Palladium-N-heterocyclic carbene (NHC)-catalyzed asymmetric synthesis of indolines through regiodivergent C(sp3)-H activation: Scope and DFT study

Article abstract of DOI:10.1002/chem.201403985

Two bulky, chiral, monodentate N-heterocyclic carbene ligands were applied to palladium-catalyzed asymmetric C-H arylation to incorporate C(sp³)-H bond activation. Racemic mixtures of the carbamate starting materials underwent regiodivergent reactions to afford different trans-2,3- substituted indolines. Although this C_Ar-C_alkyl coupling requires high temperatures (140-160°C), chiral induction is high. This regiodivergent reaction, when carried out with enantiopure starting materials, can lead to single structurally different enantiopure products, depending on the catalyst chirality. The C-H activation at a tertiary center was realized only in the case of a cyclopropyl group. No C-H activation takes place alpha to a tertiary center. A detailed DFT study is included and analyses of methyl versus methylene versus methine C-H activation is used to rationalize experimentally observed regio- and enantioselectivities.

Full text of DOI:10.1002/chem.201403985

DOI: 10.1002/chem.201403985
Full Paper
&
Asymmetric Catalysis |Hot Paper|
Palladium–N-Heterocyclic Carbene (NHC)-Catalyzed Asymmetric
Synthesis of Indolines through Regiodivergent C(sp3)ÀH
Activation: Scope and DFT Study
Dmitry Katayev,[a] Evgeny Larionov,[a] Masafumi Nakanishi,[a] Cꢀline Besnard,[b] and
E. Peter Kꢁndig*[a]
Abstract: Two bulky, chiral, monodentate N-heterocyclic car-
bene ligands were applied to palladium-catalyzed asymmet-
ric CÀH arylation to incorporate C(sp3)ÀH bond activation.
Racemic mixtures of the carbamate starting materials under-
went regiodivergent reactions to afford different trans-2,3-
substituted indolines. Although this CArÀCalkyl coupling re-
quires high temperatures (140–1608C), chiral induction is
high. This regiodivergent reaction, when carried out with
enantiopure starting materials, can lead to single structurally
different enantiopure products, depending on the catalyst
chirality. The CÀH activation at a tertiary center was realized
only in the case of a cyclopropyl group. No CÀH activation
takes place alpha to a tertiary center. A detailed DFT study is
included and analyses of methyl versus methylene versus
methine CÀH activation is used to rationalize experimentally
observed regio- and enantioselectivities.
Introduction
number of CÀH bonds in organic molecules and their high
bond energies, the biggest challenge in all direct CÀH func-
tionalization approaches is selectivity. The ortho-directing
groups have been widely used in CArÀH functionalization and
the avoidance of steric congestion is a dominant factor in
other reactions. The catalytic step for the formation of
a carbon–metal bond from a carbon–hydrogen bond generally
requires the assistance of an internal base; carboxylate ligands
are used frequently. The process is described as a carboxylate-
mediated concerted metalation–deprotonation (CMD) path-
way.[4]
Transition-metal-catalyzed CÀH functionalization has emerged
as a powerful tool in organic chemistry. Its primary impact is
that it enables heretofore inert CÀH groups to become func-
tional entities, and thereby, rendering prefunctionalization of
starting materials redundant. This reduces the number of steps
and opens the way to new strategies in synthesis. A number
of approaches have strongly contributed to the success. The
CÀH bond functionalization can take place through an outer-
or inner-sphere mechanisms. The former includes CÀH inser-
tions of metal–carbene species[1] and oxidations by metal–oxo
complexes.[2] The latter involves the direct insertion of a transi-
tion metal into a CÀH bond, leading to the formation of an or-
ganometallic intermediate (MÀC bond). Long limited to CArÀH
bonds and with the majority of examples being heteroatom-di-
rected coordination/activation processes,[3] recent results have
demonstrated that inner-sphere processes are also suitable for
the functionalization of C(sp3)ÀH bonds. Given the large
Recent reviews showed that the vast majority of these inner-
sphere C(sp3)ÀH activation reactions involved allylic- or benzyl-
ic CÀH bonds; those alpha to a heteroatom or methyl CÀH
bonds.[4e,5,6] The functionalization of unactivated methylene
CÀH bonds through metal insertion is very rare. These are,
however, arguably the most interesting cases because in RÀ
CH2ÀR’ systems the hydrogen atoms are enantiotopic. Asym-
metric catalytic CÀH functionalization in such systems is a very
ambitious project and a worthwhile challenge. A literature
precedent for a PdÀPCy3/pivalate-catalyzed (Cy=cyclohexyl)
intramolecular methylene CÀH/aryl coupling to afford 2-substi-
tuted and 2,3-trans-fused indolines attracted our attention.[7]
Subsequently, we and others demonstrated that this reaction
could be carried out to give highly enantioenriched products.
The key to success in this asymmetric methylene coupling in
our hands was the use of monodentate chiral N-heterocyclic
carbene (NHC) ligands (Scheme 1).[8] An impressively wide
array of highly enantioenriched, substituted indolines and aza-
indolines was made available by this route.
[a] D. Katayev,+ Dr. E. Larionov,+ Dr. M. Nakanishi,+ Prof. Dr. E. P. Kꢀndig
Department of Organic Chemistry, University of Geneva
30 Quai Ernest Ansermet, 1211Geneva 4 (Switzerland)
Fax: (+)4122-237-3215
E-mail: peter.kundig@chiorg.unige.ch
[b] Dr. C. Besnard
Laboratoire de Crystallographie, University of Geneva
Quai Ernest Ansermet 24, 1211 Geneva 4 (Switzerland)
[+] These authors contributed equally to this work.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201403985. It contains experimental, spec-
troscopic, analytical, computational, and X-ray data.
The groups of Kagan[9] and Cramer[10] chose the chiral phos-
phine ligands Duphos and Sagephos, respectively; the latter
Chem. Eur. J. 2014, 20, 15021 – 15030
15021
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
methylene units (Table 1, entry 1).[13] In this case, product (R)-
3a arises from the matched case of the reaction of substrate
(S)-1a and catalyst PdÀ(S,S)-4. It is formed in higher yield, but
lower enantiomeric excess (ee) than that of (2R,3S)-2a because
the reaction of (R)-1a and PdÀ(S,S)-4 is the mismatched pair
and gives not only (2R,3S)-2a, but also (S)-3a; thus lowering
the ee value of 3a. It is characteristic that, although the major
isomer is obtained in varying enantiomeric enrichment, the
minor isomer is formed uniformly in very high enantiomeric
purity with only a single exception (2g). The difference in yield
between the two indolines can be very small. The first example
is the reaction of 1b, with which CÀH insertion can occur
either in the methyl group (to give 3b) or in the benzyl meth-
ylene group (to give 2b). The enantiomer recognition by the
chiral catalyst is almost perfect here. The catalyst (S,S)-5 reacts
with (S)-1b to give, in approximately equal proportion, two dif-
ferent, highly enantioenriched indolines (Table 1, entry 2).[13] It
was of interest then to see the influence of a para-substituent
on the aryl ring on the reaction outcome. These data show
that 4-Me, -OMe, and -CF3 uniformly provided excellent yields
and selectivities (Table 1, entries 3–5). An erosion of yield and/
or selectivity is apparent with 4-F- and 4-NO2-substituted
benzyl substrates (Table 1, entries 6 and 7). High discrimination
was also observed with substrates that incorporated 2- and 3-
aryl substituents, with the exception of 2-C6H4OMe, for which
sterics may be unfavorable (Table 1, entries 8–10). Finally, in
the aromatic series, we note the regiodivergent reaction also
performs well with substrate 1k, which incorporates a furyl
group (Table 1, entry 11). The data in entries 12 and 13 in
Table 1 show the complete absence of products 2l and 2m,
but a high yield of 3l and 3m, albeit in low enantiomeric
purity. Indolines 2l and 2m are not formed because CÀH acti-
vation does not tolerate branching in the alpha position of the
methylene group. In these two cases, we see that methyl acti-
vation of both enantiomers takes place. The enantiomeric en-
richment of 3l points to a degree of kinetic resolution in the
transformation. Table 1, entries 14–19, list RRM results of the
competition of a methyl group with chains that contain heter-
oatoms or an alkene functionality. We see throughout good to
excellent divergent reactions for the two enantiomers in the
racemic mixture, with the exception of acetyl-containing sub-
strate 3q (Table 1, entry 17) and the reaction of 1s that pro-
duced 3s with excellent enantioselectivity, but gave indene,
rather than 2, presumably because of methanethiol elimina-
tion. Entries 20 and 21 in Table 1 demonstrate that this extraor-
dinary selectivity can also hold for reactions in which two dif-
ferent methylene groups are pitched against each other.
Scheme 1. Indoline synthesis through asymmetric C(sp3)ÀH activation.
outperformed the former. Related research, aimed at substitut-
ed chiral indanes, was published by Baudoin et al.[11]
Experimental studies showed that a palladium-coordinated
carboxylate ligand was important for the reaction and compu-
tational analysis pointed to an inner-sphere, carboxylate-assist-
ed CMD pathway analogous to that formulated previously for
CArÀH functionalization. The CMD step is rate and enantioselec-
tivity determining in the [Pd(NHC)*]-catalyzed reaction.[12] Stoi-
chiometric studies showed that acetate and pivalate per-
formed best, but that even carbonate could play the role of in-
ternal base, albeit less efficiently.[12]
In probing the scope and limitations of this transformation,
we found that the chiral NHC/Pd catalysts reacted with racemic
starting materials by regiodivergent reaction of a racemic mix-
ture (RRM; Scheme 2).[13,14]
Scheme 2. The RRM through C(sp3)ÀH activation.
Herein, we report on our findings on the scope and limita-
tions in the synthesis of enantioenriched indolines through the
RRM. We also demonstrate its use in synthesis with enantio-
pure starting materials, for which single compounds rather
than mixtures can be obtained in high enantiomeric purity. In
the second part of this article, we describe a DFT study to
probe the CMD step of this transformation and analyze the re-
activity of methyl versus methylene groups for the rationaliza-
tion of the selectivity observed in regiodivergent RRM.
As pointed out above, the CÀH activation reported herein
does not tolerate branching in the alpha position of the meth-
ylene group (Table 1, entries 12 and 13), nor does it engage
CÀH bonds in the tertiary positions. One exception is the cy-
clopropyl-containing substrate 1v, which undergoes the regio-
divergent reaction with the formation of products (R)-2v and
(R)-3v. Compound (S)-2v is formed through activation of the
cyclopropyl methine CÀH bond in (S)-1v and (R)-3v is formed
through activation of the CH3 group in (R)-1v (Scheme 3). Me-
thine CÀH activation has literature precedent.[10b,c16]
Results and Discussion
Starting materials 1a–v were prepared by using either reduc-
tive amination of ketones with 2-bromoaniline or Buchwald–
Hartwig N-arylation of primary amines with 1,2-dibromoben-
zene[15] followed by carbamation with methyl chloroformate.
We previously found that substrate 1a underwent an arylation
reaction through CÀH bond activation of both methyl and
Chem. Eur. J. 2014, 20, 15021 – 15030
www.chemeurj.org
15022
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
enantioenriched products can
be obtained either by using
the same catalyst on different
enantiomers of the starting
material or by switching the
chirality of the catalyst. The
former is demonstrated by
Table 1. Substrate scope for the synthesis of enantioenriched indolines through the RRM.
the
example
given
in
Scheme 4 and the latter is
shown in Table 2. The requi-
site enantiopure substrates
1b, u, and w–y were synthe-
sized through palladium-cata-
lyzed Buchwald–Hartwig N-
arylation of primary amines
with 1,2-dibromobenzene[15]
followed by carbamation with
methyl chloroformate.
Entry[a]
Starting
material
R1
R2
Ligand precursor
NHCÀHI
Product 2[b]
(yield, ee [%])
Product 3[b]
(yield, ee [%])
Entries 1–8
in
Table 2
1[13]
2[13]
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21[d]
1a
1b
1c
1d
1e
1 f
1g
1h
1i
1j
1k
1l
1m
1n
1o
1p
1q
1r
Me
Ph
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Ph
Me
(S,S)-4
(S,S)-5
(R,R)-5
(R,R)-5
(R,R)-5
(R,R)-5
(R,R)-5
(R,R)-5
(R,R)-5
(R,R)-5
(R,R)-5
(S,S)-4
(S,S)-4
(S,S)-4
(S,S)-4
(S,S)-4
(S,S)-4
(S,S)-4
(S,S)-4
(S,S)-4
(S,S)-4
(2R,3S)-2a (38, 99)
(2R,3S)-2b (44[c], 98)
(2S,3R)-2c (46, 99)
(2S,3R)-2d (49, 99)
(2S,3R)-2e (49, 98)
(2S,3R)-2 f (32, 98)
(2S,3R)-2g (18, 68)[c]
(2S,3R)-2h (38, 99)
(2S,3S)-2i (46, 98)
(2S,3R)-2j (39, 99)
(2S,3R)-2k (42, 98)
(2R)-3a (57, 77)
(2R)-3b (49[c], 95)
(2S)-3c (48, 94)
(2S)-3d (50, 93)
(2S)-3e (49, 96)
(2S)-3 f (35, 89)
(2S)-3g (46, 93)[c]
(2S)-3h (60, 66)
(2R)-3i (49, 93)
(2S)-3j (42, 94)
(2S)-3k (45, 97)
(2R)-3l (81, 25)[c]
(2R)-3m (86, 9)[c]
(2R)-3n (51, 70)
(2S)-3o (51, 85)[c]
(2S)-3p (45, 78)[c]
depict competitive methylene
CÀH coupling reactions with
extraordinarily high regio-
and diastereoselectivity. The
products are entirely under
the control of the catalyst.
Reactions with the (R,R)-cata-
lyst led to the asymmetric
4-C6H4Me
4-C6H4(OMe)
4-C6H4(CF3)
4-C6H4F
4-C6H4(NO2)
2-C6H4(OMe)
2-C6H4F
3-C6H4(OMe)
2-furyl
cHex
iPr
prenyl
CH2OMe
CH2OMOM
CH2OAc
CH2SMe
SMe
functionalization
of
[f]
–
–
[f]
a CmethyleneÀH bond adjacent
to the OMe group to afford
products 2u, w, and x,
whereas the (S,S)-catalyst en-
gaged a CmethyleneÀH bond of
the alkyl side chain to afford
products 3u, w, and x. En-
tries 7 and 8 in Table 2 lead
to analogous highly selective
coupling reactions that in-
volve either a CH3 group or
the CH2 group adjacent to
(2R,3S)-2n (29, 99)
(2R,3S)-2o (35, 99)[c]
(2R,3S)-2p (44, 99)[c]
(2R,3S)-2q (42, 96)[c]
(2R,3S)-2r (29, 95)[c]
[f]
–
(2S)-3r (56, 74)[c]
(2S)-3s (45, 97)[c,e]
(2R,3S)-3t (47, 90)
(2S,3S)-3u (40, 88)
[f]
1s
1t
1u
–
Me
OMe
(2R,3S)-2t (43, 99)
(2R,3S)-2u (40, 96)
[a] Conditions: substrate (0.2 mmol), NHCÀHI (10 mol%), [{Pd(p-cinnamyl)Cl}2] (5 mol%), cesium pivalate (1 equiv),
1
and Cs2CO3 (1.5 equiv) were used in dry xylenes at 1408C for 24 h. [b] Determined by H NMR spectroscopic analy-
sis of an isolated mixture 2/3. [c] Yield of product isolated. [d] Conditions: (S,S)-5 (5 mol%), [{Pd(p-cinnamyl)Cl}2]
(2.5 mol%), cesium pivalate (1 equiv), and Cs2CO3 (1.5 equiv) in dry mesitylene as a solvent at 1608C for 3 h.
[e] Indole was a side product. [f] No product detected.
In the optimum case, the RRM reactions afford two enantio-
pure regioisomeric products in 50% yield each. Although fasci-
nating, their use in synthesis is limited because of the need for
product separation.
However, when starting from enantiopure precursors, this
limitation disappears and in selected examples single, highly
Scheme 4. Enantiomers reacting with identical catalysts to give regiodiver-
Scheme 3. Divergent CÀH activation: methine versus methyl CÀH bonds.
Chem. Eur. J. 2014, 20, 15021 – 15030 www.chemeurj.org
gent single products.
15023
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The proposed reaction mechanism is shown in
Scheme 5. The Pd0ÀNHC catalyst is formed by treat-
ment of [{Pd(p-cinnamyl)Cl}2] and NHCÀHI with
cesium pivalate as a base and nucleophile. Oxidative
addition of aryl bromide to the Pd0Àcarbene complex
fragment produces palladacycle A as a key intermedi-
ate. Bromide exchange for pivalate sets the stage for
the asymmetric regiodivergent step by C(sp3)ÀH acti-
vation through a pivalate-assisted CMD reaction.[12,16]
The two palladacycles produce the indolines by re-
ductive elimination; thus closing the catalytic cycle.
In a preliminary article, we showed that methyl CÀH
activation was favored over methylene CÀH activa-
tion with the achiral catalysts PdÀIMes or PdÀPCy3
(IMes=1,3-dimesitylimidazol-2-ylidene; both give rac-
3a exclusively). The energy difference of the two
transition states appears to be small enough to be
overcome when a suitable chiral ligand (e.g., 4 or 5)
is employed. This warranted further analysis and led
to the computational study detailed below.
Table 2. Switching the chirality of the catalyst, which leads to regioisomeric enantio-
merically pure products.
Entry R1/R2
Starting NHCÀHI Product 2
Product 3
(yield[b], ee [%])
material
(yield[b], ee [%])
1[a]
2[a]
3[a]
4[a]
5[a]
6[a]
7[c]
8[c]
Me/OMe
1u
1u
1w
1w
1x
1x
(R,R)-5 (2S,3R)-2u (86, >99) not formed
(S,S)-5 not formed (2S,3S)-3u (85, >99)
(R,R)-5 (2S,3R)-2w (97, >99) not formed
Me/OMe
Et/OMe
Et/OMe
Ph/OMe
Ph/OMe
(S,S)-5 not formed
(2S,3S)-3w (96, >99)
(R,R)-5 (2S,3R)-2x (98, >99) not formed
(S,S)-5 not formed
(R,R)-4 (2S)-2y (88, >99)
(S,S)-4 not formed
(2S,3S)-3x (98, >99)
not formed
(2R,3S)-3y (77, >99)
3-N-Me indolyl/H 1y
3-N-Me indolyl/H 1y
[a] Conditions: substrate (0.2 mmol), NHCÀHI (5 mol%), [{Pd(p-cinnamyl)Cl}2]
(2.5 mol%), cesium pivalate (1 equiv), and Cs2CO3 (1.5 equiv) were used in dry mesity-
lene (2 mL) at 1608C for 3 h. [b] Yield of product isolated. [c] Conditions: (R)-1y
(0.2 mmol), NHCÀHI (10 mol%), [{Pd(p-cinnamyl)Cl}2] (5 mol%), cesium pivalate
(1 equiv), and Cs2CO3 (1.5 equiv) were used in dry xylenesxyleness (2 mL) at 1408C for
24 h.
Computational study
Our recent combined theoretical and experimental
study showed that the PdÀNHC-catalyzed C(sp3)ÀH
arylation proceeded through a CMD mechanism.[4,12]
The CMD step for the activation of N-isopropyl carba-
mate was shown to be selectivity determining. In
view of the results detailed in the first part of this
the indolyl fragment. As before, only 2,3-trans-disubstituted in-
dolines were formed.
article, we proceeded with a detailed DFT study of methyl
versus methylene C(sp3)ÀH activation to rationalize selectivi-
ties observed in the regiodivergent reaction of racemic mix-
Recrystallization of enantiopure indoline (2S,3S)-3x from a so-
lution in hexane and ethyl acetate afforded crystals suitable for
X-ray analysis. The structure is shown in Figure 1 and clearly
demonstrates that the trans configuration of the 2,3-disubsti-
tuted indoline is formed. This is in perfect agreement with the
results previously obtained from vibrational circular dichroism
(VCD) studies.[13]
tures.[4c–e,18] The computed activation barriers (DG°413
, gas
phase) for the CMD step with a truncated achiral NHC ligand
and calculated for the activation of 1a are as follows:
CH3 (118.4 kJmolÀ1)<CH2trans (129.4 kJmolÀ1)<CH2cis (149.3 kJ
molÀ1).[19] This is in agreement with the finding that PdÀIMes
or PdÀPCy3 exclusively give the indoline that results from CH3/
Ar coupling.[12] We note, however, that DDG4°13 CH3/CH2trans in
1a computes to merely 11.0 kJmolÀ1. A chiral ligand may over-
come this difference. For the methyl versus methine CÀH acti-
vation in 1v, the numbers are CHcyclopropyl (122.7 kJmolÀ1)<CH3
(129.0 kJmolÀ1)<CH2trans
(138.9 kJmolÀ1)<CH2cis
(150.8 kJ
molÀ1). Here, DDG°413 CHcyclopropyl/CH3 is only 6.3 kJmolÀ1. Again,
this is a small difference and a suitable ligand may change the
situation. Details of the above analyses are given in the Sup-
porting Information. In the following section, we discuss CÀH
activation by the [Pd((S,S)-4)] catalyst.
Activation of substrate 1a
The experiments show that the coupling reaction of racemic
substrate 1a formed the two products 2a and 3a (Table 1,
entry 1). The CH3 activation product 3a is present in higher
yield (57%), but with lower ee of 77% compared with CH2 acti-
vation product 2a, which is formed in 38% yield with a very
high ee of >99%. Product (R)-3a represents the matched case
Figure 1. X-ray structure of trans-indoline (2S,3S)-3x.[17] Ellipsoids are given
at the 50% probability level.
Chem. Eur. J. 2014, 20, 15021 – 15030
www.chemeurj.org
15024
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
trans-CH2
activation
are
15.9 kJmolÀ1 for (R)-CH2 and
8.8 kJmolÀ1 for (S)-CH2. In agree-
ment with experiments, CH3 acti-
vation is preferable for (S)-1a
than that for (R)-1a. The respec-
tive difference between the acti-
vation barriers for trans-CH2 acti-
vation is 16.9 kJmolÀ1 in favor of
the (R)-CH2 transition state
(Figure 2). This is again in agree-
ment with experimentally ob-
served higher ee values of prod-
uct 2a.
The structures of the transi-
tion states trans-(R)-CH2 and
trans-(S)-CH2 are also shown in
Figure 2. As observed for CÀH
activation of parent N-isopropyl
carbamate,[12] there is steric re-
pulsion between the substrate
benzene ring and an o-methyl
group of the NHC ligand in TS
Scheme 5. Proposed catalytic cycle for the title-reaction.
trans-(S)-CH2 (the distance be-
tween the center of the benzene
ring and the o-methyl hydrogen
is 2.321 ꢃ), which increases the energy of the transi-
tion state. Moreover, the interaction of the second o-
methyl group of the NHC ligand with the methyl
group connected directly to the reacting carbon
atom of the substrate further increases the energy of
trans-(S)-CH2; thus leading to high enantioselectivity
for CH2 activation. Interestingly, transition state trans-
(S)-CH2 is a late transition state compared with trans-
(R)-CH2. This is shown in more detail and in color in
the Supporting Information, in which the graph from
Figure 2 is reproduced with the inclusion of solvent
effects (xylenes). Although the numbers are different,
the order of calculated DG°413 is the same.
Next, we attempted to correlate the activation bar-
riers, which were calculated relative to the most
stable transition state (S)-CH3, with the respective ex-
perimental values. A graphical representation of this
comparison is shown in Figure 3. The results show
that the free energy barriers, DDG°413, in the gas phase correlate
Scheme 6. Rationalization of regiodivergent RRM of substrate 1a.
of substrate (S)-1a and catalyst PdÀ(S,S)-4. It is formed in
higher yield, but lower enantioselectivity because (R)-1a/PdÀ very well with the experimental data (R2 =0.9819, root-mean-
(S,S)-4, which is the mismatched pair, produces not only
(2R,3S)-2a (catalyst control, excellent enantioselectivity), but
also some (S)-3a (Scheme 6). The differences in activation ener-
gies can be calculated from the relative amounts of products
formed (Scheme 6).
square deviation (RMSD)=1.8 kJmolÀ1). The largest deviation
from experiment (3.3 kJmolÀ1) is observed for transition state
trans-(S)-CH2, which leads to product (2S,3R)-2a. Due to the
very low yield of this product (detected as a minor peak in
HPLC), the error in the experimental relative free energy barrier
is quite large. The inclusion of solvent effects in xylenes as po-
larizable continuum model (PCM) single-point calculations with
united-atom Hartree–Fock (UAHF) radii does not improve the
correlation significantly (see the Supporting Information).
The activation barriers of the CMD step for different path-
ways of C(sp3)ÀH activation of carbamate 1a were calculated
(Figure 2). The results show that, with the chiral NHC, the pref-
erence for the formation of the trans product in CH2 activation
persists, as observed for the model achiral NHC (see the Sup-
porting Information for details). The DDG°413 values for cis- and
Chem. Eur. J. 2014, 20, 15021 – 15030
www.chemeurj.org
15025
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 2. Activation barriers (DG°413, gas phase) for the CMD step with (S,S)-NHC ligand 4 and substrate 1a; structures of the transition states for CH2 activation
of substrate 1a. Selected hydrogen atoms have been removed for clarity. The (R) and (S) descriptors refer to the configuration of substrate 1a. A color version
of this figure is available in the Supporting Information.
goes CH2 activation. In contrast to substrate 1a, transition
state (R)-CH3 for the activation of 1b is higher in energy rela-
tive to other transition states (Figure 4). Inspection of the cor-
responding structure shows that, in addition to steric repulsion
between the substrate benzene ring and the o-methyl group
of the NHC ligand (the distance between the center of the
benzene ring and the o-methyl carbon is 3.313 ꢃ), there is an
unfavorable interaction between substrate phenyl and carba-
mate groups. This increases the energy of transition state (R)-
CH3; the result of which is high levels of enantioselectivity in
both in CH3 and CH2 activation of substrate 1b.
Figure 3. Correlation between experimental and calculated (DG°413, gas
The activation barriers are, in most cases, overestimated by
phase) relative activation barriers for the regiodivergent RRM of substrate
1a.
calculations, particularly when solvent effects in xylenes are
added to the gas-phase free energies (Table 3). One possible
explanation for this is the systematic underestimation of dis-
persion interactions with the DFT functional used herein,[19]
Activation of substrate 1b
since the studied systems contain several aromatic rings. Dis-
Experimental data showed that racemic phenyl-substituted
substrate 1b gave two products, 2b and 3b, in 1:1 ratio in in-
tramolecular C(sp3)ÀH activation (Table 1, entry 2). Both prod-
ucts 2b and 3b are formed in very high enantiomeric purity
(98 and 95% ee, respectively). The activation barriers of the
CMD step for the different pathways of C(sp3)ÀH activation of
persion-corrected functionals, such as B97-D, were recently
shown to predict this type of interaction more accurately.[20]
Therefore, we briefly studied whether using the B97-D/
LANL2DZ,6-31G(d,p) level of theory for single-point calcula-
tions would improve the correlation.
Indeed, the relative activation free energies are consistently
lower when calculated at the combined B97-D/M06-L level of
theory, and RMSD=4.4 kJmolÀ1 is also lower in this case (see
the Supporting Information for details). This implies that dis-
persion interactions play some role in the relative stability of
transition states. However, the overall correlation was not
better (R2 =0.7707) and full optimization at the B97-D level of
theory was consequently carried out (Table 3). The obtained
DDG° values at the B97-D level of theory are close to the rela-
substrate 1b with (S,S)-NHC ligand
4 were calculated
(Figure 4). Similarly to the activation of substrate 1a, trans-
product 2b in CH2 activation is formed exclusively: the corre-
sponding differences between the activation barriers for cis-
and trans-CH2 activation are 27.9 kJmolÀ1 for (R)-CH2 and
17.7 kJmolÀ1 for (S)-CH2 (see the Supporting Information).
In agreement with experimental results, compound (S)-1b
preferably undergoes CH3 activation, whereas (R)-1b under-
Chem. Eur. J. 2014, 20, 15021 – 15030
www.chemeurj.org
15026
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The results show that for
para-methyl-substituted
sub-
strate 1c (Ar=4-CH3C6H4)) the
difference in DDG°413(CH2) be-
tween trans-(S)-CH2 and trans-
(R)-CH2 activation barriers is
higher and the difference in
DDG°413(CH3) is lower than that
for the substrate 1b (Ar=Ph).
This implies that product 2c
should be obtained in higher
enantioselectivity and product
3c should be obtained with
a lower ee value than those of
2b and 3b, respectively
These predictions are in good
agreement with experimental re-
sults (see Table 1, entries 2 and
3). However, the differences be-
tween the diastereomeric activa-
tion barriers are overestimated
by calculations, as mentioned for
substrate 1b (see above). For
Figure 4. Activation barriers (DG°413, gas phase) for the CMD step with (S,S)-NHC ligand 4 and substrate 1b; struc-
ture of the transition state for (R)-CH3 activation. Selected hydrogen atoms have been removed for clarity. The (R)
and (S) descriptors refer to the configuration of substrate 1b. A color version of this figure is available in the Sup-
porting Information.
Table 3. Comparison between experimental and calculated relative acti-
vation barriers for the regiodivergent RRM of substrate 1b.
Table 4. Calculated relative free energy activation barriers for the regiodi-
vergent RRM of racemic substrates 1b, 1c, and 1e with (S,S)-NHC 4.
Activation mode Experimental DDG4°13
Calculated DDG°413 [kJmolÀ1
DDG°[a] DDG°[a] DDG°[b]
gas phase xylenes gas phase
]
(S)-CH3
(R)-CH3
trans-(R)-CH2
trans-(S)-CH2
0.0
9.8
À0.1
11.8
0
0
0
21.6
2.4
14.0
19.6
12.0
24.5
20.5
2.3
21.9
Activation mode
Calculated DDG°413 (gas phase)[a] [kJmolÀ1
1b 1c 1e
]
correlation coefficient R2[c]
0.8057
6.1
0.7766 0.8874
10.1 6.9
[c]
RMSD [kJmolÀ1
]
(S)-CH3
0.0 0.0 0.0
[a] Calculated at the M06-L/[LANL2DZ,6-31G(d,p)] level of theory. [b] Cal-
culated at the B97-D/[LANL2DZ,6-31G(d,p)] level of theory. [c] Correlation
coefficients, R2, and RMSD between calculated and experimental relative
activation barriers.
(R)-CH3
trans-(R)-CH2
trans-(S)-CH2
21.6
2.4
14.0
9.6
À8.6
17.8
18.1
À3.8
9.7
DDG°413(CH2)[a]
DDG°413(CH3)[a]
11.6
21.6
26.4
9.6
13.5
18.1
[a] DDG°413(CH3)=DG°((R)-CH3)ÀDG°((S)-CH3);
CH2)ÀDG°((S)-CH2).
DDG°413(CH2)=DG°((R)-
tive activation barriers at M06-L level, except for transition
state trans-(S)-CH2. Although the overall correlation is slightly
better (R2 =0.8874), the activation barriers are, in most cases,
overestimated by calculations (RMSD=6.9 kJmolÀ1).
para-trifluoromethyl-substituted substrate 1e (Ar=4-CF3C6H4),
the difference between trans-(S)- and trans-(R)-CH2 activation
barriers is close to the corresponding DDG°413(CH2) value for un-
substituted 1b (Ar=Ph), so that products 3b and 3e are
formed with the same ee values (Table 1, entries 2 and 5). The
CH3 activation of substrate 1e is the only example of when
computational results do not agree with those obtained exper-
imentally: the calculated DDG°413(CH3) value is lower, but the ee
value of 2e is higher than that for unsubstituted product 2b.
One possible reason for this disagreement could be the differ-
ent NHC ligands used in experiments (NHC 5) and in calcula-
tions (NHC 4).
Variation of the para substituent in substrate 1b
To study the influence of electronic effects on enantioselectivi-
ties and product ratios in regiodivergent RRM, the activation
barriers for two para-substituted derivatives of 1b were calcu-
lated and compared with experimental results. The activation
free energies calculated at the M06-L/[LANL2DZ,6-31G(d,p)]
level in the gas phase are shown in Table 4 (for free energies in
xylenes, see the Supporting Information).
Chem. Eur. J. 2014, 20, 15021 – 15030
www.chemeurj.org
15027
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
For full experimental data of all compounds, see the Supporting In-
formation.
Conclusion
We probed the scope and limitations of PdÀNHC-catalyzed
asymmetric C(sp3)ÀH activation that led to enantioenriched 2-
substituted and trans-2,3-disubstituted indolines. Readily syn-
thesized substrates that incorporated a stereogenic center
were investigated and found to undergo highly regiodivergent
reactions of racemic mixtures. A number of cases with enantio-
pure starting materials demonstrated the synthetic utility of
this approach to produce single regioisomeric, enantiopure
trans-2,3-disubstituted indolines. This is all the more remark-
able because high temperatures are required to engage
C(sp3)ÀH bonds in coupling reactions. Limitations were that
CÀH bonds at, or adjacent to, tertiary centers were not
touched. One exception was the methine CÀH bond of a cyclo-
propyl moiety. A detailed DFT analysis of these reactions pro-
vided a good match to experimental data. The computational
data confirmed earlier conclusions that we were in the pres-
ence of a CMD pathway and it also provided insight into the
energetics of C(sp3)ÀH activation of methylÀ, methyleneÀ, and
methineÀCÀH bonds.
Synthesis of (rac)-1v
2-Bromoaniline (3.4 g, 20 mmol, 1 equiv), 4 ꢃ molecular sieves
(6 g), and 1-cyclopropylethan-1-one (3.4 g, 40 mmol, 2 equiv) were
dissolved in benzene (50 mL). The reaction mixture was heated at
reflux with a Dean-Stark apparatus for 3 days and filtered through
Celite. The filtrate was evaporated by rotary evaporator and dried
under vacuum. The crude imine was dissolved in dry methanol
(50 mL) and NaBH4 (2.28 g, 60 mmol, 3 equiv) was added slowly
under nitrogen. The reaction mixture was stirred for 2 h followed
by addition of 1n KOH (aq) and extraction with dichloromethane.
The organic phase was dried over Na2SO4. After filtration and evap-
oration, the residue was purified by column chromatography (silica
gel; diethyl ether/pentane as the eluent) to afford 2-bromo-N-(1-
cyclopropylethyl)aniline as a yellow oil (0.72 g, 15%). 1H NMR
(400 MHz, CDCl3): d=7.46 (dd, J=7.9, 1.5 Hz, 1H), 7.19 (ddd, J=
8.2, 7.3, 1.5 Hz, 1H), 6.69–6.65 (m, 1H), 6.57 (ddd, J=7.9, 7.3,
1.5 Hz, 1H), 4.36 (brs, 1H), 3.10 (m, 1H), 1.31 (d, J=6.3 Hz, 3H),
1.10–0.99 (m, 1H), 0.62–0.54 (m, 2H), 0.41–0.34 (m, 1H), 0.34–
0.27 ppm (m, 1H).
2-Bromo-N-(1-cyclopropylethyl)aniline (0.72 g, 3.0 mmol, 1 equiv)
was dissolved in methyl chloroformate (3.5 mL, 45 mmol, 15 equiv).
The reaction mixture was heated at reflux for 10 h before the reac-
tion mixture was poured into ice/water and extracted with di-
chloromethane. The organic phase was dried over MgSO4 and
evaporated by means of a rotary evaporator after filtration. The
crude product was purified by flash column chromatography (silica
gel; ethyl acetate/pentane=1:30 as the eluent) to afford (rac)-1v
Experimental Section
General
1
Solvents were purified by filtration on drying columns by using
a Solvtek system. Mesitylene was distilled over CaH2 under nitro-
gen. Reactions and manipulations that involved organometallic or
moisture-sensitive compounds were carried out under purified ni-
trogen and glassware was further dried by heating under high
vacuum as necessary. Flash chromatography (FC) was performed
on silica gel 60 (40 mm). Analysis with chiral HPLC was performed
by using an Agilent 1100 series chromatograph with a JASCO PU-
980 pump and an Agilent 1100 Series detection system.1H and
13C NMR spectra were recorded on a Bruker AMX-400 spectrome-
ter; d in ppm, pattern abbreviations: broad (b), quartet (q), quintet
(quint). The Fourier transform (FT) spectrometers used an internal
deuterium lock. Chemical shifts are quoted in ppm downfield of
tetramethylsilane. IR spectra were recorded on a Perkin–Elmer
Spectrum One spectrophotometer. Electron impact (EI) mass spec-
tra were obtained by using Varian CH-4 or SM-1 instruments oper-
ating at 40–70 eV and electrospray ionization (ESI) HRMS analyses
were measured on a VG analytical 7070E instrument. Optical rota-
tions were measured at 208C on a Perkin–Elmer 241 polarimeter
by using a quartz cell (l=10 cm) with a Na high-pressure lamp (l=
589 nm). Melting points were determined on a Bꢁchi 510 appara-
tus and were uncorrected.
as a colorless oil (0.54 g, 60%). H NMR (400 MHz, CDCl3): d=7.64
(d, J=8.0 Hz, 1H), 7.45 (d, J=6.9 Hz, 0.6H), 7.36–7.28 (m, 1H),
7.28–7.24 (m, 0.5H), 7.22–7.12 (m, 1H), 3.80 (m, 0.9H), 3.63 (s,
2.9H), 3.51 (d, 0.4H), 1.44 (d, J=6.8 Hz, 1.2H), 1.17 (d, J=6.9 Hz,
1.9H), 0.91–0.80 (m, 0.7H), 0.68–0.46 (m, 2.8H), 0.44–0.35 (m,
0.8H), 0.35–0.27 (m, 0.7H), 0.17–0.08 ppm (m, 0.4H); 13C NMR
(101 MHz, CDCl3): d=156.0, 155.5, 138.3, 133.65, 133.60, 131.2,
130.6, 129.0, 128.9, 127.91, 127.87, 126.3, 126.1, 60.9, 59.7, 53.1,
53.0, 20.5, 18.0, 17.8, 15.1, 6.2, 5.6, 4.5, 4.1 ppm; IR (neat): n˜ =2953
(w), 1704 (vs), 1585 (w), 1475 (m), 1440 (s), 1375 (m), 1315 (vs),
1262 (m), 1193 (m), 1112 (m), 1070 (m), 1044 (m), 1025 (m), 765 (s),
730 cmÀ1 (s); HRMS: m/z calcd for C13H16BrN1ONa [M+Na]+:
320.0256; found: 320.0260.
Catalytic asymmetric synthesis of (2S)-2v and (2R)-3v
Compound (rac)-1v (0.2 mmol, 59.4 mg), [{Pd(p-cinnamyl)Cl}2]
(2.6 mg, 0.005 mmol), cesium carbonate (97.5 mg, 0.3 mmol),
cesium pivalate (46.8 mg, 0.2 mmol), and (R,R)-5 (0.01 mmol,
5.88 mg) were successfully filled into a Schlenk flask. After the flask
was evacuated and backfilled with nitrogen, dry xylenes (2 mL)
was added and stirred at 1408C for 24 h. The reaction mixture was
cooled to room temperature and diluted with dichloromethane
(2 mL) followed by filtration through Celite. The solvent was re-
moved in vacuum and the residue was purified by flash column
chromatography (silica gel; ethyl acetate/pentane system as the
eluent) to afford a mixture of indolines (2S)-2v and (2R)-3v
(38.2 mg, 88%). 2v+3v: 1H NMR (400 MHz, CDCl3): d=8.00–7.36
(m, 1.4H), 7.24–7.13 (m, 2.3H), 7.03–6.92 (m, 1.6H), 6.66 (dd, J=
7.5, 1.3 Hz, 1H), 4.18 (brs, 1H), 4.03 (t, J=8.6 Hz, 0.6H), 3.86 (two s,
3H+1.8H), 3.33 (ddt, J=15.9, 9.4, 1.2 Hz, 0.6H), 2.89–2.79 (m,
0.6H), 1.27 (d, J=6.3 Hz, 3H), 1.20–1.06 (m, 2.6H), 1.03–0.96 (m,
1.2H), 0.88–0.80 (m, 1.2H), 0.73–0.56 (m, 0.6H), 0.58–0.40 (m,
Dry xylenes, benzene, cesium carbonate, cesium pivalate, pivalic
acid, methyl chloroformate, benzylmethylketones, and 2-bromoani-
line were purchased from Sigma-Aldrich, Fluka, or Acros and used
without further purification. 4-Trifluroomethylbenzylmethylketone,
N-alkyl-2-bromoaniline derivatives,[21] and methyl N-alkyl-2-bromo-
phenyl carbamates 1[7] were prepared by following the general
procedure or previously reported procedures. Carbamates 1 were
obtained as mixtures of rotamers. [{Pd(p-cinnamyl)Cl}2] was pre-
pared according to a procedure reported in the literature.[22] The
synthesis of the ligand precursors 4 and 5 was described previous-
ly.[23]
Chem. Eur. J. 2014, 20, 15021 – 15030
www.chemeurj.org
15028
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
1.2H), 0.28–0.18 ppm (m, 0.6H); 2v+3v: 13C NMR (101 MHz,
CDCl3): d=154.4, 130.8, 127.5, 127.1, 125.01, 123.2, 123.0, 118.8,
115.8, 115.2, 77.0, 63.5, 62.2, 52.7, 52.6, 34.4, 29.9, 28.7, 20.7, 18.7,
16.6, 9.8, 4.5, 1.5, 1.3 ppm; 2v+3v: IR (neat): n˜ =2954 (w), 1702
(vs), 1604 (w), 1485 (m), 1440 (s), 1386 (s), 1314 (m), 1273 (m), 1192
(m), 1140 (m), 1120 (m), 1044 (m), 1055 (m), 1022 (m), 749 cmÀ1 (s);
2v+3v HRMS: m/z calcd for C13H16N1O2 [M+H]+: 218.1175; found:
218.1165.
The reaction mixture was cooled to RT and diluted with dichloro-
methane (2 mL) followed by filtration through a pad of Celite. The
filtrate was evaporated under high vacuum. The residue was puri-
fied by flash column chromatography (silica gel; ethyl acetate/pen-
tane=1:30 as the eluent) to afford (2S,3R)-2x as a colorless oil
(58.2 mg, 98%, >99% ee). [a]2D0 =À24.18 (c=1.0 in CH2Cl2);
1H NMR (400 MHz, CDCl3): d=2.47–2.58 (m, 1H), 3.08 (brs, 3H),
3.15–3.47 (m, 1H), 3.87 (brs, 3H), 4.43 (s, 1H), 4.50–4.61 (m, 1H),
7.12 (t, J=7.2 Hz, 1H), 7.20–7.51 (m, 2H), 7.51–8.14 ppm (m, 1H);
13C NMR (100 MHz, CDCl3): d=38.0, 52.7, 55.2, 59.2, 66.2, 81.8,
115.9, 122.9, 126.6, 126.8, 127.4, 128.1, 128.6, 129.5, 130.5,
137.0 ppm; IR (neat): n˜ =2952, 1704, 1603, 1481, 1440, 1389, 1346,
1316, 1272, 1190, 1136, 1087, 1055, 1034, 968, 755 cmÀ1; HRMS (EI):
m/z calcd for C18H19NO3 [M]+: 297.1360; found: 297.1359.
Synthesis of (S)-1x
[Pd2(dba)3] (2 mol%, 0.127 mmol, 116 mg; dba=dibenzylideneace-
tone), rac-2,2’-bis(diphenylphosphino)-1,1’-binaphthyl (rac)-BINAP;
6 mol%, 0.381 mmol, 237 mg), and tBuONa (0.85 g, 8.89 mmol)
were successfully filled into a Schlenk flask. After the flask was
evacuated and backfilled with nitrogen, dry toluene (10 mL), (S)-1-
methoxy-3-phenylpropan-2-amine[24] (1.15 g, 6.98 mmol), and 1,2-
dibromobenzene (1.5 g, 6.35 mmol) were added under nitrogen.
The resulting reaction mixture was stirred at 1108C in the Schlenk
tube for 24 h. The reaction mixture was cooled to RT and diluted
with ethyl acetate (20 mL) followed by filtration through a pad of
Celite. The solvent was removed in vacuum and the residue was
purified by flash column chromatography (silica gel; diethyl ether/
pentane as the eluent) to afford (S)-2-bromo-N-(1-methoxy-3-phe-
Catalytic asymmetric synthesis of (2S,3S)-3x
Carbamate (S)-1x (75.6 mg, 0.2 mmol), cesium carbonate (97.5 mg,
0.3 mmol), [{Pd(p-cinnamyl)Cl}2] (2.6 mg, 0.005 mmol), cesium piva-
late (46.8 mg, 0.2 mmol), and (S,S)-5 (5.9 mg, 0.01 mmol) were
placed in a Schlenk flask. After the flask was evacuated and backfil-
led with nitrogen, dry mesitylene (2 mL) was added under nitro-
gen. The resulting reaction mixture was stirred at 1608C for 3 h.
The reaction mixture was cooled to RT and diluted with dichloro-
methane (2 mL) followed by filtration through a pad of Celite. The
filtrate was evaporated under high vacuum. The residue was puri-
fied by flash column chromatography (silica gel; ethyl acetate/pen-
tane=1:30 as the eluent) to afford (2S,3S)-3x as a colorless solid
(58.2 mg, 98%, >99% ee). M.p.=117–1198C; [a]2D0 = +69.59 (c=
1.0 in CH2Cl2); 1H NMR (400 MHz, CDCl3): d=3.46 (s, 3H), 3.53 (t,
J=9.0 Hz, 1H), 3.76 (dd, J=9.4, 3.8 Hz, 1H), 3.89 (s, 3H), 4.49–4.52
(m, 2H), 7.01–7.18 (m, 4H), 7.24–7.34 (m, 4H), 7.46–8.13 ppm (brd,
1H); 13C NMR (100 MHz, CDCl3): d=49.3, 52.7, 59.2, 67.2, 72.7,
115.4, 123.4, 126.0, 126.8, 127.4, 128.1, 128.7, 144.1 ppm; IR (neat):
n˜ =2929, 2818, 1703, 1599, 1481, 1440, 1377, 1302, 1261, 1189,
1
nylpropan-2-yl)aniline (1.36 g, 67% yield) as a colorless oil. H NMR
(400 MHz, CDCl3): d=3.02 (d, J=6.5 Hz, 1H), 3.40–3.50 (m, 3H),
3.81–3.87 (m, 1H), 4.67 (d, J=8.1 Hz, 1H), 6.62 (td, J=7.7, 1.4 Hz,
1H), 6.78 (dd, J=8.2, 1.1 Hz, 1H), 7.24 (ddd, J=8.4, 7.3, 1.4 Hz, 1H),
7.27–7.35 (m, 3H), 7.35–7.44 (m, 2H), 7.49 ppm (dd, J=7.9, 1.5 Hz,
1H); 13C NMR (100 MHz, CDCl3): d=37.2, 54.0, 59.1, 72.5, 110.3,
111.7, 117.7, 126.5, 128.5, 129.5, 132.7, 138.1, 144.0 ppm; IR (neat):
n˜ =3402, 3026, 2923, 1739, 1593, 1505, 1454, 1431, 1366, 1319,
1231, 1202, 1119, 1069, 1016, 968, 738, 698, 638 cmÀ1; HRMS (ESI):
m/z calcd for C16H19BrNO [M+H]+: 320.0644; found: 320.0641.
(S)-2-Bromo-N-(1-methoxy-3-phenylpropan-2-yl)aniline
(1.36 g,
1136, 1122, 1047, 964, 760 cmÀ1
; HRMS (ESI): m/z calcd for
C18H20NO3 [M+H]+: 298.1437; found: 298.1438.
4.25 mmol, 1 equiv) was dissolved in methyl chloroformate (3.5 mL,
45 mmol, 15 equiv). The reaction mixture was heated at reflux for
6 h before the reaction mixture was poured into ice/water and ex-
tracted with dichloromethane. The organic phase was dried over
MgSO4 and evaporated by means of a rotary evaporator after filtra-
tion. The crude product was purified by flash column chromatogra-
phy (silica gel; ethyl acetate/pentane=1:30 as eluent) to afford (S)-
1x as a colorless oil (1.54 g, 96%). [a]2D5 =À4.77 (c=1.0 in CH2Cl2);
1H NMR (400 MHz, CDCl3): d=2.62 (dd, J=13.3, 9.5 Hz, 1H), 3.19 (s,
1.8H), 3.25 (dd, J=13.2, 6.5 Hz, 1H), 3.37 (s, 4H), 3.45–3.68 (m, 1H),
3.65 (dd, J=10.0, 4.6 Hz, 1H), 3.71 (s, 1H), 3.72 (s, 3H), 3.77–3.94
(m, 1H), 4.22–4.29 (m, 0.5H), 4.45–4.52 (m, 1H), 6.97 (d, J=7.7 Hz,
1H), 7.13–7.43 (m, 11H), 7.64 (dd, J=8.0 Hz, 1H), 7.66 ppm (d, J=
11.3, 0.5H); 13C NMR (100 MHz, CDCl3): d=14.2, 21.1, 35.5, 37.3,
52.9, 53.0, 58.5, 58.6, 60.4, 61.4, 62.5, 70.6, 72.0, 124.8, 125.5, 126.4,
126.5, 128.0, 128.0, 128.5, 128.5, 128.6, 128.7, 129.2, 129.3, 130.9,
133.1, 138.7, 139.2, 155.1, 155.6 ppm; IR (neat): n˜ =3027, 2950,
1706, 1584, 1474, 1440, 1297, 1190, 1116, 1067, 1028, 951, 761, 747,
700 cmÀ1; HRMS (EI): m/z calcd for C18H20BrNO3 [M]+: 337.0621;
found: 337.0615.
Computational methods
Density functional (M06-L/LANL2DZ,6-31G(d,p)) calculations were
used to study the CMD step of the C(sp3)ÀH arylation. Single-point
PCM calculations in xylenes of all gas-phase optimized structures
yielded the solvent-corrected energy, Esolv.
[25] The united atom topo-
logical model with UAHF radii was used. For the PCM calculations,
the same level of theory as that used in the gas-phase calculations
was used. Gas-phase Gibbs free energy corrections, Gcorr (P=1 atm,
T=298 K), were considered for each species and the total Gibbs
free energy, G, of each optimized structure was taken as G=Esolv
+
Gcorr. The calculations were performed with the Gaussian 09 pack-
age.[19]
Acknowledgements
We thank the Swiss National Science Foundation and the Uni-
versity of Geneva for financial support. E.L. expresses his
thanks to Prof. H. Zipse for providing access to the Linux clus-
ter at the Faculty of Chemistry and Pharmacy of the LMU
Munich.
Catalytic asymmetric synthesis of (2S,3R)-2x
Carbamate (S)-1x (75.6 mg, 0.2 mmol), cesium carbonate (97.5 mg,
0.3 mmol), [{Pd(p-cinnamyl)Cl}2] (2.6 mg, 0.005 mmol), cesium piva-
late (46.8 mg, 0.2 mmol), and (R,R)-5 (5.9 mg, 0.01 mmol) were
placed in a Schlenk flask. After the flask was evacuated and backfil-
led with nitrogen, dry mesitylene (2 mL) was added under nitro-
gen. The resulting reaction mixture was stirred at 1608C for 3 h.
Keywords: asymmetric catalysis · carbenes · CÀH activation ·
chirality · density functional calculations
Chem. Eur. J. 2014, 20, 15021 – 15030
www.chemeurj.org
15029
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[1] For recent reviews on CÀH insertion, see: a) M. P. Doyle, R. Duffy, M. Rat-
nikov, L. Zhou, Chem. Rev. 2010, 110, 704–724; b) H. M. L. Davies, A. R.
Dick, Top. Curr. Chem. 2010, 292, 303–345.
2820; Angew. Chem. Int. Ed. 2012, 51, 2763–2766; c) C. Tsukano, M.
Okuno, Y. Takemoto, Chem. Lett. 2013, 42, 753–755; d) Q. Zhang, H.-Z.
Yu, Y. Fu, Organometallics 2013, 32, 4165–4173.
[2] A. R. Dick, M. S. Sanford, Tetrahedron 2006, 62, 2439–2463.
[3] For reviews on CÀH bond functionalization, see: a) L. Ackermann, Acc.
Chem. Res. 2014, 47, 281–295; b) Q.-Z. Zheng, N. Jiao, Tetrahedron Lett.
2014, 55, 1121–1126; c) P. B. Arockiam, C. Bruneau, P. H. Dixneuf, Chem.
Rev. 2012, 112, 5879–5918; d) B.-J. Li, Z.-J. Shi, Chem. Soc. Rev. 2012, 41,
5588–5598; e) S. R. Neufeldt, M. S. Sanford, Acc. Chem. Res. 2012, 45,
936–946; f) N. Kuhl, M. N. Hopkinson, J. Wencel-Delord, F. Glorius,
Angew. Chem. 2012, 124, 10382–10401; Angew. Chem. Int. Ed. 2012, 51,
10236–10254; g) L. Ackermann, Chem. Rev. 2011, 111, 1315–1345; h) X.
Chen, K. M. Engle, D.ÀH. Wang, J.-Q. Yu, Angew. Chem. 2009, 121, 5196–
5217; Angew. Chem. Int. Ed. 2009, 48, 5094–5115.
[4] For selected references, see: a) B. Biswas, M. Sugimoto, S. Sakaki, Orga-
nometallics 2000, 19, 3895–3908; b) S. Pascual, P. de Mendoza, A. M.
Echavarren, Org. Biomol. Chem. 2007, 5, 2727–2734; c) D. Balcells, E.
Clot, O. Eisenstein, Chem. Rev. 2010, 110, 749–823; d) S. Rousseaux, S. I.
Gorelsky, B. K. W. Chung, K. Fagnou, J. Am. Chem. Soc. 2010, 132,
10692–10705; e) R. Jazzar, J. Hitce, A. Renaudat, J. Sofack-Kreutzer, O.
Baudoin, Chem. Eur. J. 2010, 16, 2654–2672.
[17] CCDC-1008423 ((2S,3S)-3x) contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge from
The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
[18] a) M. Lafrance, K. Fagnou, J. Am. Chem. Soc. 2006, 128, 16496–16497;
b) M. Lafrance, S. I. Gorelsky, K. Fagnou, J. Am. Chem. Soc. 2007, 129,
14570–14571.
[19] Gaussian 09, Revision A.02, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
gar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, ꢄ. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski, D. J. Fox, Gaussian, Inc. Wallingford CT, 2009.
[5] C. Zheng, S.-L. You, RSC Adv. 2014, 4, 6173–6214.
[6] N. Dastbaravardeh, M. Christakakou, M. Haider, M. Schnꢁrch, Synthesis
2014, 1421–1439.
[7] T. Watanabe, S. Oishi, N. Fujii, H. Ohno, Org. Lett. 2008, 10, 1759–1762.
[8] M. Nakanishi, D. Katayev, C. Besnard, E. P. Kꢁndig, Angew. Chem. 2011,
123, 7576–7579; Angew. Chem. Int. Ed. 2011, 50, 7438–7441.
[9] S. Anas, A. Cordi, H. B. Kagan, Chem. Commun. 2011, 47, 11483.
[10] a) T. Saget, S. J. Lemouzy, N. Cramer, Angew. Chem. 2012, 124, 2281–
2285; Angew. Chem. Int. Ed. 2012, 51, 2238–2242; b) T. Saget, N.
Cramer, Angew. Chem. 2012, 124, 13014–13017; Angew. Chem. Int. Ed.
2012, 51, 12842–12845; c) T. Saget, D. Perez, N. Cramer, Org. Lett. 2013,
15, 1354–1357.
[11] N. Martin, C. Pierre, M. Davi, R. Jazzar, O. Baudoin, Chem. Eur. J. 2012,
18, 4480–4484.
[12] E. Larionov, M. Nakanishi, D. Katayev, C. Besnard, E. P. Kꢁndig, Chem. Sci.
2013, 4, 1995–2005.
[20] a) S. Grimme, J. Comput. Chem. 2004, 25, 1463–1473; b) S. Grimme, J.
Comput. Chem. 2006, 27, 1787–1799.
[21] a) A. F. Abdel-Magid, K. G. Carson, B. D. Harris, C. A. Maryanoff, R. D.
Sarah, J. Org. Chem. 1996, 61, 3849–3862; b) D. J. Cho, C. J. Wu, S. S,
W.-S. Han, S. O. Kang, B. Y. Lee, Organometallics 2006, 25, 2133–2134.
[22] N. Marion, O. Navarrro, J. Mei, E. D. Steven, N. M. Scott, S. P. Nolan, J.
Am. Chem. Soc. 2006, 128, 4101–4111.
[23] For the synthesis of (S,S)- and (R,R)-4, see: E. P. Kꢁndig, T. M. Seidel, Y. X.
Jia, G. Bernardinelli, Angew. Chem. 2007, 119, 8636–8639; Angew. Chem.
Int. Ed. 2007, 46, 8484–8487; (S,S)- and (R,R)-5: D. Katayev, Y.-X. Jia, A. K.
Sharma, D. Banerjee, C. Besnard, R. B. Sunoj, E. P. Kꢁndig, Chem. Eur. J.
2013, 19, 11916–11927; (S,S)-4 and (S,S)-5 are available from Sigma-Al-
drich Corporation.
[13] D. Katayev, M. Nakanishi. T. Bꢁrgi, E. P. Kꢁndig, Chem. Sci. 2012, 3,
1422–1425.
[14] For a review on divergent reactions on racemic mixtures, see: L. C.
Miller, R. Sarpong, Chem. Soc. Rev. 2011, 40, 4550–4562.
[24] (S)-1-Methoxy-3-phenylpropan-2-amine was synthesized according to
a procedure reported in the literature, see: B. Deng, Q. Wu, Y. Pan, B.
Hu, L. Zhang, Patent CN 102060732 (A), 2011.
[25] J. Tomasi, B. Mennucci, R. Cammi, Chem. Rev. 2005, 105, 2999–3094.
[15] a) J. F. Hartwig, Acc. Chem. Res. 2008, 41, 1534–1544; b) D. S. Surry, S. L.
Buchwald, Chem. Sci. 2011, 2, 27–50.
Received: June 16, 2014
Published online on September 26, 2014
[16] a) S. Rousseaux, B. Liegault, K. Fagnou, Chem. Sci. 2012, 3, 244–248; b)
C. Tsukano, M. Okuno, Y. Takemoto, Angew. Chem. 2012, 124, 2817–
Chem. Eur. J. 2014, 20, 15021 – 15030
www.chemeurj.org
15030
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim