysts at 10 mol % catalyst loading and
incomplete conversions in order to determine whether (a)
substrate control could be significantly amplified or overridden
C
dx.doi.org/10.1021/ja407241d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
significantly altered the substrate control (and thereby
confirmed their participation in the stereodetermining step)14
were identified. Catalysts E, F, and G gave the anti-anti
diastereomer 3a as the predominant diastereomer (entries 6−8,
Table 1). But despite overriding substrate control, the
enantioselectivities of the cyclized products and the unreacted
olefin were poor with all three catalysts; i.e., these catalysts were
neither competent in resolving the racemic substrate nor
capable of imparting face selectivity of the chlorenium ion
delivery to the olefin. In sharp contrast, catalyst H
(DHQD2PHAL) served to significantly amplify the substrate
control by affording 2a as the major diastereomer (2a:3a =
92:8, entry 9, Table 1). More importantly, when the reaction
was stopped at 61% conversion, the major diastereomer 2a was
isolated in 94:6 er and the unreacted olefin was highly enriched
in the S-enantiomer (97:3 er).15 While tremendous rate
acceleration was observed with all four catalysts, only
(DHQD)2PHAL was identified as a competent catalyst for
kinetic resolution. Stoichiometric NMR experiments revealed
that (DHQD)2PHAL was likely behaving as a hydrogen
bonding catalyst to promote the kinetic resolution (vide infra).
Further studies focused on improving some of the practical
aspects of this reaction, such as the reaction temperature,
concentration, and catalyst loading. The identity of the
chlorenium source did not significantly influence the diastereo-
and enantioselectivity of this transformation. N-chlorophthali-
mide (NCP) was eventually chosen for further studies due to
the practical ease of removing the phthalimide byproduct using
a basic workup as well as the consistently higher isolated yields
of the cyclized products (entry 11, Table 1) (a detailed account
of results with different chlorenium sources and other reaction
variables’ optimization can be found in the SI). Remarkably, no
loss in efficiency was seen even at ambient temperatures and at
higher concentrations (0.10 M) (entries 12 to 14, Table 1)
while maintaining sub 1.0 mol % catalyst loadings, indicating
exquisite specificity of the catalyst for the transformation of the
R enantiomer of the substrate. The catalyst loading could be
reduced to as low as 0.25 mol % (entry 13, Table 1). The quasi-
enantiomeric (DHQ)2PHAL catalyst gave practically identical
results favoring the opposite enantiomers of products and the
recovered olefin (entry 15, Table 1).
presence of stoichiometric amounts of N-chlorophthalimide
and 0.50 mol % of (DHQD)2PHAL, (R)-1a was almost
completely consumed in 5 min to give exclusively the anti-syn
diastereomer 2a (99:1 dr) indicating exquisite α-face selectivity
in the delivery of the chlorenium ion to the olefin of (R)-1a.
Under identical conditions, (S)-1a showed only 27%
conversion to product. Notably, low olefin face selectivity was
seen in the chlorenium delivery still favoring the α face by a
ratio of 70:30 indicating a much attenuated reagent control for
(S)-1a.
Substrate Scope. The generality of the kinetic resolution
was investigated by subjecting numerous trans disubstituted
allylic amides to optimized reaction conditions. As seen in
Table 2, the reaction was tolerant of electronically and sterically
diverse aryl substituents on the olefin. Electron deficient and
halogenated aryl rings presented no difficulties, nor did
substrates with ortho substituents on the aromatic ring.
These substrates were resolved in comparable efficiency as
the test substrate (entries 1−6, 8, and 9 Table 2). In all of these
instances, the cyclized products were formed with excellent
diastereoselectivity (93:7 to >99:1 dr) at ∼50% conversion.
Furthermore, the enantioselectivity for the major diastereomer
was >95:5 for most products and the unreacted substrates were
isolated in >90:10 er for most substrates (a single crystallization
was usually sufficient to upgrade the enantioselectivity to
>97.5:2.5 er). The effect of increasing steric demand of the
olefin substituent as well as the substituent at the α-carbon (i.e.,
the C4 and C6 substituents of the cyclized products) was then
studied. Substrate 1j with a sterically demanding naphthyl
substituent was efficiently resolved (92:8 er for unreacted
substrate at 55% conversion), albeit the diastereoselectivity of
the cyclized product was poor (entry 10, 74:26 dr). This result
indicates that the catalyst effectively discriminates the
enantiomers of the racemic substrate; but a poor facial
selectivity in the chlorenium delivery to the olefin leads to
the diminished diastereoselectivity.
The effect of increasing steric demand of the substituent at α-
position of the amide had a more dramatic effect. Substrates 1k
and 1l with conformationally flexible allyl and n-C5H11
substituent, respectively, were excellent substrates for the
resolution (entries 11 and 12, Table 2). Much diminished
efficiency was observed when the α-substituent was a
homobenzyl, phenyl, or a t-butyl substituent (entries 13 to
15, Table 2). Cyclization of substrates 1m, 1n, and 1o required
higher catalyst loadings (3.0 mol %) and longer reaction times
(90 min) to reach ∼50% conversion. In the case of substrate 1n
with a C6H5 α-substituent, practically no resolution of the
racemic substrate was seen (1n was recovered in 60:40 er at
55% conversion). The cyclization exhibited poor diastereose-
lectivity (2n/3n = 60:40) and the products exhibited only
moderate levels of enantioselectivity (84:16 er for 2n and 69:31
er for 3n). Likewise, substrate 1o with the bulky tBu substituent
also fared poorly. The lower reaction rates and resolution
efficiency with these substrates is attributed to poor substrate-
catalyst interaction due to increased steric impediments
(stoichiometric substrate-catalyst NMR studies support this
hypothesis; see below for discussion and the SI for the NMR
studies).
Predictably, a clear preference was observed for the
chlorocyclization of enantiomerically pure (R)-1a in compar-
ison to (S)-1a under optimized conditions (Scheme 1). In the
Scheme 1. Observed Difference in Reagent Control by
(DHQD)2PHAL in Mediating the Chlorocyclization of (R)-
1a and (S)-1a
NMR Analysis of Stoichiometric Substrate-Catalyst
Mixtures. As outlined in Figure 1c, the first and crucial level of
selectivity is defined by the catalyst’s ability to preferentially
accelerate (or inhibit) the reaction of one of the two
enantiomers presumably via hydrogen-bonding interactions.
D
dx.doi.org/10.1021/ja407241d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Table 2. Substrate scope for kinetic resolution via chlorocyclization
a
b
c
d
d
entry
substrate
R1
R2
conversion (Yield of 2+3)
dr (2:3)
er (2a−2o)
er (1a−1o)
1
2
3
4
1a
1b
1c
1d
1e
1f
C6H5
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
57% (55%)
55% (54%)
55% (46%)
50% (50%)
42% (36%)
55% (45%)
54% (36%)
55% (52%)
55% (48%)
53% (44%)
54% (51%)
54% (49%)
50% (45%)
55% (52%)
50% (46%)
95:5
97:3
96:4
98.5:1.5
99.5:0.5
99:1
4-Cl−C6H4
4-F−C6H4
2-Cl−C6H4
3-OMe-C6H4
4-CF3−C6H4
4-CH3−C6H4
2-CH3−C6H4
2-F−C6H4
1-naphthyl
C6H5
95.5:4.5
95:5
93.5:6.5
96.5:3.5
98.5:1.5
96:4
97:3
91:9
e
5
>99:1
96:4
85:15
89:11
76:24
97:3
6
f
7
1g
1h
1i
80:20
93:7
82:18
96:4
8
9
93:7
94:6
99:1
f g
,
h
10 ,
1j
74:26
98:2
ND
92:8
11
1k
1l
CH2CHCH2
98:2
88:12
92:8
12
C6H5
n-C5H11
97:3
94:6
f g
,
h
h
13 ,
1m
1n
1o
C6H5
CH2CH2C6H5
C6H5
89:11
60:40
80:20
88:12 (2m)
77:23
60:40
70:30
f g
14 ,
,
C6H5
84:16 (2n)/69:31 (3n)
72:28 (2o) /60:40 (3o)
f g
15 ,
C6H5
t-Bu
a
b
Based on GC yields of the unreacted substrate using undecane as internal standard (see SI). Isolated yield after column chromatography.
c
d
e
f
Determined by crude NMR and/or GC analysis. Determined by chiral HPLC. Reaction was run at −30 °C for 1 h and then at 0 °C for 2 h. 3.0
g
h
mol % catalyst was used. Reaction time was 90 min. Reaction was run in (CF3)2CHOH to improve solubility of substrate.
The absolute magnitudes as well as the relative difference in the
association constants (KaR vs KaS) of the two enantiomers with
the chiral catalyst are crucial in enabling the kinetic resolution.
Implicit is the assumption that only the “bound” enantiomer
will react, regardless of whether it is the stronger or the weaker
bound enantiomer (scenarios where the stronger bound olefin
enantiomer is stereoelectronically incapable of capturing the
halenium ion cannot be ruled out at this stage). Given the
excellent selectivity factors even at ambient temperatures,
stoichiometric mixtures of the test substrate rac-1a and catalyst
(DHQD)2PHAL were evaluated by 1H NMR in CF3CH2OH in
order to elucidate the nature of substrate-catalyst interactions
that lead to differential reaction rates for the two enantiomers
(Figure 2).16 Most of the protons of rac-1a exhibited upfield
shifts suggesting an intimate association of the substrate and the
catalyst. Surprisingly, a clean formation of diastereomeric
complexes (in a 1:1 ratio) was seen even at ambient
temperatures. With enantiomerically pure (R)-1a and (S)-1a,
diastereomerically pure complexes were seen, ruling out the
possibility of nonenantiospecific fluxional processes on a NMR
time scale (see Figure 2a). Diminished or no diastereotopicity
was observed in other deuterated solvents such as CDCl3,
C6D6, acetone-d6, and CD3CN (see SI for spectra). It is perhaps
not surprising that other solvents fare poorly as a reaction
medium.
3.52 ppm after catalyst protonation. Likewise, small downfield
shift is seen for the methylene protons He). This behavior is in
sharp contrast to that seen in CDCl3, where Hd and He have
distinct chemical shifts depending on the protonation state of
the catalyst. Catalyst protonation leads to significant downfield
shifts of Hd and He (Hd shifts from 3.39 ppm to 3.52 ppm; He
protons shift from 2.60 to 2.80 ppm to 2.94−3.20 ppm, see
CDCl3 column in Figure 2b). Furthermore, the chemical shifts
for Hd and He in CF3CH2OH without the benzoic acid additive
are similar to those of the protonated catalyst in CDCl3 (3.52
ppm for Hd and 2.94−3.20 ppm for He); also there is no further
downfield shift in the presence of stoichiometric quantities of
benzoic acid in CF3CH2OH. These results suggest that the
catalyst is protonated in CF3CH2OH even without an external
proton source. The protonation of the quinuclidine nitrogen
atoms may cause a change in the conformation of the catalyst
and allow for better substrate-catalyst interactions in
CF3CH2OH as opposed to other solvents. Although significant,
catalyst protonation cannot be the only role for CF3CH2OH,
since mere incorporation of protic additives in different
solvents does not recapitulate the results with CF3CH2OH.
Besides its enhanced acidity, CF3CH2OH is a good hydrogen
bond donor, a weak nucleophile, a noncoordinating counter-
anion, and also a highly polar solvent. The latter set of features
is not easily duplicated with other solvent systems. In any event,
the protonated catalyst can serve as a hydrogen bond donor to
bind to the amide functional group (whereas the non-
protonated catalyst can only serve as a hydrogen bond
acceptor). These NMR studies support the hypothesis that
stereodiscrimination likely results from asymmetric general acid
catalysis (i.e., hydrogen-bonding catalysis) by protonated
(DHQD)2PHAL in CF3CH2OH, although a chiral Lewis-base
assisted Bronsted acid catalysis (LBBA) by CF3CH2OH cannot
be ruled out at this stage.17
We postulate that the quinuclidine nitrogen atoms of the
chiral catalyst (pKa ≈ 10) are protonated in CF3CH2OH (pKa
of CF3CH2OH = 12.5; mole ratio of catalyst/CF3CH2OH ≥
1:25,000 for reactions and ∼1:1,000 for NMR studies). This
hypothesis is supported by NMR studies that reveal little
change in the chemical shifts of the methylene and methine
protons adjacent to the quinuclidine nitrogen atoms of the
catalyst before and after addition of stoichiometric quantities
(2.0 equiv) of benzoic acid (see right side column in Figure 2b;
methine proton Hd shows negligible shift from 3.49 ppm to
E
dx.doi.org/10.1021/ja407241d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
lead to the formation of diastereomeric mixture of products.
The excellent diastereo- and enantioselectivity for the cyclized
products and recovered substrates at <1 mol % catalyst loading
suggests that high selectivity factors are in operation in addition
to exquisite face selectivity in the chlorenium delivery. In order
to quantify the efficiency of the resolution, Krel values were
calculated for four of these reactions on the basis of the yields
and ers of the cyclized products.18 The resolution of substrate
1a proceeded with a Krel value of 113. Likewise, Krel values of
89, 90, and 56 were calculated for substrates 1b, 1l, and 1k,
respectively. It must also be emphasized that conversions used
in the calculations were based on isolated yields of the products
on 2.0 mmol (∼0.5 g) scale reactions, and therefore, the values
obtained for Krel represent the lower limits. As such, this kinetic
resolution has the potential to simultaneously access the
products and the substrates in highly enantioenriched form
and at yields approaching 50% on a preparative scale.
Most of the reactions were rapid (∼10 min) and run in open
reaction vessels at up to 200 mM concentrations. The
resolution is conveniently scaled to gram quantities with no
detrimental effect on the drs and ers (Scheme 2). The catalyst
was found to be stable to the reaction conditions and was
isolated using routine silica gel chromatography. The catalyst
could be recycled up to three times with negligible loss in
activity (catalyst recovery and recycling studies are detailed in
the SI). Routine hydrolytic and oxidative transformations of the
products and the unreacted substrates are shown in Scheme 2a.
Acid hydrolysis of 2a gave amino alcohol 5. The oxidative
cleavage of the recovered olefin (S)-1a gave protected alanine
6. Taken together with the recyclability of the catalyst and
solvent, this reaction paves the way for efficient, multigram
scale synthesis of densely functionalized chiral building blocks.
Attempts to uncover analogous kinetic resolution phenom-
ena in iodocyclization reactions of the racemic compounds led
to intriguing results. While the resolution was inefficient, the
cyclization exhibited a complementary diastereoselectivity to
the chlorocyclization reaction and favored the formation of the
anti-anti diastereomer in a 93:7 ratio (Scheme 2b).19 The
iodocyclized products 2ab and 4 exhibited low enantioenrich-
ment. Notably, unlike the chlorocyclization reaction, the
noncatalyzed and catalyzed reactions exhibited similar levels
of diastereoselectivity for the iodocyclization (see SI). The
transformation was rapid; no significant rate acceleration or
change in product distribution was observed even at up to 3
mol % catalyst loadings. Taken together with the comple-
mentary diastereoselectivity and negligible reagent control at
ambient temperatures, a tandem one-pot kinetic resolution/
diastereoselective iodocyclization cascade was conceptualized
(Scheme 2c) that would allow for a 100% conversion of the
racemic substrate into densely functionalized sterotriads. rac-1a
was resolved via a chlorocyclization reaction under optimized
reaction conditions; this was followed by the addition of 0.50
equiv of NIS in to the reaction vessel to initiate the
diastereoselective iodocyclization of the enantioenriched
substrate. This tandem protocol gave chlorocyclized product
2a in 97.5:2.5 dr with the major anti-syn diastereomer being
formed in 97:3 er and the iodocyclized product 4 in a 90:10 dr
with the major anti-anti diastereomer being formed in a 92:8 er
(Scheme 2c). Products 2a and 4 were readily separable by
column chromatography and were isolated in 48% and 43%
yields, respectively.
Figure 2. (a) Stoichiometric NMR studies of substrate-catalyst
mixtures in CF3CH2OH at ambient temperature and 0.02 M
concentration of substrate and catalyst: Both enantiomers bind to
the chiral catalyst. (b) Diagnostic quinuclidine proton shifts of free and
protonated catalyst in CDCl3 and CF3CH2OH: Catalyst is likely
protonated in CF3CH2OH even without acid additive.
Numerous characteristics of enzyme-type catalysis is
evidenta remarkable selectivity for binding one enantiomer
of the substrate, extensive preorganization of the substrate-
catalyst-reagent triad leading to rapid reaction rates for the
catalyzed process for the “matched” enantiomer and finally,
exquisite levels of face selectivity in the delivery of the
chlorenium ion on to the olefin functionality of the substrates.
An in-depth analysis of the enthalpic and entropic drivers for
this transformation is currently underway.
Features and Utility of the Resolution. A number of
aspects of this kinetic resolution warrant emphasis. Kinetic
resolutions are seldom used for the synthesis of the
enantioenriched products (as opposed to the unreacted
substrate in enantioenriched form). This is because Krel values
(selectivity factors) of ∼50 are required to obtain products with
>95:5 ers and in yields approaching 50%. For example, for a Krel
value of 10, the theoretical maximum er for the products even at
conversions as low as 10% is only ∼90:10. At 65% conversion
for the same reaction, the unreacted substrate can be recovered
in 97:3 er and 35% yield. The resolution presented here can
F
dx.doi.org/10.1021/ja407241d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
a
Scheme 2. Synthetic Utility of the Kinetic Resolution
a
(a) Gram-scale reaction and transformation of cyclized products and recovered substrate. (b) Diastereoselectivity of analogous iodocyclization
reaction under catalyzed and non-catalyzed conditions. (c) Development of a one-pot kinetic resolution-diastereoselective iodocyclization reaction.
CONCLUSIONS
ASSOCIATED CONTENT
■
■
S
In summary, the first example of a highly diastereoselective
kinetic resolution in a chlorofunctionalization reaction of
olefins has been developed. Two distinct and highly stereo-
selective events mediated by the same catalyst are crucial to the
success of this reaction. The catalyst not only ensures rate
acceleration for one of the two enantiomers of a chiral
substrate, but also serves to impart exceptional stereoselectivity
in the alkene chlorination and cyclization events, resulting in
stereotriad products with excellent relative and absolute
stereocontrol starting from easily accessed racemic starting
materials. Low catalyst loadings, ambient reaction temperatures,
open reaction vessels, short reaction times, and recyclability of
the catalyst are some of the features of this chemistry. The Krel
values for many substrates are sufficiently high to permit the
use of the resolution protocol for the simultaneous synthesis of
products and substrates in highly enantioenriched form. NMR
studies have uncovered a substrate-catalyst hydogen-bonding
interaction as a potentially key molecular recognition event in
enabling the resolution; these studies have also hinted at the
transformation of a chiral Lewis base catalyst into a hydrogen-
bonding catalyst in CF3CH2OH, thereby opening the doors to
hitherto unknown modes of activation for this class of catalyst.
The results serve as a starting point for detailed mechanistic
studies. Efforts to further improve the scope and utility of this
transformation and its application in natural products’
syntheses are currently being pursued.
* Supporting Information
Detailed experimental procedures, optimization, and character-
ization of all new compounds. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
babak@chemistry.msu.edu
Author Contributions
The manuscript was written through contributions of all
authors.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was funded by NSF Grant CHE-0957462. The
authors thank the MSU mass spectrometry facility for HR-MS
analyses. Ms. Yanmen Yang and Dr. Daniel Holmes are
acknowledged for experimental assistance and NMR assistance
respectively. A.J. thanks the College of Natural Sciences at
MSU for the Tracy Hammer graduate student award. A.J. was a
Harold Hart Endowed Fellow in the Chemical Sciences at
MSU during the commencement of this project (June−
September 2012).
G
dx.doi.org/10.1021/ja407241d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(e) Olah, G. A.; Westerman, P. W.; Melby, E. G.; Mo, Y. K. J. Am.
Chem. Soc. 1974, 96, 3565. (f) Solling, T. I.; Radom, L. Int. J. Mass
Spectrom. 1999, 185, 263.
REFERENCES
■
(1) (a) Gao, J.; Ma, S.; Major, D. T.; Nam, K.; Pu, J.; Truhlar, D. G.
Chem. Rev. 2006, 106, 3188. (b) Warshel, A.; Sharma, P. K.; Kato, M.;
Xiang, Y.; Liu, H.; Olsson, M. H. M. Chem. Rev. 2006, 106, 3210.
(2) For a review of enzymatic and biomimetic polyene cascades
leading to terpene skeletons, see: (a) Yoder, R. A.; Johnston, J. N.
Chem. Rev. 2005, 105, 4730. For application of cascade reactions in
natural product synthesis, see: (b) Nicolaou, K. C.; Edmonds, D. J.;
Bulger, P. G. Angew. Chem., Int. Ed. 2006, 45, 7134.
(3) Knowles, R. R.; Jacobsen, E. N. Proc. Natl. Acad. Sci. 2010, 107,
20678.
(4) (a) Pellissier, H. Adv. Synth. Catal. 2011, 353, 1613. (b) Robinson,
D. E. J. E.; Bull, S. D. Tetrahedron: Asymmetry 2003, 14, 1407.
(5) Keith, J. M.; Larrow, J. F.; Jacobsen, E. N. Adv. Synth. Catal. 2001,
343, 5.
(6) (a) Castellanos, A.; Fletcher, S. P. Chem.Eur. J. 2011, 17, 5766.
(b) Chen, G.; Ma, S. Angew. Chem., Int. Ed. 2010, 49, 8306.
(c) Denmark, S. E.; Kuester, W. E.; Burk, M. T. Angew. Chem., Int. Ed.
2012, 51, 10938. (d) Hennecke, U. Chem. Asian J. 2012, 7, 456.
(e) Murai, K.; Fujioka, H. Heterocycles 2013, 87, 763.
(7) (a) Jaganathan, A.; Garzan, A.; Whitehead, D. C.; Staples, R. J.;
Borhan, B. Angew. Chem., Int. Ed. 2011, 50, 2593. (b) Garzan, A.;
Jaganathan, A.; Salehi Marzijarani, N.; Yousefi, R.; Whitehead, D. C.;
Jackson, J. E.; Borhan, B. Chem.Eur. J. 2013, 19, 9015. (c) White-
head, D. C.; Yousefi, R.; Jaganathan, A.; Borhan, B. J. Am. Chem. Soc.
2010, 132, 3298.
(8) (a) Chen, J.; Zhou, L.; Yeung, Y.-Y. Org. Biomol. Chem. 2012, 10,
3808. (b) Chen, Z.-M.; Zhang, Q.-W.; Chen, Z.-H.; Li, H.; Tu, Y.-Q.;
Zhang, F.-M.; Tian, J.-M. J. Am. Chem. Soc. 2011, 133, 8818.
(c) Denmark, S. E.; Burk, M. T. Org. Lett. 2012, 14, 256. (d) Dobish,
M. C.; Johnston, J. N. J. Am. Chem. Soc. 2012, 134, 6068. (e) Fang, C.;
Paull, D. H.; Hethcox, J. C.; Shugrue, C. R.; Martin, S. F. Org. Lett.
2012, 14, 6290. (f) Haas, J.; Bissmire, S.; Wirth, T. Chem.Eur. J.
2005, 11, 5777. (g) Hennecke, U.; Muller, C. H.; Frohlich, R. Org.
Lett. 2011, 13, 860. (h) Huang, D.; Wang, H.; Xue, F.; Guan, H.; Li,
L.; Peng, X.; Shi, Y. Org. Lett. 2011, 13, 6350. (i) Jiang, X.; Tan, C. K.;
Zhou, L.; Yeung, Y.-Y. Angew. Chem., Int. Ed. 2012, 51, 7771. (j) Kang,
S. H.; Lee, S. B.; Park, C. M. J. Am. Chem. Soc. 2003, 125, 15748.
(k) Kwon, H. Y.; Park, C. M.; Lee, S. B.; Youn, J.-H.; Kang, S. H.
Chem.Eur. J. 2008, 14, 1023. (l) Li, H.; Zhang, F.-M.; Tu, Y.-Q.;
Zhang, Q.-W.; Chen, Z.-M.; Chen, Z.-H.; Li, J. Chem. Sci. 2011, 2,
1839. (m) Lozano, O.; Blessley, G.; del Campo, T. M.; Thompson, A.
L.; Giuffredi, G. T.; Bettati, M.; Walker, M.; Borman, R.; Gouverneur,
V. Angew. Chem., Int. Ed. 2011, 50, 8105. (n) Murai, K.; Matsushita, T.;
Nakamura, A.; Fukushima, S.; Shimura, M.; Fujioka, H. Angew. Chem.,
Int. Ed. 2010, 49, 9174. (o) Murai, K.; Nakamura, A.; Matsushita, T.;
Shimura, M.; Fujioka, H. Chem.Eur. J. 2012, 18, 8448. (p) Nicolaou,
K. C.; Simmons, N. L.; Ying, Y.; Heretsch, P. M.; Chen, J. S. J. Am.
Chem. Soc. 2011, 133, 8134. (q) Paull, D. H.; Fang, C.; Donald, J. R.;
Pansick, A. D.; Martin, S. F. J. Am. Chem. Soc. 2012, 134, 11128.
(r) Phipps, R. J.; Hiramatsu, K.; Toste, F. D. J. Am. Chem. Soc. 2012,
134, 8376. (s) Rauniyar, V.; Lackner, A. D.; Hamilton, G. L.; Toste, F.
D. Science 2011, 334, 1681. (t) Tan, C. K.; Le, C.; Yeung, Y.-Y. Chem.
Commun. 2012, 48, 5793. (u) Tan, C. K.; Zhou, L.; Yeung, Y.-Y. Org.
Lett. 2011, 13, 2738. (v) Veitch, G. E.; Jacobsen, E. N. Angew. Chem.,
Int. Ed. 2010, 49, 7332. (w) Wang, Y.-M.; Wu, J.; Hoong, C.; Rauniyar,
V.; Toste, F. D. J. Am. Chem. Soc. 2012, 134, 12928. (x) Wilking, M.;
Muck-Lichtenfeld, C.; Daniliuc, C. G.; Hennecke, U. J. Am. Chem. Soc.
2013, 135, 8133. (y) Zhang, W.; Liu, N.; Schienebeck, C. M.; Decloux,
K.; Zheng, S.; Werness, J. B.; Tang, W. Chem.Eur. J. 2012, 18, 7296.
(z) Zhang, W.; Zheng, S.; Liu, N.; Werness, J. B.; Guzei, I. A.; Tang,
W. J. Am. Chem. Soc. 2010, 132, 3664. (aa) Zhou, L.; Chen, J.; Tan, C.
K.; Yeung, Y.-Y. J. Am. Chem. Soc. 2011, 133, 9164. (ab) Zhou, L.; Tan,
C. K.; Jiang, X.; Chen, F.; Yeung, Y.-Y. J. Am. Chem. Soc. 2010, 132,
15474.
(10) Unpublished mechanistic studies of the analogous chlorolacto-
nization reaction have uncovered strong evidence for the intermediacy
of β-chlorocarbenium ions as intermediates in this chemistry.
(11) (a) Corey, E. J.; Noe, M. C.; Guzman-Perez, A. J. Am. Chem. Soc.
1995, 117, 10817. (b) Yun, J.; Buchwald, S. L. J. Org. Chem. 2000, 65,
767. (c) Brimble, M. A.; Johnston, A. D. Tetrahedron-Asymmetry 1997,
8, 1661. (d) Xing, X.; Zhao, Y.; Xu, C.; Zhao, X.; Wang, D. Z.
Tetrahedron 2012, 68, 7288.
(12) Ikeuchi, K.; Ido, S.; Yoshimura, S.; Asakawa, T.; Inai, M.;
Hamashima, Y.; Kan, T. Org. Lett. 2012, 14, 6016.
(13) (a) Brunel, J. M. Chem. Rev. 2005, 105, 857. (b) Chen, Y.;
Yekta, S.; Yudin, A. K. Chem. Rev. 2003, 103, 3155.
(14) Denmark, S. E.; Burk, M. T. Proc. Natl. Acad. Sci. 2010, 107,
20655.
(15) The absolute stereochemistry of 2a and 3a was determined by
single crystal X-ray diffraction (CCDC 948806−948807 are the
corresponding Cambridge Structural Database deposition numbers for
the crystal structures of 2a and 3a).
(16) No-D 1H NMR were run in CF3CH2OH using gradient
shimming and solvent suppression; see SI for details.
(17) For a leading reference of chiral Lewis base assisted Bronsted
acid catalysis, see: Sakakura, A.; Sakuma, M.; Ishihara, K. Org. Lett.
2011, 13, 3130.
(18) A straightforward calculation of selectivity factors is precluded
by the formation of a diastereomeric mixture of products; although
such calculations have been used previously for analogous kinetic
resolutions, they may lead to significant errors if contributions of the
minor diastereomer are ignored or when diastereoselectivities are not
very high. See SI for a brief discussion and calculation of Krel values for
selected transformations.
(19) The bromocyclization reaction with NBS had a 57:43
diastereoselectivity marginally favoring the anti-syn diastereomer.
(9) (a) Ohta, B. K.; Hough, R. E.; Schubert, J. W. Org. Lett. 2007, 9,
2317. (b) Olah, G. A.; Bollinger, M. J. Am. Chem. Soc. 1968, 90, 947.
(c) Olah, G. A.; Bollinger, M.; Brinich, J. J. Am. Chem. Soc. 1968, 90,
6988. (d) Olah, G. A.; Peterson, P. E. J. Am. Chem. Soc. 1968, 90, 4675.
H
dx.doi.org/10.1021/ja407241d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX