Iridium-catalyzed allylic alkylation reaction with N-aryl phosphoramidite ligands: Scope and mechanistic studies

Article abstract of DOI:10.1021/ja210923k

A series of N-aryl phosphoramidite ligands has been synthesized and applied to iridium-catalyzed allylic alkylation reactions, offering high regio- and enantioselectivities for a wide variety of substrates. These ligands feature the synthetic convenience and good tolerance of the ortho-substituted cinnamyl carbonates. Mechanistic studies, including DFT calculations and X-ray crystallographic analyses of the (π-allyl)-Ir complexes, reveal that the active iridacycle is formed via C(sp²)-H bond activation.

Full text of DOI:10.1021/ja210923k

Article
pubs.acs.org/JACS
Iridium-Catalyzed Allylic Alkylation Reaction with N-Aryl
Phosphoramidite Ligands: Scope and Mechanistic Studies
Wen-Bo Liu, Chao Zheng, Chun-Xiang Zhuo, Li-Xin Dai, and Shu-Li You*
State Key Laboratory of Organometallic Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, 345
Lingling Lu, Shanghai 200032, People's Republic of China
*
S Supporting Information
ABSTRACT: A series of N-aryl phosphoramidite ligands has
been synthesized and applied to iridium-catalyzed allylic
alkylation reactions, offering high regio- and enantioselectiv-
ities for a wide variety of substrates. These ligands feature the
synthetic convenience and good tolerance of the ortho-
substituted cinnamyl carbonates. Mechanistic studies, includ-
ing DFT calculations and X-ray crystallographic analyses of the
(
π-allyl)−Ir complexes, reveal that the active iridacycle is
2
formed via C(sp )−H bond activation.
INTRODUCTION
substituted cinnamyl substrates in general could not be
tolerated, with much decreased enantioselectivities observed
■
During the past decade significant progress has been made in
the field of iridium-catalyzed allylic substitution reactions,
which has been developed into a reliable method to construct
chiral centers in organic synthesis. The reaction allows the
preparation of highly enantioenriched allylic compounds from
readily available monosubstituted allylic substrates (Scheme 1).
4
b,8,9
for many different types of nucleophiles.
explanation is the steric hindrance between ligands and
A plausible
7b
1
,2
substrates. Recently, we synthesized a series of phosphor-
amidite ligands from the readily available secondary amines and
enantiopure BINOL, which resulted in good regio- and
enantioselectivities for a broad range of substrates, including
9
ortho-substituted phenyl allylic substrates. In addition, these
Scheme 1. Ir-Catalyzed Allylic Substitution Reactions
ligands were also found to be effective in the iridium-catalyzed
asymmetric intramolecular allylic dearomatization of indoles
10
and pyrroles. DFT calculation and X-ray analyses of crystal
structures of the (π-allyl)−Ir intermediates showed that the
2
active catalytic species was formed via C(sp )−H bond
activation of the ligand. Inspired by these results, various
phosphoramidite ligands were synthesized to confirm the
importance of the N-aryl moiety, accounting for the formation
of a catalytic iridacycle. Moreover, these ligands were found to
be highly efficient for Ir-catalyzed allylic alkylation reactions.
Here, we report the details of this study.
Since the first report on iridium-catalyzed asymmetric allylic
3
alkylation reactions with oxazolinyl−phosphine ligands, many
4
catalytic systems involving chiral ligands such as binaphthol-
4
a−h
4i
based phosphoramidites,
diene ligands, phosphorami-
4j,k
4l−n
dite−olefin ligands, and others
afford excellent levels of regio- and enantioselectivity.
In particular, the Feringa-type ligand 1a, first employed in
Ir-catalyzed allylic amination by the Hartwig group, was used
most frequently, providing high regio- and enantioselectivity for
various nucleophiles. Then the Alexakis group developed the o-
methoxyl-substituted ligand 1b, which induced higher
selectivities in many cases. The mechanistic studies carried
out by Hartwig and co-workers disclosed that the active
catalytic species of the iridium complex was formed via an
have been developed to
RESULTS AND DISCUSSION
■
5
Design and Synthesis of N-Aryl Phosphoramidite
Ligands. The initial design of phosphoramidite ligands was
focused on binaphthol-based species derived from amines
bearing an α-methyl group, imitating the phenethyl amino
moiety in the Feringa ligand, as we have thought this is critical
4b
4
d
3
to form the active iridacycle catalyst by Ir-mediated C(sp )−H
1c
bond insertion. For this purpose, enantiopure 2-methylindo-
11
12
line and 2-methyltetrahydroquinoline were chosen as the
3
amine moieties. Phosphoramidite ligands 1c−f were synthe-
insertion of the Ir center into the C(sp )−H bond in the
methyl group of the amine part of the ligand.
13
1c,6a
sized in a one-pot procedure (Scheme 2). After further
The (π-
allyl)−Ir complex intermediates were also synthesized, isolated,
6
b,7
and confirmed recently.
Despite the broad substrate scope,
Received: November 21, 2011
Published: February 7, 2012
with the iridium catalyst derived from ligand 1a or 1b, ortho-
©
2012 American Chemical Society
4812
dx.doi.org/10.1021/ja210923k | J. Am. Chem. Soc. 2012, 134, 4812−4821

Journal of the American Chemical Society
Article
Scheme 2. Synthesis of the Phosphoramidite Ligands
ligands could be obtained after recrystallization in ethyl
15
acetate. Most of the ligands synthesized above are crystalline
16
solids and can be readily purified.
Optimization of Reaction Conditions. To test the
suitability of these N-aryl ligands in Ir-catalyzed allylic
substitution reactions, we began our study by choosing
cinnamyl carbonate and sodium dimethyl malonate as
substrates, along with the catalysts derived from [Ir(cod)Cl]2
and phosphoramidites. The results are summarized in Table 1.
Ligands 1c−f were screened following procedure A, with LiCl
4a
as the additive, developed by Helmchen and co-workers. The
catalysts derived from 1c,d were ineffective for this reaction,
with very low conversion (entries 1 and 2). The reaction with
ligand 1e proceeded in excellent yield, regioselectivity, and
enantioselectivity (3a/4a = 97/3, 97% ee for 3a, entry 3).
Ligand 1f, a diastereoisomer of 1e, only led to the formation of
the product in 20% yield and 78% ee with reversed absolute
configuration (entry 4). Comparison of the results obtained
from ligands 1e,f not only clearly shows the match of two kinds
of chiralities in 1e but also indicates that the enantiocontrol is
dominated by the axial chirality of the ligands (entries 3 and 4).
Due to the fact that the reaction is much faster with 1e than
4f
with 1f, the diastereomeric mixture of ligands 1e*, prepared
from (R )-BINOL and racemic 2-methyltetrahydroquinoline,
were also tested in the reaction. As expected, the reaction with
a
understanding the catalysts derived from these ligands, we
realized that the active catalyst might be formed via Ir-mediated
1
e* gave the alkylation product in 92% yield and 92% ee (entry
2
3
C(sp )−H bond insertion rather than C(sp )−H bond
insertion. Therefore, enantiopure 2-ethyltetrahydroquinoline
and several commercially available achiral N-aryl amines were
5
).
Different procedures of catalyst preparation and reaction
conditions were also investigated. The combination of dry
triazabicyclodecene (TBD), tetrahydrothiophene (THT), and
CuI (procedure B) was found previously to be very efficient for
allylic alkylation reactions, particularly in conjunction with
14
also used, giving phosphoramidite ligands 1g and 1h−1j
with only axial chirality). Ligands 1k,l were prepared by using
the racemic amines and (R )-BINOL, and the enantiopure
(
a
a
Table 1. Investigation of Ligands in Ir-Catalyzed Allylic Alkylation Reactions
b
c
d
e
entry
ligand
procedure
t (h)
3a/4a
yield (%)
ee (%)
f
f
1
2
3
4
5
6
1c
1d
1e
1f
A
A
A
A
A
B
24
24
5
92/8
ND
ND
90/10
97/3
83/17
97/3
97/3
99/1
97/3
97/3
97/3
86/14
92/8
94/6
73/27
76/24
16
91
20
92
94
88
95
96
95
95
73 (S)
97 (R)
78 (S)
92 (R)
96 (R)
96 (R)
98 (R)
95 (R)
95 (R)
95 (R)
92 (R)
85 (R)
13 (S)
10 (S)
24
12
<1
1
(rac-C,R )-1e*
a
1e
1a
1e
1e
1g
1h
1i
4
e
7
8
9
1
1
1
1
1
1
B
C
C
C
C
C
C
C
C
<1
12
2
g
0
1
2
3
4
5
12
12
12
48
48
h
42
h
1j
61
i
i
h
1k
1l
30
h
49
a
The reactions were conducted at room temperature on a 0.25 mmol scale in THF (2.5 mL) with a relative mole ratio 2a:NaCH(CO Me) :
2
2
b
[
Ir(cod)Cl] :1 of 1:2:0.02:0.04. Procedure A: [Ir(cod)Cl] and 1 were stirred in THF for 20 min at room temperature, and then LiCl (100 mol %)
2
2
and substrates were added. Procedure B: [Ir(cod)Cl] , 1, TBD (12 mol %), and THT (20 mol %) were stirred in THF for 2 h at room temperature,
2
and then CuI (20 mol %) and substrates were added. Procedure C: [Ir(cod)Cl] , 1, and n-PrNH (0.3 mL) were stirred in THF (0.5 mL) for 30 min
2
2
c
1
at 50 °C, and then the volatiles were removed under vacuum to yield the catalyst. See the Supporting Information for details. Determined by H
NMR of the crude reaction mixture. Isolated yield of 3a and 4a. Determined by HPLC analysis; the absolute configuration is indicated in
parentheses. Not determined. 0.5 mol % [Ir(cod)Cl] and 1 mol % 1e were used. Conversion of 2a. 10 mol % [Ir(cod)Cl] and 20 mol % 1 were
d
e
f
g
h
i
2
2
used.
4
813
dx.doi.org/10.1021/ja210923k | J. Am. Chem. Soc. 2012, 134, 4812−4821

Journal of the American Chemical Society
Article
4
d
a
ligand 1a. Further investigations by Hartwig and co-workers
Table 2. Scope of the Allylic Carbonates
8a
led to procedure C, a general method for catalyst preparation
using n-propylamine to induce the formation of cyclometalated
6
a
complex K1a via a C−H bond activation at the methyl group
of ligand 1a (Scheme 3). Procedures B and C could both
Scheme 3. Cyclometalated Ir Complexes
obviously increase the rate of the reaction (entries 6 and 8).
The reaction with ligand 1e proceeded in higher yield and ee
than those with Feringa ligand 1a (entries 7 and 8). Even with a
reduced catalyst loading (0.5 mol % of catalyst), the product
could be obtained in 96% yield together with 97/3
regioselectivity and 95% ee (entry 9).
Encouraged by these results, we next examined the rest of the
phosphoramidite ligands synthesized above. Ligand 1g, bearing
a 2-ethyltetrahydroquinoline moiety, could afford regioselectiv-
ity and enantioselectivity (95% ee, entry 10) comparable to
those of ligand 1e under the same reaction conditions (entry
8).
Next, phosphoramidites 1h−j, containing achiral aromatic
amines and (R )-BINOL as the only stereochemical element,
a
were surveyed. The reaction with 1h occurred smoothly in
excellent yield and ee, indicating that the methyl group β to
4f
nitrogen is not necessary for the cyclometalation (entry 11).
The more steric bulky ligands 1i,j also afforded the desired
product in good regioselectivity and enantioselectivity, with
moderate yields (entries 12 and 13). In contrast, ligands 1k,l,
containing chiral aliphatic amines, were screened in the
reaction, affording the product in low conversion and
enantioselectivity (entries 14 and 15). Notably, the absolute
configuration of branched products was reversed in both cases.
All the experimental data above suggest that the phenyl group
on the nitrogen in the ligands plays an important role in this
reaction. It is likely that the cyclometalation of iridium
complexes generated from ligands 1c−j might be formed via
a
The reactions were conducted under the conditions of entry 8, Table
b
1
c
1. Determined by H NMR of the crude reaction mixture. Isolated
yield of 3 and 4. The ee was determined by HPLC analysis, and the
absolute configuration (or the sign of optical rotation) is indicated in
parentheses.
d
Heteroaryl allyl carbonate 2j also reacted smoothly with
ligand 1e to produce 3j in 92% yield and 93% ee (entry 22).
Excellent regioselectivity was obtained in most cases (95/ 5→
2
3
C(sp )−H bond activation instead of C(sp )−H bond
9
9/1), except for aliphatic allyl carbonate 2k, which afforded
activation observed with the Feringa ligand.
Reaction Scope. With the optimized reaction conditions in
product 3k in 90% yield but with a branched to linear ratio of
5
0/50 (entry 23).
Next our efforts were focused on expanding the reaction
hand (2 mol % of [Ir(cod)Cl] , 4 mol % of 1e, and 200 mol %
2
of NaCH(CO Me) in THF at room temperature), the
2
2
scope with various ortho-substituted phenyl allyl carbonates,
which in general afford poor enantioselectivity with ligands
1
9
1
was employed, only moderate ee values of 36−79% were
obtained (entries 9, 10, and 12). 1-Naphthyl allyl carbonate 2f
gave branched product in 92% ee with 1e, but in only 15% ee
with 1a (entries 14 and 15). Halogen-containing substrates,
allowing versatile subsequent transformations, were also
investigated. Allylic carbonates 2g−i, bearing F, Cl, and Br at
the ortho position of the phenyl group, could give products
3g−i in 97−98% ee with 1e (entries 17, 19, and 21). However,
substrate scope was examined by varying the substituents on
the allylic carbonates (Table 2). The results from ligands 1a,b
are also included for comparison.
With ligand 1e, the reactions of allylic carbonates bearing an
electron-donating group (−OMe) at the para- or meta
positions of the aromatic ring both gave excellent yields,
regioselectivities, and ee (entries 6 and 8, Table 2), almost the
same as those obtained from ligands 1a (entry 4) and 1b
a,b. With o-OMe- and o-Me-substituted cinnamyl carbonates,
9% and 98% ee were obtained respectively by utilizing ligand
e (entries 11 and 13). In sharp contrast, when ligand 1a or 1b
(
entries 5 and 7). It should be noted that the stereochemistry of
the branched products remains the same, although the
binaphthol chirality of (R,R )-1e is different from that of
a
(
S,S,S )-1a and (S,S,S )-1b (entries 1−8).
a
a
4
814
dx.doi.org/10.1021/ja210923k | J. Am. Chem. Soc. 2012, 134, 4812−4821

Journal of the American Chemical Society
Article
Scheme 4. Possible C−H Bond Activation Pathways in the Preparation of the Active Catalytic Species with Model Ligand 1e′
a
Calculated at the M06-2X/SDD/6-31G(d,p) Level of Theory
a
The ΔG and ΔE values (in parentheses) are in kcal/mol.
these corresponding alkylation products were obtained in only
−80% ee in the presence of ligand 1a with much longer
reaction time (entries 16−21). In addition to the significant
improvement of the enantioselectivity, the yields of the
products were increased dramatically by 1e for many cases,
such as m-methoxyphenyl (98% vs 58%, entries 7 and 8)-, o-
methylphenyl (90% vs 26%, entries 12 and 13)-, 1-naphthyl
product in good ee (entries 11−13), but ligands 1k,l, with only
3
3
possible C(sp )−H activation sites, afforded the S product with
poor enantioselectivity (entries 14 and 15). These results
indicate that the cyclometalation of ligands 1a,e might occur
through different manners. To shed light on the mechanism,
computational studies and characterizations of iridium
complexes have been carried out.
(
84% vs 31%, entries 14 and 15)-, 2-chlorophenyl (85% vs 25%,
entries 18 and 19)-, and 2-bromophenyl (82% vs 21%, entries
0 and 21)-substituted allyl carbonates. This kind of broad
DFT Calculations. DFT calculations were performed to
2
investigate the chemoselectivity between the C(sp )−H bond
2
3
and the C(sp )−H bond activation in the formation of active
scope tolerating ortho-substituted phenyl allyl carbonates has
never been realized before.
18
catalytic species with the phosphoramidite ligands. In our
simplified model system, ligands (R,R )-1e and (R,R,R )-1a are
a
a
Formation of Cyclometalated Ir Complexes. The
pioneering work by the Hartwig group disclosed that the
active catalyst contains the five-membered cyclometalated
replaced by their biphenyl-derived analogues 1e′ and 1a′. The
19
calculations were conducted with the Gaussian09 package
20
6a
using the M06-2X functional. For Ir, the SDD basis sets with
the associated effective core potential were employed, while
the standard 6-31G(d,p) basis sets were used for all the other
iridium complex K1a, formed by C−H activation at the
21
methyl group of ligand 1a. A similar C−H activation of
P(OPh) was also observed earlier by Helmchen and co-
3
22
1
7
atoms. All the structures were optimized in THF (ε = 7.4257)
workers. For most of the N-aryl ligands, 1e as an example,
23
with the SMD model. Frequency analyses were conducted to
confirm each structure being a minimum (no imaginary
frequency) or a transition state (only one imaginary frequency).
The gas-phase energies of the intermediates and transition
states for the C−H bond activation as well as the allylic
complexes with 1e′ are presented in the Supporting
Information. The key structural parameters and the energetic
trends are similar to those obtained in THF solution. In this
article, we mainly discuss the Gibbs free energies in THF (ΔG)
unless specified. The electronic energies in THF (ΔE) are also
given in parentheses for reference.
there are two possible sites to form the cyclometalated species
2
3
K1e-sp and K1e-sp , which could be formed respectively via
2
3
the cleavage of the C(sp )−H bond at C8 and the C(sp )−H
bond at the methyl group of the tetrahydroquinoline moiety
(
Scheme 3).
From the data shown in Table 1, two trends could be
extracted. First, the absolute configuration of branched
products was principally controlled by the BINOL scaffold,
and those obtained with R axial chirality ligands 1e,g are the
a
same as that obtained with S ligand 1a. Second, ligands 1h−j,
a
2
containing only potential C(sp )−H activation sites, gave the R
4
815
dx.doi.org/10.1021/ja210923k | J. Am. Chem. Soc. 2012, 134, 4812−4821

Journal of the American Chemical Society
Article
Figure 1. Optimized structures and relative energies of intermediates and transition states for the C−H bond activation with model ligand 1e′,
calculated at the M06-2X/SDD/6-31G(d,p) level of theory. The ΔG and ΔE values (in parentheses) are in kcal/mol. The bond distances are in Å.
2
The chemistry of Ir is different from that of Rh, where the
chloride is usually still bound to the metal center in the active
These results indicate that the aromatic C(sp )−H bond
activation is both kinetically feasible and thermodynamically
24
7
catalytic species. Helmchen and co-workers showed that
adding an Ag(I) salt could facilitate the cyclometalation by
chloride abstraction. This method for generation of the active
catalytic species was used in the preparation of allyl−Ir
complexes with ligand 1e. The catalysts synthesized in this
way exhibit reactivity similar to that obtained in situ by
procedures A−C in the catalytic reactions described previously
favored for ligand (R,R )-1e.
a
Subsequently, the thermodynamic stabilities of the allylic
complexes with different types of metallacycles were compared.
2
3
In each case (generated via C(sp )−H bond or C(sp )−H
bond activation), the allylic moiety can adopt two different
configurations. We define in this article that the complex with
the C2−H bond of the allylic moiety pointing toward the
amine part of the ligand as endo, while the complex with the
C2−H bond of the allylic moiety pointing toward cod as exo
(Scheme 5). The calculation results showed that allyl-1e′-sp -
endo is the most stabilized isomer. The relative energy of allyl-
1e′-sp -exo is 2.9 kcal/mol. On the other hand, the allyl
(vide infra). Therefore, we set the cationic complex C-1
without a chloride bound to the Ir(I) center as the starting
25
2
point of the calculation. As depicted in Scheme 4 and Figure
2
6
1,
the Ir center in C-1 has a vacant coordination site; the
2
distance between Ir and the nearest proton on the phenyl ring
2
3
of the ligand is 2.589 Å. Either the aromatic C(sp )−H bond or
complexes with the metallacycle derived from the C(sp )−H
bond activation are generally very unstable compared to the
3
the methyl C(sp )−H bond of the 2-methyltetrahydroquinoline
2
moiety in C-1 can interact with the Ir center to form two types
C(sp )−H bond activation counterparts. The relative energies
2
3
2
3
3
of agostic intermediates (I-1e′-sp and I-1e′-sp ). I-1e′-sp is
of allyl-1e′-sp -endo and allyl-1e′-sp -exo are 8.2 and 11.8 kcal/
27
the most stabilized isomer and is set as the energy reference.
mol, respectively. These data suggest that the allylic complex
3
2
The calculated ΔG values of C-1 and I-1e′-sp are 1.4 and 2.8
with the C(sp )−H bond activation is the favored intermediate
2
kcal/mol relative to that of I-1e′-sp , respectively. The
with ligand (R,R )-1e.
a
3
transition states and the direct products of the following C−
The next question to be addressed is why methyl C(sp )−H
bond activation occurs when Feringa-type ligands are used,
since they also have a potential C(sp )−H bond activation site.
Similar calculations were carried out on the simplified Feringa-
type ligand 1a′. As shown in Scheme 6 and Figure 2, complex
C-2 is the direct product of chloride abstraction with an Ag(I)
salt. The Ir center in this complex also has a vacant
coordination site. The distance between Ir and the nearest
proton on the methyl group of the ligand is 2.522 Å. Different
H bond cleavage from these agostic intermediates were also
2
2
located. The energetic barrier of aromatic C(sp )−H bond
2
activation (TS-1e′-sp ) is 12.6 kcal/mol, leading to the Ir−H
2
26
species II-1e′-sp (0.3 kcal/mol). On the other hand, the
3
methyl C(sp )−H bond activation has a much higher energetic
3
25
barrier (TS-1e′-sp , 19.7 kcal/mol) and the Ir−H species
3
generated (II-1e′-sp , 4.0 kcal/mol) is also unstable in
2
comparison to its C(sp )−H bond activation counterpart.
4
816
dx.doi.org/10.1021/ja210923k | J. Am. Chem. Soc. 2012, 134, 4812−4821

Journal of the American Chemical Society
Article
complex III-1a′-sp2.29 This complex is set as the energy
reference, and the relative energies of C-2 and the agostic
Scheme 5. Possible (π-Allyl)−Ir Complexes with Model
Ligand 1e′ Calculated at the M06-2X/SDD/6-31G(d,p)
Level of Theory
a
3
complex (III-1a′-sp ) in which the Ir center interacts with the
3
methyl C(sp )−H bond of the ligand are 1.0 and 7.4 kcal/mol,
respectively. The transition states and the direct products of the
C−H bond cleavage processes were also located. The energetic
3
barrier of the C(sp )−H bond activation of the phenyl ring
3
(
TS-1a′-sp ) is 20.1 kcal/mol, affording the corresponding Ir−
3
H intermediate IV-1a′-sp (5.3 kcal/mol). However, the
2
C(sp )−H bond activation of ligand 1a′ is less feasible. The
2
ΔG value of TS-1a′-sp (23.9 kcal/mol) is higher than that of
3
TS-1a′-sp by 3.8 kcal/mol. In addition, the Ir−H intermediate
2
2
generated through C(sp )−H bond activation (IV-1a′-sp , 11.0
kcal/mol) is also thermodynamically unstable in comparison to
3
3
IV-1a′-sp . These results indicate that the C(sp )−H bond
activation is the favored process with a Feringa-type ligand,
which is consistent with the experimental observations.
6a
The structures of the (π-allyl)−Ir complexes with different
types of metallacycles of ligand 1a′ were also optimized. The
relative energy of the (π-allyl)−Ir complex with the metalla-
a
28
The ΔG and ΔE values (in parentheses) are in kcal/mol.
2
2
cycle derived from C(sp )−H bond activation (allyl-1a′-sp -
3
endo) is 12.5 kcal/mol higher than the C(sp )−H bond
from the case for ligand 1e′, the Ir center in this system is
favored to coordinate with the phenyl ring of the chiral amine
part rather than either of the C−H bonds to form the stabilized
3
activation counterpart allyl-1a′-sp -endo (Scheme 7), which
3
indicates that the (π-allyl)−Ir complex with the C(sp )−H
bond activation should be the thermodynamically more
Scheme 6. Possible C−H Bond Activation Pathways in the Preparation of the Active Catalytic Species with Model Ligand 1a′
a
Calculated at the M06-2X/SDD/6-31G(d,p) Level of Theory
a
The ΔG and ΔE values (in parentheses) are in kcal/mol.
4
817
dx.doi.org/10.1021/ja210923k | J. Am. Chem. Soc. 2012, 134, 4812−4821

Journal of the American Chemical Society
Article
Figure 2. Optimized structures and relative energies of intermediates and transition states for the C−H bond activation with model ligand 1a′,
calculated at the M06-2X/SDD/6-31G(d,p) level of theory.The ΔG and ΔE values (in parentheses) are in kcal/mol. The bond distances are in Å.
2
Scheme 7. Possible (π-Allyl)−Ir Complexes with Model
Ligand 1a′ Calculated at the M06-2X/SDD/6-31G(d,p)
Level of Theory
six-membered iridacycle (A(Ir−P−N) = 119.4° in TS-1a′-sp ).
3
The C(sp )−H bond activation of one methyl group affords a
a
3
more favorable situation (A(Ir−P−N) = 109.1° in TS-1a′-sp ),
making the formation of a five-membered iridacycle a faster
process.
Preparation of the (π-Allyl)−Ir Complexes. To obtain
more solid evidence, allyliridium complex K2a, cinnamyliridium
complex K2b, and crotyliridium complex K2c were synthesized
using a one-pot procedure developed by Helmchen and co-
7
workers (Scheme 8). These complexes were obtained as
Scheme 8. Synthesis of (π-Allyl)−Ir Complexes K2a−c
a
28
The ΔG and ΔE (in parentheses) are in kcal/mol.
stabilized intermediate for ligand (R,R,R )-1a. These results are
a
in accord with the single-crystal X-ray diffraction analyses
1c,6b
7
reported by Hartwig
and Helmchen.
The possible origination leading to such different chemo-
selectivities with these two systems is summarized as follows.
Electronically, the benzene ring in (R,R,R )-1a is not electron-
a
rich compared to the case for (R,R )-1e, where the benzene ring
a
is substituted by an N atom. Therefore, the C−H bond
activation at the electron-rich aromatic ring through an
oxidative addition mechanism should be more favored. In
addition, the ring size of the metallacycles also contributes to
the selectivity between two C−H bond activation pathways.
mixtures of endo and exo isomers, in ratios of 9:1 (K2a), 1:4
K2b), and 5:1 (K2c), respectively. The crystal structures of
(
30
these complexes clearly show that the metallacycle is formed
2
31,32
via C(sp )−H bond activation
of 1e (Figure 3), which is
2
3
Both the C(sp )−H and the C(sp )−H bond activation of
consistent with the above DFT calculations.
(
(
R,R )-1e will generate five-membered iridacycles. The C-
Some selected bond lengths and angles of complexes K2a−c
are given in Table 3. These complexes can be considered to be
severely distorted octahedra with the allyl ligand, an olefinic
portion (C4C5) of the cod ligand, and the phosphorus
a
2
sp )−H bond activation should be the kinetically favored
2
process. However, in the case of (R,R,R )-1a the C(sp )−H
a
bond activation of the phenyl ring will provide a highly strained
4
818
dx.doi.org/10.1021/ja210923k | J. Am. Chem. Soc. 2012, 134, 4812−4821

Journal of the American Chemical Society
Article
phosphorus atom of K2a−c are 358.5−362.1°, indicating the
relative planarity of these groups with the metal. The Ir−C3
bond of K2a is longer than the Ir−C1 bond by 0.08 Å, likely
3
3
due to the strong trans influence of the P atom. The C2−H
bond of the allyl ligand is pointing away from cod (endo),
1c
similar to the case for the Feringa ligand derived complex.
However, for cinnamyl complex K2b, the C2−H bond is
pointing toward cod (exo). The difference between Ir−C3 and
Ir−C1 amounts to 0.17 Å, probably a result of the strong trans
influence of the P atom as well as the ability of the phenyl ring
to stabilize the neighboring positive charge (see the Supporting
Information for the calculated FMOs and NBO charge
distribution in the allylic parts of complexes K2a−c). These
results suggest that the Ir−C3 bond is weaker and C3 is more
electropositive than C1. In addition, the Re face of C3 is
shielded with the chiral pocket generated by the ligand. Thus,
the nucleophile prefers to attack from the Si face at C3 to give
the R-branched product, in accord with the experimental
observation. The pocket for the allyl ligand in cinnamyl
complex K2b is much larger than that in the Feringa ligand
derived complex, which might allow the bulky ortho-substituted
cinnamyl derivatives to generally proceed with excellent yield
and enantioselectivity (see the Supporting Information for a
CPK model of complex K2b).
Interestingly, the crotyl moiety of complex K2c is in the
endo-in position, much different from that of K2b. Effected by
both the trans P ligand and the methyl substituent, the bond
distances of iridium to allyl temini are almost equal. This likely
accounts for the poor regioselectivity observed with aliphatic
allylic carbonate (entry 23, Table 2).
Reaction of (π-Allyl)−Ir Complexes K2a−c. The
stoichiometric reactions of (π-allyl)−Ir complexes K2a−c
with sodium dimethyl malonate were studied (Scheme 9).
Figure 3. Crystallographic representation of (π-allyl)−Ir complexes
K2a−c. Anions and hydrogen atoms are omitted for clarity (ellipsoids
are shown at the 30% probability level).
Scheme 9. Stoichiometric Reactions of (π-Allyl)−Ir
Complexes K2a−c
Table 3. Selected Bond Lengths and Bond Angles of (π-
Allyl)−Ir Complexes K2a−c
K2a
K2b
K2c
Bond Lengths (Å)
Ir−P
2.2733(13)
2.211(5)
2.210(5)
2.291(5)
2.097(5)
1.419(8)
1.395(8)
2.2547(16)
2.227(6)
2.241(6)
2.400(6)
2.101(6)
1.414(8)
1.392(9)
2.268(2)
2.266(7)
2.219(7)
2.267(7)
2.095(7)
1.410(12)
1.409(12)
Ir−C1
Ir−C2
Ir−C3
Ir−C13
C1−C2
C2−C3
The yields and regioselectivities were comparable to the
reactions of corresponding allylic carbonates catalyzed by
iridium catalysts prepared in situ. The reaction with 1 equiv of
K2b (4:1 exo and endo mixture) afforded 3a in 58% ee, while
Bond Angles (deg)
a
P−Ir−C1
88.43(15)
65.9(2)
80.30(16)
91.6(2)
C1−Ir−C3
C3−Ir−C5
C5−Ir−C4
C4−Ir−P
sum of Ir−plane
P−Ir−C3
63.9(2)
65.8(3)
b
84.1(2)
89.1(2)
81.6(3)
5
0 and 4 mol % of K2b gave the product in 88% and 95% ee,
36.37(19)
87.46(14)
362.06
37.0(2)
36.8(3)
86.2(2)
362.0
respectively (entry 4 and entry 2, Table 4). The stoichiometric
reaction with K2c (5:1 endo and exo mixture) provided 3k in
nearly quantitative yield. In general, the π−σ−π transformation
of the allyliridium complex is very slow. These data are in
agreement with the theoretical results on the assumption that
the transformation between the endo and exo forms of the allyl
motif is much slower than the nucleophilic attack toward the
88.20(17)
358.5
c
150.38(16)
79.82(14)
121.1(5)
141.85(16)
79.64(16)
153.8(2)
P−Ir−C13
C1−C2−C3
79.1(3)
121.7(6)
121.8(8)
a
b
c
The P−Ir−C3 angle. The C1−Ir−C5 angle. The P−Ir−C1 angle.
(
π-allyl)−Ir complex. The higher enantioselectivities under
occupying one plane and the other olefinic unit (C8C9) of
catalytic conditions might be explained by the kinetically
favorable formation of the exo form of the cinnamyliridium
complex that is attacked by the nucleophile before the π−σ−π
interconversion.
the cod and a phenyl group of the metallacycle lying apical to
1
c,6b
this plane.
The summed angles involving the olefinic
portion (C4C5) of cod, two allyl termini, and the
4
819
dx.doi.org/10.1021/ja210923k | J. Am. Chem. Soc. 2012, 134, 4812−4821

Journal of the American Chemical Society
Article
Table 4. Catalytic Reactions of Cinnamyl Methyl Carbonate 2a Using (π-Allyl)-Ir Complexes K2a−c as Catalysts
entry
K2 (x)
3a/4a
yield (%)
ee (%)
1
2
3
4
K2a (4)
K2b (4)
K2c (4)
K2b (50)
98/2
99/1
99/1
99/1
91
92
90
99
96 (R)
95 (R)
97 (R)
88 (R)
The application of (π-allyl)−Ir complexes K2a−c as the
catalysts in allylic alkylation reactions with sodium dimethyl
malonate is shown in Table 4. With 4 mol % of K2a−c, the
alkylated products were obtained in 90−92% yield with 95−
(3 mL × 3) to afford (π-allyl)−Ir complex K2a (922.5 mg, 97% yield)
as a pale yellow powder, existing as two isomers in a ratio of 9:1.
Full details may be found in the Supporting Information.
ASSOCIATED CONTENT
Supporting Information
97% ee and up to 99:1 branched to linear ratio (entries 1−3,
■
*
S
Table 4). The absolute configuration of the branched product
obtained here is also the same with that obtained with in situ
prepared catalyst. These results support that the (π-allyl)−Ir
complexes are intermediates of the catalytic cycle.
Text, figures, tables, and CIF files giving experimental
procedures, compound characterization data, crystallographic
data for the ligands and iridium complexes, and computational
details. This material is available free of charge via the Internet
at http://pubs.acs.org.
CONCLUSION
■
In summary, a series of N-aryl phosphoramidites has been
synthesized and applied in Ir-catalyzed asymmetric allylic
alkylation reactions. The reactions employing these ligands
feature broad substrate scope. Excellent regio- and enantiose-
lectivities could be achieved for ortho-substituted phenyl allylic
substrates. Mechanistically, the active iridium catalysts derived
AUTHOR INFORMATION
Corresponding Author
E-mail: slyou@sioc.ac.cn.
■
*
Notes
The authors declare no competing financial interest.
2
from the phosphoramidite ligand 1e are formed by a C(sp )−H
activation of the N-aryl group in the ligands. The DFT
ACKNOWLEDGMENTS
The National Basic Research Program of China (973 Program
009CB825300) and the National Natural Science Foundation
■
2
calculations and X-ray crystallographic analyses of the (π-
2
allyl)−Ir complexes explain and support this C(sp )−H
of China (Nos. 21121062, 20923005, 20932008, 21025209) are
acknowledged for generous financial support. We are also
grateful to Prof. Gunter Helmchen for helpful discussion, Prof.
̈
Xuebing Leng for assistance with X-ray crystallographic
analyses, and Prof. Yong-Gui Zhou and Prof. Qinhua Fan for
providing the chiral amines.
activation mode and the excellent selectivities obtained in the
allylic alkylation reactions. The studies here highly broaden the
substrate scope in Ir-catalyzed allylic substitution reactions and
provide a better understanding of the reaction mechanism as
well as a basis for future ligand design.
EXPERIMENTAL SECTION
General Procedure for Ir-Catalyzed Allylic Alkylation (Entry
, Table 1). In a dry Schlenk tube filled with argon, [Ir(cod)Cl] (3.3
■
REFERENCES
1) For recent reviews: (a) Helmchen, G.; Dahnz, A.; Du
■
(
̈
bon, P.;
8
2
Schelwies, M.; Weihofen, R. Chem. Commun. 2007, 675. (b) Helm-
chen, G. In Iridium Complexes in Organic Synthesis; Oro, L. A., Claver,
C., Eds.; Wiley-VCH: Weinheim, Germany, 2009; pp 211−250.
c) Hartwig, J. F.; Stanley, L. M. Acc. Chem. Res. 2010, 43, 1461.
d) Hartwig, J. F.; Pouy, M. J. Top. Organomet. Chem. 2011, 34, 169.
e) Liu, W.-B.; Xia, J.-B.; You, S.-L. Top. Organomet. Chem. 2012, 38,
55.
2) For the first example: Takeuchi, R.; Kashio, M. Angew. Chem., Int.
Ed. Engl. 1997, 36, 263.
3) Janssen, J. P.; Helmchen, G. Tetrahedron Lett. 1997, 38, 8025.
4) (a) Bartels, B.; Helmchen, G. Chem. Commun. 1999, 741.
(b) Ohmura, T.; Hartwig, J. F. J. Am. Chem. Soc. 2002, 124, 15164.
c) Bartels, B.; García, Y. C.; Helmchen, G. Eur. J. Org. Chem. 2003,
097. (d) Tissot-Croset, K.; Polet, D.; Alexakis, A. Angew. Chem., Int.
Ed. 2004, 43, 2426. (e) Lipowsky, G.; Miller, N.; Helmchen, G. Angew.
Chem., Int. Ed. 2004, 43, 4595. (f) Leitner, A.; Shekhar, S.; Pouy, M. J.;
Hartwig, J. F. J. Am. Chem. Soc. 2005, 127, 15506. (g) Onodera, G.;
Watabe, K.; Matsubara, M.; Oda, K.; Kezuka, S.; Takeuchi, R. Adv.
Synth. Catal. 2008, 350, 2725. (h) Fuji, K.; Kinoshita, N.; Tanaka, K.;
Kawabata, T. Chem. Commun. 1999, 2289. (i) Fischer, C.; Defieber,
C.; Suzuki, T.; Carreira, E. M. J. Am. Chem. Soc. 2004, 126, 1628.
(j) Defieber, C.; Ariger, M. A.; Moriel, P.; Carreira, E. M. Angew.
Chem., Int. Ed. 2007, 46, 3139. (k) Roggen, M.; Carreira, E. M J. Am.
Chem. Soc. 2010, 132, 11917. (l) Miyabe, H.; Matsumura, A.;
mg, 0.005 mmol, 2 mol %), phosphoramidite ligand 1e (4.6 mg, 0.01
mmol, 4 mol %), and n-propylamine (0.3 mL) were dissolved in THF
0.5 mL). The reaction mixture was heated at 50 °C for 30 min, and
then the volatile solvents were removed under vacuum to give a yellow
solid. In this tube, allylic carbonate 2a (0.25 mmol), THF (1.5 mL),
and sodium dimethyl malonate solution (0.5 mmol in 1.0 mL THF, 2
equiv) were added and stirred at room temperature until the reaction
was complete. The reaction was quenched by saturated aqueous
NH Cl (2 mL) and extracted with Et O (5 mL × 3). The combined
(
(
(
(
1
(
4
2
(
(
organic layer was washed with brine, separated, and dried over
anhydrous Na SO . The solvent was evaporated, and the ratio of
2
4
1
regioisomers (branched to linear b:l) was determined by H NMR of
the crude reaction mixture. The residue was purified by silica gel
column chromatography using 10/1 PE/EA as the eluent to give the
desired product 3a.
(
1
General Procedure for the Preparation of (π-Allyl)−Ir
Complexes K2 (K2a). A solution of [Ir(cod)Cl] (335.5 mg, 0.5
2
mmol) and ligand 1e (464.1 mg, 1.0 mmol) in dry THF (10 mL) was
added to a flame-dried Schlenk tube under an argon atmosphere. After
the mixture was stirred for 30 min at room temperature, AgOTf (258.2
mg, 1.0 mmol) and allyl methyl carbonate (220.5 mg, 3.8 mmol) were
added subsequently and the reaction mixture was stirred overnight to
give a pale suspension. The precipitate was filtered, and the filtrate was
removed under reduced pressure. The residue was washed with Et O
2
4
820
dx.doi.org/10.1021/ja210923k | J. Am. Chem. Soc. 2012, 134, 4812−4821

Journal of the American Chemical Society
Article
Moriyama, K.; Takemoto, Y. Org. Lett. 2004, 6, 4631. (m) Nemoto, T.;
Sakamoto, T.; Fukuyama, T.; Hamada, Y. Tetrahedron Lett. 2007, 48,
(22) (a) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972,
56, 2257. (b) Hariharan, P.; Pople, J. A. Theor. Chim. Acta. 1973, 28,
213.
4
(
2
(
977. (n) Kimura, M.; Uozumi, Y. J. Org. Chem. 2007, 72, 707.
5) For a review: Teichert, J. F.; Feringa, B. L. Angew. Chem., Int. Ed.
010, 49, 2486.
6) (a) Kiener, C. A.; Shu, C.; Incarvito, C. D.; Hartwig, J. F. J. Am.
(23) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B
2
(
2
009, 113, 6378.
24) (a) Yu, Z.-X.; Wender, P. A.; Houk, K. N. J. Am. Chem. Soc.
004, 126, 9154. (b) Lee, S. I.; Fukumoto, Y.; Chatani, N. Chem.
Commun. 2010, 46, 3345.
25) Chloride anion can readily be abstracted with Ag(I) salt from
Chem. Soc. 2003, 125, 14272. (b) Madrahimov, S. T.; Markovic, D.;
Hartwig, J. F. J. Am. Chem. Soc. 2009, 131, 7228.
(
(
7) (a) Spiess, S.; Raskatov, J. A.; Gnamm, C.; Bro
Helmchen, G. Chem. Eur. J. 2009, 15, 11087. (b) Raskatov, J. A.;
Spiess, S.; Gnamm, C.; Brodner, K.; Rominger, F.; Helmchen, G.
Chem. Eur. J. 2010, 16, 6601.
8) (a) Lopez, F.; Ohmura, T.; Hartwig, J. F. J. Am. Chem. Soc. 2003,
̈
dner, K.;
the neutral Ir(I) complex L(cod)IrCl. The calculated Gibbs free
energy changes of the chloride-abstraction process are −10.0 kcal/mol
̈
(
L = 1e′) and −12.7 kcal/mol (L = 1a′).
(
26) All the figures of the optimized structures in this article were
prepared by CYLview: Legault, C. Y. CYLView, 1.0b; Universite de
Sherbrooke, Montreal, Quebec, Canada, 2009; http://www.cylview.
org.
27) For a computational study of the stability of π-allyl
(
́
́
125, 3426. (b) Shu, C.; Leitner, A.; Hartwig, J. F. Angew. Chem., Int. Ed.
́
2
004, 43, 4797. (c) Alexakis, A.; Polet, D. Org. Lett. 2004, 6, 3529.
(
d) Polet, D.; Alexakis, A.; Tissot-Croset, K.; Corminboeuf, C.;
(
Ditrich, K. Chem. Eur. J. 2006, 12, 3596. (e) Yamashita, Y.;
Gopalarathnam, A.; Hartwig, J. F. J. Am. Chem. Soc. 2007, 129,
intermediates in molybdenum-catalyzed asymmetric allylic alkylations,
see: Luft, J. A. R.; Yu, Z.-X.; Hughes, D. L.; Lloyd-Jones, G. C.; Krska,
S. W.; Houk, K. N. Tetrahedron: Asymmetry 2006, 17, 716.
7508. (f) Pouy, M. J.; Leitner, A.; Weix, D. J.; Ueno, S.; Hartwig, J. F.
Org. Lett. 2007, 9, 3949. (g) Liu, W.-B.; He, H.; Dai, L.-X.; You, S.-L.
Org. Lett. 2008, 10, 1815. (h) Liu, W.-B.; Zheng, S.-C.; He, H.; Zhao,
X.-M.; Dai, L.-X.; You, S.-L. Chem. Commun. 2009, 6604. (i) Stanley,
L. M.; Hartwig, J. F. J. Am. Chem. Soc. 2009, 131, 8971.
(
28) The optimized structures of these complexes are given in the
Supporting Information.
(29) Mezzetti et al. reported an NMR study on the structure of an
Ir(I) complex with a Feringa ligand in solution, which showed that the
most stable species formed upon chloride abstraction of [Ir(cod)
(
9) Liu, W.-B.; He, H.; Dai, L.-X.; You, S.-L. Synthesis 2009, 10, 2076.
10) (a) Wu, Q.-F.; He, H.; Liu, W.-B.; You, S.-L. J. Am. Chem. Soc.
+
(
(R,R,R )-1a]Cl is Ir(cod)[(R,R,R )-1a] . These results do not conflict
a
a
2
2
010, 132, 11418. (b) Zhuo, C.-X.; Liu, W.-B.; Wu, Q.-F.; You, S.-L.
Chem. Sci. 2012, 3, 205.
11) Krasnov, V. P.; Levit, G. L.; Bukrina, I. M.; Andreeva, I. N.;
with our calculations because the second molecule of the Feringa
ligand should dissociate when the cyclometalation proceeds. Osswald,
(
̈
T.; Ruegger, H.; Mezzetti, A. Chem. Eur. J. 2010, 16, 1388.
Sadretdinova, L. S.; Korolyova, M. A.; Kodess, M. I.; Charushin, V. N.;
Chupakhin, O. N. Tetrahedron: Asymmetry. 2003, 14, 1985.
(
(30) The racemic iridium complexes K2a−c were used for the single-
crystal X-ray diffraction analyses; see the Supporting Information for
details.
12) (a) Lu, S.-M.; Han, X.-W.; Zhou, Y.-G. Adv. Synth. Catal. 2004,
2
(31) For the C(sp )−H activation at Cp*−iridium complexes to
346, 909. (b) Lu, S.-M.; Wang, Y.-Q.; Han, X.-W.; Zhou, Y.-G. Angew.
form iridacycles, see: (a) Kisenyi, J. M.; Sunley, G. J.; Cabeza, J. A.;
Smith, A. J.; Adams, H.; Salt, N. J.; Maitlis, P. M. J. Chem. Soc., Dalton
Trans. 1987, 2459. (b) Davies, D. L.; Donald, S. M. A.; Al-Duaij, O.;
Chem., Int. Ed. 2006, 45, 2260. (c) Wang, Z.-J.; Deng, G.-J.; Li, Y.; He,
Y.-M.; Tang, W.-J.; Fan, Q.-H. Org. Lett. 2007, 9, 1243. (d) Wang, X.-
B.; Zhou, Y.-G. J. Org. Chem. 2008, 73, 5640. (e) Li, Z.-W.; Wang, T.-
L.; He, Y.-M.; Wang, Z.-J.; Fan, Q.-H.; Pan, J.; Xu, L.-J.. Org. Lett.
Macgregor, S. A.; Po
̈
lleth, M. J. Am. Chem. Soc. 2006, 128, 4210.
(c) Davies, D. L.; Donald, S. M. A.; Al-Duaij, O.; Fawcett, J.; Little, C.;
2
008, 10, 5265. (f) Zhou, H.; Li, Z.; Wang, Z.; Wang, T.; Xu, L.; He,
Y.; Fan, Q.-H.; Pan, J.; Gu, L.; Chan, A. S. C. Angew. Chem., Int. Ed.
008, 47, 8464. (g) Gou, F.-R.; Li, W.; Zhang, X. M.; Liang, Y.-M. Adv.
Macgregor, S. A. Organometallics 2006, 25, 5976. (d) Li, L.;
Brennessel, W. W.; Jones, W. D. Organometallics 2009, 28, 3492.
2
(
32) For an enantioselective transfer hydrogenative coupling reaction
Synth. Catal. 2010, 352, 2441. (h) Zhang, D.-Y.; Wang, D.-S.; Wang,
M.-C.; Yu, C.-B.; Gao, K.; Zhou, Y.-G. Synthesis 2011, 12, 2796.
(
2
using a C(sp )−H activated iridacycle complex as catalyst, see:
a) Kim, I. S.; Ngai, M.-Y.; Krische, M. J. J. Am. Chem. Soc. 2008, 130,
4891. For a recent review: (b) Hassan, A.; Krische, M. J. Org. Process
Res. Dev. 2011, 15, 1236.
33) Appleton, T. G.; Clark, H. C.; Manzer, L. E. Coord. Chem. Rev.
1973, 10, 335.
(
13) (a) Alexakis, A.; Rosset, S.; Allamand, J.; March, S.; Guillen, F.;
1
Benhaim, C. Synlett 2001, 1375−1378. (b) Naasz, R.; Arnold, L. A.;
Minnaard, A. J.; Feringa, B. L. Angew. Chem., Int. Ed. 2001, 40, 927−
(
9
30. (c) Polet, D.; Alexakis, A. Synthesis 2004, 2586−2590.
14) Ligand 1j has been previously synthesized; see: Toselli, N.;
Martin, D.; Achard, M.; Tenaglia, A.; Burgi, T.; Buono, G. Adv. Synth.
Catal. 2008, 350, 280.
15) Isolated yield of single enantiomer after recrystallization of the
diastereoisomers; see the Supporting Information for details.
16) The structures of 1e,g,k,l were all confirmed by X-ray
crystallographic analyses; see the Supporting Information for details.
17) (a) Bedford, R. B.; Castillon, S.; Chaloner, P. A.; Claver, C.;
Fernandez, E.; Hitchcock, P. B.; Ruiz, A. Organometallics 1996, 15,
990. (b) Bartels, B.; Garcıa-Yebra, C.; Rominger, F.; Helmchen, G.
Eur. J. Inorg. Chem. 2002, 2569.
18) See the Supporting Information for computational details.
(
̈
(
(
(
̀
3
́
(
Helmchen et al. has conducted computational studies on Ir-catalyzed
7
allylic substitution reactions with a Feringa-type ligand.
(
19) Frisch, M. J. et al. Gaussian09, Revision A.01; Gaussian, Inc.,
Wallingford, CT, 2009.
20) (a) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215.
b) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157.
21) (a) Bergner, A.; Dolg, M.; Kuechle, H.; Stoll, H.; Preuss, H. Mol.
(
(
(
Phys. 1993, 80, 1431. (b) Kaupp, M.; Schleyer, P. v. R.; Stoll, H.;
Preuss, H. J. Chem. Phys. 1991, 94, 1360. (c) Dolg, M.; Stoll, H.;
Preuss, H.; Pitzer, R. M. J. Phys. Chem. 1993, 97, 5852.
4
821
dx.doi.org/10.1021/ja210923k | J. Am. Chem. Soc. 2012, 134, 4812−4821