Ir-catalysed asymmetric allylic substitutions with cyclometalated (phosphoramidite)Ir complexes - resting states, catalytically active (π-Allyl)Ir complexes and computational exploration

Article abstract of DOI:10.1002/chem.200903465

Mechanistic aspects of allylic substitutions with iridium catalysts derived from phosphoramidites by cyclometalation were investigated. The determination of resting states by ³¹P NMR spectroscopy led to the conclusion that the cyclometalation process is reversible. A novel, one-pot procedure for the preparation of (πallyl)Ir complexes was developed, and these complexes were characterised by X-ray crystal structure analyses and spectral data. They are fully active catalysts of the allylic substitution reaction. DFT calculations on the allyl complexes, transition states of the allylic substitution and product olefin complexes gave further mechanistic insight.

Full text of DOI:10.1002/chem.200903465

FULL PAPER
DOI: 10.1002/chem.200903465
Ir-Catalysed Asymmetric Allylic Substitutions with Cyclometalated
(Phosphoramidite)Ir Complexes—Resting States, Catalytically Active
ACHTUNGTRENNUNG(p-Allyl)Ir Complexes and Computational Exploration
Jevgenij A. Raskatov, Stephanie Spiess, Christian Gnamm, Kerstin Brçdner,
Frank Rominger, and Gꢀnter Helmchen*[a]
Abstract: Mechanistic aspects of allylic
substitutions with iridium catalysts de-
rived from phosphoramidites by cyclo-
metalation were investigated. The de-
termination of resting states by
31P NMR spectroscopy led to the con-
clusion that the cyclometalation pro-
cess is reversible. A novel, one-pot pro-
cedure for the preparation of (p-
allyl)Ir complexes was developed, and
these complexes were characterised by
X-ray crystal structure analyses and
spectral data. They are fully active cat-
alysts of the allylic substitution reac-
tion. DFT calculations on the allyl
complexes, transition states of the allyl-
ic substitution and product olefin com-
plexes gave further mechanistic insight.
Keywords: allylic compounds · al-
lylic substitution · asymmetric catal-
ysis · iridium · reaction mechanisms
Introduction
The Ir-catalysed allylic substitution has been developed into
a reliable tool of organic synthesis over the last few years.
The reaction allows preparation of branched, chiral allylic
derivatives 2 from readily available monosubstituted allylic
substrates 1 with high selectivity (Scheme 1). A broad range
of C-, N- and O-nucleophiles can be employed.[1] After
probing a variety of Ir complexes the presently best catalysts
ꢀ
are generated by base-induced C H activation from [{Ir-
A
R
(dbcot=dibenzocyclooctatetraene)[4] and a chiral phosphor-
amidite. Several experimental protocols have been devel-
oped for the reaction. The salt-free version has proven par-
ticularly reliable;[5] this method was originally worked out
by Tsuji and co-workers for Rh-catalysed allylic substitu-
tions.[6] It involves production of the nucleophile from its
conjugate acid by reaction with carbonate or alkoxide gener-
ated in the substitution step. Because of a low content of
Scheme 1. Ir-catalysed allylic substitution.
base and usually homogeneous reaction mixtures in allylic
alkylations, the method is also well suited for mechanistic
investigations.
We present here a full account of a mechanistic investiga-
tion, which addresses several aspects of the Ir-catalysed al-
lylic substitution: 1) Determination of the resting state of
the alkylation reaction, which previously was not clearly es-
tablished. 2) Isolation of a catalytically active (allyl)Ir com-
plex that was elusive at the outset of our investigation. 3)
Quantum chemical calculations on experimentally estab-
lished intermediates of the catalytic cycle.[7]
One result of this study is a remarkably simple and broad-
ly applicable method for preparation of (p-allyl)Ir com-
plexes. Their characterisation by X-ray crystal structure
analysis and use as catalysts has furnished considerable in-
[a] Dr. J. A. Raskatov, S. Spiess, Dr. C. Gnamm, K. Brçdner,
Dr. F. Rominger, Prof. Dr. G. Helmchen
Organisch-Chemisches Institut
Ruprecht-Karls-Universitꢀt Heidelberg
Im Neuenheimer Feld 270, 69120 Heidelberg (Germany)
Fax : (+49)6221-544205
E-mail: g.helmchen@oci.uni-heidelberg.de
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200903465.
Chem. Eur. J. 2010, 16, 6601 – 6615
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6601

sight into the catalytic cycle. We have reported some of the
results in a communication.[8] In the meantime we have ex-
tended our studies considerably. As ligand, we have mainly
employed the particularly effective phosphoramidite L2 that
was introduced by Alexakis et al.[9] Parallel work was car-
ried out by Hartwig et al., who have been using the ligand
L1.[10,11] They also succeeded in preparing and characterising
(p-allyl)Ir complexes, using a preparative procedure, which
is different from ours.
in situ generated K2 with weak acids, for example acetic
acid, effects the reformation of K1. Now we have found that
reformation of K1 also proceeds with allylic substrates and
that the catalytically inactive complex K1 constitutes the
resting state of the allylic substitution.
A typical series of experiments was conducted as follows
(cf. 31P NMR spectra in Figure 1). Mixing [{Ir
ACTHNUGTRNE(NUG cod)Cl}2] and
L2 (2 equiv) in THF produced the orange complex K1 (A in
A summary of our current view on the mechanism of the
Ir-catalysed allylic substitution is given in Scheme 2. Com-
plexes K1 are formed by simply combining [{IrACTHNURGTNEUNG(cod)Cl}2]
and a phosphoramidite. These complexes are catalytically
Figure 1. Reversibility of the C H activation. 31P NMR spectra: A) Com-
ꢀ
plex K1, generated by reaction of [{Ir
A
ꢀ
(0.02 mmol) in [D8]THF (0.5 mL); B) C H activated species K2 after
1 h, generated by addition of TBD (0.04 mmol) to solution A (K1);
C, D) Reformed complex K1, measurement 10 min after the addition of
cinnamyl acetate or cinnamyl diethyl phosphate (0.5 mmol) to solution B
(K2).
Figure 1). This complex displays a peak at 120.2 ppm and an
additional small peak at 131.6 ppm, which corresponds to
the hydrido complex K1H (see Scheme 3 and the Support-
Scheme 2. Catalysis cycle for an allylic substitution reaction using a cata-
lyst prepared in situ from K1 by treatment with an amine base B; the
complex K1 constitutes the resting state.
ing Information).[12] Addition of TBD effected C H activa-
ꢀ
tion, yielding complex K2 as a pair of diastereoisomers. The
31P NMR spectrum (B in Figure 1) displayed two pairs of
doublets that were broadened. Both K1 and K2 have been
previously described.[4] Addition of an allylic substrate to
the solution of K2 rapidly led to reformation of the complex
K1. As examples, reactions with cinnamyl acetate (C) and
cinnamyl diethyl phosphate (D) are illustrated by 31P NMR
spectra in Figure 1.
inactive, as they do not react with an allylic substrate. Start
of a substitution reaction requires the presence of a base,
which effects the formation of the cyclometalated complexes
K2 (18 valence electrons (VE)).[3] If the nucleophile is suffi-
ciently basic, as is true, for example, for benzylamine and
propylamine, an additional base is not required. Among
non-nucleophilic bases TBD (1,5,7-triaza-bicycloACTHNUTRGENUG[N 4.4.0]dec-
5-ene) is preferred in our laboratory, because it allows the
activation to be carried out in a few minutes. Complex K2 is
not a directly catalytically active species because it is coordi-
natively saturated; furthermore the reaction rate of an allyl-
ic substitution is optimal with a ratio L*/Ir=1:1, an excess
of L* leads to lowering of the reaction rate. As the ligand
L* of K2 can be easily replaced by other ligands, the most
likely active species in the catalytic cycle is the 16 VE inter-
mediate K3.
Results and Discussion
Resting states and (p-allyl)Ir complexes
Resting state of the allylic substitution upon use of an in situ
Scheme 3. Formation of (hydrido)Ir complexes K3H and allyl complexes
K4.
catalyst: Some time ago we have shown[4] that treatment of
6602
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 6601 – 6615

Asymmetric Catalysis
FULL PAPER
A further significant observation concerning the resting
state of the allylic substitution was made when 31P NMR
spectra were measured during and after completed alkyla-
tion and amination reactions (Figure 2). The addition of cin-
tion of the (p-allyl)Ir complexes of type K4. Evidently, a
chloride ion and a proton are required for the reformation
of K1; these must originate from residual [{IrACTHNURTGNENG(U cod)Cl}2] and
the ammonium salt (BH+Clꢀ in Scheme 2) formed upon cat-
alyst activation by treating K1 with TBD. We wondered
what would be the result of scavenging chloride ions. This
was carried out with silver salts; unexpectedly, this directly
led to the allyl complexes K4 (Scheme 4).
Scheme 4. Synthesis of (p-allyl)Ir complexes from simple components.
Mechanistic aspects of the new method are described for
the formation of K4a. The preparation was monitored by
31P NMR spectra (Figure 3). First, the complex K1, R=Ph,
was prepared by mixing [{IrACHTNUTRGNENG(U cod)Cl}2] and L2 in THF (A in
Figure 3).[4] Addition of AgClO4 gave rise to a precipitate of
AgCl. The corresponding solution displayed two singlets
with an intensity ratio of 3:1 (B in Figure 3) in the 31P NMR
spectrum. Diastereomers with the constitution K3H, X=
ClO4, are assigned to the corresponding compounds
Figure 2. Resting state of the allylic substitution. A solution of K2 was
generated by addition of TBD (0.04mmol) to a solution of [{IrACHTNUTRGNEUNG(cod)Cl}2]
(0.01mmol) and L2 (0.02mmol) in [D8]THF (0.5mL). 31P NMR spectra:
A) Reformed complex K1, generated by addition of cinnamyl methyl car-
bonate (0.50mmol) and CH2ACHTNUGTRNE(GNU CO2CH3)2 (0.60mmol) to the solution of
(Scheme 3). This assignment is based on the following data.
K2; B) Spectrum after 40min at RT. The reaction was complete; A’) Re-
formed complex K1, generated by addition of cinnamyl methyl carbonate
(0.50mmol) and aniline (0.60mmol) to K2; B’) Spectrum after>10h at
RT. The reaction was complete.
1
ꢀ
H NMR spectra (not shown) displayed Ir H resonances at
namyl methyl carbonate and a nucleophile, dimethyl malo-
nate or aniline, to a solution of K2 in THF led to the refor-
mation of the complex K1 (A and A’ in Figure 2) in both
cases. K1 remained the major species after completion of
the alkylation reaction (B). In the case of the completed
amination (B’ in Figure 2), K1 was accompanied by a
second species showing a peak at dP =147 ppm. Hartwig
et al.[10] have found
a product complex of type K5
(Scheme 2) for the corresponding reaction upon use of L1
as ligand (dP =148-150 ppm). Thus, the same type of com-
pound is assumed for the mixture B’ in Figure 2.[13]
Figure 3. Preparation of K4a. 31P NMR spectra: A) [{Ir
ACHTUNGTRENNUNG(cod)Cl}2]
(0.01 mmol)+(S,S,S)-L2 (0.02 mmol) (K1) in [D8]THF (0.5 mL) after
30 min; B) Addition of AgClO4 (0.02 mmol) to solution A; C) Addition
of cinnamyl methyl carbonate (0.04 mmol) to solution B, spectrum after
4.5 h at RT; D) Spectrum after 14 h at RT.
Preparation of the (p-allyl)Ir complexes K4: The observation
that complex K1 constitutes the resting state of the allylic
substitution led us to a very simple method for the prepara-
Chem. Eur. J. 2010, 16, 6601 – 6615
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6603

G. Helmchen et al.
Table 1. Selected NMR data of the allylic fragment of the (p-allyl)Ir complexes K4 (numbering of atoms see Figure 5).
1H NMR [ppm] 13C NMR [ppm]
C1A C2A C3A
2.76–2.90 (m) 2.35 (d, J=10.8 Hz) 86.51
K4b 5.74 (d, J=12.4 Hz) 5.05 (ddd, J=11.6 Hz, 11.6 Hz, 9.7 Hz) 2.77–2.93 (m) 2.09 (d, J=11.3 Hz) 86.26
31P NMR [ppm]
major (minor) isomer
1A-H
2A-H
3A,syn-H
3A,anti-H
K4a 5.91 (d, J=12.0 Hz) 4.93–4.99 (m)
98.21 41.17 120.0
98.15 40.63 119.0
K4c 4.85–4.94 (m)
K4d 4.68–4.82 (m)
K4e 4.88–4.95 (m)
K4 f 4.79–4.86 (m)
4.22 (ddd, J=10.7 Hz, 10.7 Hz, 8.5 Hz) 2.84–2.96 (m) 2.20 (d, J=11.3 Hz) 76.76 107.72 44.71 123.1 (126.6, 4%)
4.22–4.32 (m)
4.18–4.24 (m)
2.79–2.98 (m) 1.94 (d, J=11.2 Hz) 76.51 107.70 44.07 122.3 (126.0, 3%)
2.84–2.94 (m) 2.22–2.41 (m)
76.68 107.72 44.64 123.2
4.53 (ddd, J=11.5 Hz, 11.5 Hz, 7.7 Hz) 2.86–2.94 (m) 2.32–2.41 (m)
77.03 107.64 45.20 122.8 (126.2, 2%)
K4h 5.98 (d, J=12.1 Hz) 4.98 (ddd, J=11.7 Hz, 11.6 Hz, 7.8 Hz) 2.74–2.91 (m) 2.43 (d, J=10.9 Hz) 86.52
98.26 41.16 120.2
ꢀ24.18 ppm and ꢀ21.35 ppm (3:1); the most intense signal
of the ESI-FT-ICR-MS possessed the mass of K3H+ (m/z
calcd and found 900.28). The viability of these (hydrido)Ir
complexes was assessed by DFT calculations that are de-
scribed later. Two likely reaction pathways for their forma-
tion are described in Scheme 3, which cannot be distinguish-
ed presently.[14]
The solution B (Figure 3) was treated with cinnamyl
methyl carbonate (1a), whereupon the two 31P NMR sin-
glets vanished over a period of 14 h and a singlet at
120.0 ppm appeared (D in Figure 3). The corresponding
compound was identified by NMR experiments as well as
ESI-MS, and a considerable time later also by X-ray crystal-
lography, as the (p-allyl)Ir complex K4a (Scheme 4).[15]
Methanol formed from the carbonate was detected by
1H NMR spectroscopy. The (p-allyl)Ir complex K4a could
also be obtained by reaction of K3H with cinnamyl acetate
or cinnamyl phosphate instead of the carbonate.
119.0 ppm), K4e (dP =123.2 ppm) and K4h (dP =120.2 ppm)
were obtained isomerically pure (31P NMR spectroscopy). In
THF as solvent, the (p-allyl)Ir complex K4c displayed a
main peak at dP =123.1 ppm and a second peak (4%) at
dP =126.6 ppm, as was the case for K4d (dP =122.3,
126.0 ppm) and K4 f (dP =122.8, 126.2 ppm) (Table 1). The
second peak (2-4%) originates from an isomer as observed
1
from H NMR spectroscopy.
Allylic substitutions and resting state upon use of (p-allyl)Ir
complexes K4 as catalysts: Availability of the (p-allyl)Ir
complexes allowed their use as catalysts of the allylic substi-
tution. Results of representative substitution reactions ob-
tained with the in situ catalyst and complexes K4 as catalysts
are presented in Table 2. Upon use of complexes K4 as cata-
lysts under salt free conditions, that is, with CH2ACHTNUTRGEN(UGN CO2CH3)2
or HNPhth as pronucleophiles, an additional base (TBD)
was added in a catalytic amount in order to start the reac-
tion. Under the in situ conditions, TBD was introduced in
the activation step. The air stability of the complexes K4 al-
The preparation of the (p-allyl)Ir complex K4a is de-
scribed above as a two-step procedure. A one-pot prepara-
tion was also feasible and al-
lowed the convenient prepara-
Table 2. Allylic substitutions corresponding to Scheme 1 using an in situ generated catalyst and complexes K4
tions of the additional com-
plexes K4b–h (Scheme 4). The
as catalysts.[a]
R
Pronucleophile
[MNu]
Catalyst
Base
[mol%]
t [h]
Yield
2/3[c]
ee [%][d]
(p-allyl)Ir
complexes
were
G
N
2+3 [%][b]
found to be stable against
oxygen and water, and it was
possible to remove trace impur-
ities by conventional column
chromatography. Structures of
several complexes were deter-
mined by X-ray crystal struc-
ture analysis (see below).
1
2
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
n-Pr
n-Pr
n-Pr
n-Pr
CH
CH
CH
NaCH
NaCH
NaCH
NaCH
NaCH
NaN
NaN
HNPhth
HNPhth
NaCH
NaCH
HNPhth
HNPhth
CH
CH
G
in situ
K4c
K4c
in situ
K4a
K4a
K4b
K4h
in situ
K4c
in situ
K4c
in situ
K4c
in situ
K4c
in situ
K4a
–
2
89
79
60
92
84
86
85
84
80
92
95
93
98
92
82
88
90
83
99:1
98:2
98:2
99:1
99:1
99:1
99:1
99:1
97:3
99:1
96:4
96:4
89:11
91:9
94:6
95:5
82:18
79:21
97
96
96
98
96
98
97
98
99
96
98
97
98
97
96
96
99
98
TBD (8)
TBD (8)
–
–
TBD (8)
–
–
–
0.5
1.5
0.5
0.5
0.5
0.5
0.5
3[e]
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
0.75
The formation of K4a from
K3H requires the abstraction
–
–
3
2.5
1
TBD (8)
–
–
–
TBD (8)
–
TBD (8)
of
a proton (cf. Scheme 3).
3
Therefore, an additional base
might accelerate this step.
Indeed, with an excess of NEt3
as additive complete conversion
to K4a could be obtained after
a reaction time of 15 min rather
than 14 h without NEt3 (see the
Supporting Information).
10.5
2.5
2
18
5
CH2OCPh3
CH2OCPh3
[a] Allylic substitution at RT: a) NuH as pronucleophile (“salt-free conditions”): A solution of [{IrACHTUNGTRENNUNG(cod)Cl}2]
(2 mol%), L2 (4 mol%) and TBD (8 mol%) (in situ procedure) or K4 (4 mol%) in abs. THF (0.5 mL) was
prepared; then carbonate 1 (0.5 mmol), a pronucleophile (0.6 mmol) and possibly TBD (8 mol%) were added.
b) NaNu as pronucleophile: As in a), however, a freshly prepared solution of NaNu in THF (0.5 mL) was
used. [b] Isolated yield. [c] Determined by 1H NMR spectroscopy of crude products. [d] Determined by chiral
The complexes K4a (dP =
120.0 ppm),
K4b
(dP = HPLC (s. Experimental Section). [e] This reaction was carried out without inert atmosphere.
6604
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 6601 – 6615

Asymmetric Catalysis
FULL PAPER
lowed allylic substitutions to be run without an inert atmos-
phere, which is not possible with the in situ catalysts derived
from cod (entry 3). The generally good agreement of the re-
sults supports the view that the (p-allyl)Ir complexes are in-
termediates of the catalytic cycle.
Determination of the resting state of reactions catalysed
by a pure allyl complex gave further interesting results
(Figure 4). Thus, a reaction mixture prepared from cinnamyl
plexes, respectively (Figure 5, R=H and R¼H, respective-
ly). Configurational details could not be determined reliably
on the basis of spectroscopic data, and DFT calculations
(see below) indicated fairly small energy differences be-
tween some of the isomers. Fortunately, our new procedure
for preparation of the complexes allowed us to vary the sub-
stituent at the allyl moiety and the anion of the complexes
quite readily. Crystals of several complexes could be ob-
Figure 5. Atom numbering of complexes K4 and model structures as de-
scribed in Figure 8.
tained under appropriate crystallisation conditions
(Table 3).[16]
Figure 4. 31P NMR spectra: A) K4c (0.02 mmol) in [D8]THF (0.5 mL); B)
Addition of the carbonate 1a (0.5 mmol), CH2ACHTNUTRGNE(UNG CO2Me)2 (0.6 mmol) and
As example the structure of the cinnamyl complex K4b is
displayed in Figure 6. As clearly visible from the CPK
model, the 1-phenylallyl moiety sits in a pocket, which sur-
rounds the moiety C1A–C3A very closely. The pocket is
formed by cod, the oxygen of the (OCH3)C6H4 group a, one
of the O-(1-naphthyl) groups and the aryl group b. The sub-
stituent R, here phenyl, protrudes partially; the ortho-hydro-
gen atoms are located at the rim of the pocket. This is likely
the cause for the fact that allylic substitutions with ortho-
substituted cinnamyl derivatives generally proceed with
comparatively low degrees of enantioselectivity.[9a,17]
Concerning the steric course of the Ir-catalysed allylic
substitution, the following observation is significant. Attack
of a nucleophile at C1A of a p-allyl complex conformational-
ly corresponding to the crystal structure would yield a prod-
uct with the absolute configuration that has been generally
observed experimentally.[1a,18]
TBD (0.04 mmol) to solution A. Measurement after 10 min at RT; C)
Measurement after 1.5 h at RT. The reaction was complete after 30 min;
D) Measurement after 20 h at RT. The spectrum was printed with sensi-
tivity doubled compared to A–C.
methyl carbonate (1a), dimethyl malonate, and catalytic
amounts of TBD (8 mol%) and K4c (4 mol%) only dis-
played the 31P NMR peak of the (p-allyl)Ir complex K4a
after mixing of the reaction partners (Figure 4, B). The peak
of K4a persisted during the reaction and vanished slowly
after consumption of the carbonate 1a (C); a very broad
peak at dP =148 ppm remained in case of the salt-free reac-
tion (D). This peak became sharp upon cooling (E). Accord-
ingly, the allyl complex is the resting state of the alkylation,
if return to a complex of type K1 is not possible.
X-ray crystal structures of (p-allyl)Ir complexes: The number
of isomeric complexes of the type K4 amounts to 8 and 16
for the parent p-allyl and monosubstituted p-allyl com-
Of particular interest are the bond lengths of iridium to
the allylic carbon atoms (Table 3). In all examples, the bond
Table 3. Selected bond lengths [ꢂ] and bond angles [8] of the (p-allyl)Ir complexes K4b,d,e,f.
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
Ir C1A
Ir C2A
Ir C3A
C1A C2A C2A C3A C1A-C2A-C3A P-Ir-CH2 Sum of the
t(C8’-C7’-N-C7) t(Ir-C8-C7-C1)
bond angles
at N
K4b/1[a]
K4b/2[a]
K4d
2.413(8)
2.452(9)
2.226 (8)
2.228(8)
2.168(8)
2.173(8)
1.378(12) 1.406(11) 124.4
1.385(13) 1.431(13) 122.5
74.5
74.6
73.1
73.7
74.3
75.5
358.4
358.4
354.3
356.1
355.9
356.6
163.8
166.5
153.3
153.6
161.4
145.6
ꢀ166.4
ꢀ163.3
ꢀ169.7
ꢀ169.6
ꢀ161.9
ꢀ164.5
2.327(12) 2.180(11) 2.201(13) 1.362(16) 1.414(17) 124.8
K4e/1[a,b] 2.32(2)
2.172(16) 2.21(2)
1.40(3)
1.39(3)
1.404(7)
1.44(3)
1.36(3)
1.409(6)
121.8
123.7
121.9
K4e/2[a,b] 2.377(15) 2.21(2)
K4 f 2.355(3) 2.216(3)
2.20(2)
2.189(3)
[a] The designations “/1” and “/2” refer to independent molecules of the unit cell. [b] The crystal structure analysis was carried out with ent-K4e.
Chem. Eur. J. 2010, 16, 6601 – 6615
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6605

G. Helmchen et al.
Figure 7. Conformations of the C,P chelate ligand of the complexes K4
and K4’ (Ar=Ph or o-(MeO)C6H4). K4’ is defined in Figure 9.
Figure 6. X-ray crystal structure of the cation of complex K4b (top: ball
and stick model, bottom: CPK model). For a colour figure see the Sup-
porting Information.
Figure 8. Superposition of the X-ray structures of all (p-allyl)Ir com-
plexes with best fit of the coordination plane [P,Ir,C8]. For a colour
figure see the Supporting Information.
to the substituted carbon C1A is longer than the bond to the
unsubstituted C3A, likely due to the strong trans influence of
P. The difference amounts to 0.15 and 0.26 ꢂ for R=CH3
and Ph, respectively. Thus, the preferred addition of a nucle-
ophile at the higher substituted allylic carbon might be due
to preferred breaking of the weaker bond. Bonding within
the allylic moiety is almost symmetric as best seen for the
crotyl ligand of K4 f. Even for the cinnamyl ligand, the bond
mational data with respect to these two units are similar.
Considerable conformational flexibility with respect to the
ꢀ
bond C7’ N is apparent from the fact that the torsion angles
t(C8’-C7’-N-C7) vary over a range of 218, and even for the
two structures of K4e in the same crystal there is a differ-
ence of 88.
ꢀ
ꢀ
lengths C1A C2A and C2A C3A differ by maximally 0.04 ꢂ.
A similar observation was made previously for a (p-allyl)-
ACHTUNGTRENNUNG
(PHOX)Ir complex (PHOX=phoshpino-oxazoline).[15]
The five-membered ring adopts an envelope conformation
DFT calculations: p-allyl complexes K4’
with an almost planar nitrogen unit and the aryl group a in
a pseudoequatorial disposition. The near planarity of the
bonds around N is apparent from the value of 356ꢁ28 for
the sum of the three bond angles. The bond angle P-Ir-C8 of
74ꢁ28 is very small because of constraints set by the chelate
structure.
The five-membered ring and the 2-arylethyl unit are of
further interest since they could exhibit conformational flex-
ibility (cf. Figure 7). A superposition of the corresponding
partial structures of all the available X-ray structures con-
taining these units is presented in Figure 8, and geometric
data are displayed in Table 3. It is apparent that the confor-
DFT calculations furnished an interesting insight into the
conformational and energetic aspects of the allyl complexes
and their reactions. The cyclometalated complexes K4’ with
L*=L3 were primarily used as models; that is, the bi-
naphthyl group was replaced by a biphenyl group and cod
by cot (1,3,5,7-cyclooctatetraene) (Figure 9). The latter
change was induced by the observation that cod complexes
display a twisted, chiral cod moiety requiring the number of
models to be doubled in order to account for the helicity of
cod. The simplification appears justified because catalysts
derived from 2,2’-dihydroxybiphenyl[9,19] or dbcot[4] and cata-
6606
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 6601 – 6615

Asymmetric Catalysis
FULL PAPER
our preliminary calculations the partial structures B/D
(Figure 7) were used, while in the crystal structures A/C
were found.
Complexes K4’, R1 =R2 =H: First, a conformational curve
concerning the phenethyl substituent within the D-P-endo
diastereomer was calculated (Figure 10). It exhibits two
minima. The energetically lower one corresponds to the con-
former C.
Figure 10. The torsional energy curve of the phenethyl substituent within
conformer A of the D-P-endo complex K4’ (R1 =R2 =H); for definition
of the torsion angle see Figure 7. Energies refer to gas-phase geometry
optimisations; zero-point or THF single-point corrections were not car-
ried out. The arrow denotes the torsional state found in the crystal struc-
ture (C of Figure 7).
Figure 9. Schematic representations of the configurations of the isomeric
(p-allyl)Ir complexes of type K4 and K4’. Formulas in the upper line
define the constitution used for DFT calculations; the structure corre-
sponding to the X-ray crystal structure is represented in the box with
broken lines. ···=cot; P and C series: one of the allylic carbon atoms is
located trans to P and the CH2 group of the C,P ligand, respectively; exo
ꢀ
and endo: the C2A H bond of the allyl ligand is pointing towards cot and
away from cot, respectively.
Using conformational orientations A and C derived from
the crystal structures as starting models, calculations on the
eight (p-allyl)Ir complexes (K4’, R1 =R2 =H) were repeated.
The results are presented in Table 4 (entry 1). The confor-
mer corresponding to the crystallographically observed one
was found to be significantly more stable than the one used
in our previous studies (Table 4, entry 6). An exception is
the L-P-endo diastereomer, for which the pseudo-axial ori-
entation B was favoured.
The P isomers were found to be more stable than C iso-
mers in the series of the eight (p-allyl)Ir complexes
(Table 4). The reason is the lower trans influence of P com-
pared to C. Two additional trends are apparent: the endo
isomers are favoured over exo isomers, and there is a clear
preference for the D- over the L-isomers in all cases.
lysts prepared from L1 or L2 and cod produce similar re-
sults. Ammonia was used as the probe nucleophile. Recent-
ly, successful monosubstitution reactions with ammonia as
nucleophile have been realised.[20]
The allyl complexes are schematically represented in
Figure 9. They are classified into two series, which are desig-
nated P- and C-isomers, according to whether one of the al-
lylic termini is located in trans-disposition to P or C of the
P,C-chelate ligand, respectively.
For our communication,[8] the crystal structure of [Ir
(C,P-L1)PMe3] (conformer B/D, cf. Figure 7), as reported by
Hartwig et al,[3,21] was used as the starting point for the ge-
ometry of the fragment {Ir(C,P-
ACHTUNGTRNE(NUNG cod)-
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
L3)}. The L-P-endo isomer was
found to be the most stable dia-
stereomer in the series of the
parent complexes, R1 =R2 =H.
However, all the subsequently
crystallised allylic complexes
displayed the D-P-endo configu-
ration with A/C as the pre-
ferred conformation. To clear
this discrepancy, the influence
of the conformational degrees
of freedom discussed above was
Table 4. DFT energies [kcalmolꢀ1] of allylic complexes K4’.[a]
K4’ D-P-endo D-P-exo L-P-endo L-P-exo D-C-endo D-C-exo L-C-endo L-C-exo
conformers A and C (cf. Figure 7)
1
2
3
4
5
R1 =R2 =H
0.0
0.0
3.0
0.0
8.3
2.2
3.2
4.7
2.5
3.7
3.9
6.6
4.7
4.6
7.0
4.9
4.1
6.3
5.4
3.9
6.7
6.2
7.2
9.1
7.4
7.0
9.1
R1 =CH3, R2 =H
R1 =H, R2 =CH3
R1 =Ph, R2 =H
R1 =H, R2 =Ph
conformers B and D (cf. Figure 7)
6
7
8
R1 =R2 =H
5.6
4.9
4.4
4.6
7.2
3.2
3.4
8.1
6.6
7.7
9.3
8.8
7.4
12.9
8.0
6.8
9.8
11.2
9.4
12.3
9.1
8.1
10.6
R1 =CH3, R2 =H
R1 =H, R2 =CH3 10.1
[a] Energies were obtained from gas-phase geometry optimisations with subsequent gas phase zero-point and
investigated computationally. In THF single-point (scrf-cpcm) corrections.
Chem. Eur. J. 2010, 16, 6601 – 6615
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6607

G. Helmchen et al.
Bonding of the allylic moiety to Ir is non-symmetric
lated to be significantly longer in the D-P-exo than in D-P-
endo isomer (2.79 vs. 2.63 ꢂ), which seems to result from
the repulsive interaction between the phenyl group of the
cinnamyl moiety and cot. This is similarly found for the cor-
responding crotyl complexes, R1 =Me, albeit to a lesser
extent (2.46 vs. 2.55 ꢂ), whereas for the allyl complexes,
ꢀ
(Table 5). In case of the unsubstituted p-allyl ligand, the C
ꢀ
ꢀ
Ir bond trans to P or C (C1A Ir) is longer than the C Ir
ꢀ
bond trans to a double bond of the cot ligand (C3A Ir) for
all isomers. The geometric distortion is more pronounced in
the C series, as should be expected considering the trans-in-
fluences of C and P.
1
2
ꢀ
R =R =H, the distance Ir C1A is almost identical for the
D-P-exo and the D-P-endo isomer (2.31 vs. 2.32 ꢂ).
The D-P-endo isomer with R1 =H, R2 =Ph was
8.3 kcalmolꢀ1 less stable than the D-P-endo isomer with the
phenyl group at C1A. This is likely due to repulsive interac-
tions between the phenyl group and the C3A substituent and
Ph(b).
Complexes K4’, R1 or R2 =CH3: The 16 model structures
were generated from the eight diastereomeric p-allyl com-
plexes by replacing terminal syn-hydrogen atoms at C1A or
C3A of the p-allyl ligand by a methyl substituent (cf.
Figure 9). The results of the DFT calculations are presented
in Tables 4 and 5. The 1-methylallyl isomers were found to
be more stable than the 3-methylallyl isomers by 1.5–
3.0 kcalmolꢀ1.
DFT calculations: transition states and olefin complexes de-
rived from complexes K4’
Concerning bond lengths of the crotyl ligand to Ir, in
most cases the bond of the methyl-substituted carbon to Ir
is longer than that of the unsubstituted terminal carbon (cf.
Table 5, entries 9–27).
Amination transition states were computed for reactions of
the thermodynamically favoured isomers D-P-endo-K4’ con-
taining the allyl, the crotyl or the cinnamyl ligand. NH3 was
used as the probe nucleophile.
Complexes K4’, R1 or R2 =Ph: Only three of the isomers
were investigated (Tables 4 and 5). The D-P-exo isomer,
R1 =Ph, R2 =H, was found to be disfavoured by
2.5 kcalmolꢀ1 relative to the D-P-endo isomer; this is consis-
tent with the results obtained for the corresponding allyl
complexes K4’ with R1 =R2 =H (2.2 kcalmolꢀ1) and crotyl
complexes R1 =Me, R2 =H (3.2 kcalmolꢀ1). Interestingly,
Reactions at the complex D-P-endo-K4’, R1 =R2 =H: Transi-
tion states (TS) and products were located for the addition
of NH3 at C1A, C2A and C3A in the case of the unsubstituted
allyl complex (Figure 11, R1 =H). The transition-state
energy of the reaction at C1A was calculated to be
2.8 kcalmolꢀ1 lower than that of the reaction at C3A. In ad-
ꢀ
for the cinnamyl complexes the distance C1A Ir was calcu-
Table 5. Selected bond lengths [ꢂ] and bond angles [8] of the (p-allyl)Ir complexes K4’.[a]
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
K4’
Configuration
Ir C1A
Ir C2A
Ir C3A
C1A C2A
C2A C3A
C1A-C2A-C3A
P-Ir-CH2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
R1 =R2 =H
D-P-endo
D-P-exo
2.31
2.32
2.32
2.35
2.46
2.50
2.42
2.43
2.46
2.55
2.45
2.59
2.73
2.72
2.68
2.63
2.25
2.28
2.24
2.31
2.36
2.41
2.32
2.40
2.63
2.79
2.18
2.25
2.26
2.25
2.28
2.30
2.32
2.30
2.31
2.28
2.32
2.28
2.33
2.37
2.38
2.35
2.36
2.26
2.27
2.28
2.29
2.30
2.31
2.30
2.30
2.31
2.38
2.35
2.24
2.24
2.24
2.22
2.20
2.18
2.23
2.22
2.21
2.19
2.20
2.16
2.14
2.15
2.16
2.16
2.36
2.35
2.41
2.31
2.30
2.25
2.35
2.26
2.19
2.16
2.74
1.41
1.41
1.41
1.40
1.39
1.39
1.40
1.39
1.40
1.39
1.40
1.39
1.38
1.38
1.38
1.38
1.42
1.41
1.43
1.41
1.41
1.40
1.42
1.40
1.40
1.39
1.44
1.42
1.42
1.42
1.42
1.43
1.44
1.43
1.44
1.43
1.44
1.43
1.44
1.45
1.46
1.45
1.45
1.41
1.41
1.40
1.41
1.42
1.43
1.41
1.44
1.43
1.45
1.39
121.5
121.4
121.6
121.5
123.2
122.4
122.9
123.2
123.0
123.5
123.2
123.2
124.9
124.6
124.0
124.1
122.0
122.9
122.6
123.6
123.2
122.9
122.5
124.3
124.5
124.3
123.2
75.7
78.1
73.9
75.9
73.5
75.3
74.1
75.1
75.6
77.1
73.5
75.6
74.5
75.8
74.4
74.7
75.7
77.0
73.8
74.9
73.6
74.6
75.3
75.4
75.5
76.7
76.6
R1 =R2 =H
R1 =R2 =H
L-P-endo
L-P-exo
D-C-endo
D-C-exo
L-C-endo
L-C-exo
D-P-endo
D-P-exo
L-P-endo
L-P-exo
D-C-endo
D-C-exo
L-C-endo
L-C-exo
D-P-endo
D-P-exo
L-P-endo
L-P-exo
D-C-endo
D-C-exo
L-C-endo
L-C-exo
D-P-endo
D-P-exo
R1 =R2 =H
R1 =R2 =H
R1 =R2 =H
R1 =R2 =H
R1 =R2 =H
R1 =CH3, R2 =H
R1 =CH3, R2 =H
R1 =CH3, R2 =H
R1 =CH3, R2 =H
R1 =CH3, R2 =H
R1 =CH3, R2 =H
R1 =CH3, R2 =H
R1 =CH3, R2 =H
R1 =H, R2 =CH3
R1 =H, R2 =CH3
R1 =H, R2 =CH3
R1 =H, R2 =CH3
R1 =H, R2 =CH3
R1 =H, R2 =CH3
R1 =H, R2 =CH3
R1 =H, R2 =CH3
R1 =Ph, R2 =H
R1 =Ph, R2 =H
R1 =H, R2 =Ph
D-P-endo
[a] Calculations were carried out using conformational orientations A/C according to Figure 7.
6608
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 6601 – 6615

Asymmetric Catalysis
FULL PAPER
Figure 11. Energy profiles of reactions of complexes D-P-endo-K4’, R2 =
H, with NH3. Energies were obtained from gas-phase geometry optimisa-
tions and subsequent gas-phase zero-point and THF single-point (scrf-
cpcm) corrections. TSC1 and TSC3 =transition states of amination at C1A
and C3A, respectively; OCC1 and OCC3 =corresponding olefin complexes.
For a colour figure see the Supporting Information.
dition, the energy of the resultant olefin complex OCC1 of
the former reaction is slightly lower.
Reactions at C2A to give iridacyclobutanes have been ob-
served for Ir complexes by others and by us.[15,22] Interesting-
ly, the activation barrier of the reaction at C2A was found to
be lower than that of the reaction at C1A (6.0 vs.
7.4 kcalmolꢀ1). However, the resulting iridacyclobutane was
found to be energetically unfavorable in comparison with
the olefin complexes from the aminations at C1A and C3A.
A rapid, reversible background reaction at C2A, therefore,
seems possible. With R1 =Me and R1 =Ph it was not possi-
ble to locate a transition state or a product of amination at
C2A (see the Supporting Information).
Figure 12. Graphical representation of the reaction of the complex D-P-
endo-K4’, R1 =CH3, R2 =H, with NH3 at C1A. For a colour figure see the
Supporting Information.
Reactions at complexes D-P-endo-K4’, R1 =CH3 or Ph, R2 =
H: These calculations also revealed a distinct preference for
the addition of NH3 at C1A, yielding the branched product,
over the addition at C3A, yielding the linear product
(Figure 11). Of interest is also the relative stability of the
product olefin complexes. In the case R1 =R2 =H product
energies are almost identical, in the case R1 =CH3, R2 =H,
the branched product is preferred over the linear one by
3.8 kcalmolꢀ1; for the cinnamyl derivative, R1 =Ph, R2 =H,
the Ir complex of the linear product is more stable than that
of the branched one by 1.9 kcalmolꢀ1. Thus, there is no cor-
relation between the energies of transition states and the
product olefin complexes. In other words, according to these
calculations, the regioselectivity of the allylic substitution is
not due to product development control.
the allylic moiety to iridium, the rotations are clockwise and
counter-clockwise for the amination at C1A and C3A, respec-
tively. A further graphical representation of these move-
ments is given in the Supporting Information.
Possible origin of the minor enantiomer: Possible reaction
channels for the formation of the minor enantiomer of the
allylic amination with NH3 were explored for the two p-allyl
complexes with D-P configuration nearest in energy to the
favoured complex D-P-endo-K4’.
The amination at C3A of the complexes D-P-endo-K4’,
R1 =H, R2 =Me or Ph, appeared generally unlikely as com-
peting pathways, because of steric crowding in the vicinity
of Ph(b) (Figure 6). However, in the case of the crotyl com-
plexes, the relative stabilities of the isomers D-P-endo-K4’,
R1 =H, R2 =Me and D-P-exo-K4’, R1 =Me, R2 =H, are
almost identical, 3.0 and 3.2 kcalmolꢀ1, respectively (see
Table 4), and the calculated activation barriers are similar,
6.2 and 7.6 kcalmolꢀ1 for the reactions of the D-P-exo
isomer at C1A and the D-P-endo at C3A, respectively
(Figure 13).
A description of the geometric change of the system
along the reaction path is given by Figure 12 and Table 6.
The allylic moiety undergoes a geometric change with only
small displacements of the atoms C2A and C3A relative to
the fragment {IrACHTUNGTRENNUNG(C,P-L3)AHCTUNGTRENN(UGN cot)}. This movement is best de-
scribed by a superposition of two rotations. Most pro-
ꢀ
nounced is an out-of-plane rotation about the axis C2A C3A,
as apparent from the torsion angle t(Ir-C3A-C2A-C1A),
which varies by 50–558. A second component of the allylic
motion is clearly seen from the dihedral angle t(C8-Ir-C2A-
H2A) describing an in-plane rotation, which varies over a
range of about 158 (Table 6). Viewed in the direction from
After this somewhat unexpected result the cinnamyl de-
rivative D-P-endo-K4’, R1 =H, R2 =Ph, was more closely ex-
amined, considering that the steric crowding might be re-
Chem. Eur. J. 2010, 16, 6601 – 6615
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6609

G. Helmchen et al.
Table 6. Selected bond lengths [ꢂ] and torsion angles [8] for the amination of D-P-endo-K4’, R2 =H.[a] TSC1 and TSC3 =transition states of amination at
C1A and C3A, respectively; OCC1 and OCC3 =corresponding olefin complexes.
R1
N Creact Ir C1A Ir C2A Ir C3A C8-Ir-C2A-C3A C8-Ir-C2A-C1A C8-Ir-C2A-H2A Ir-C3A-C2A-C1A Ir-C1A-C2A-C3A
ꢀ
ꢀ
ꢀ
ꢀ
1
2
3
4
5
H
H
H
H
H
allyl complex
TSC1
OCC1
TSC3
OCC3
1
2.31
3.03
3.22
2.18
2.18
2.25
2.27
2.20
2.24
2.18
2.24
2.15
2.17
2.87
3.18
123.9
111.1
106.0
ꢀ104.3
10.8
ꢀ1.6
ꢀ6.6
18.5
24.7
ꢀ57.3
ꢀ96.0
56.0
1.93
1.58
2.01
1.56
ꢀ113.8
ꢀ95.4
ꢀ88.5
88.0
112.8
6
7
8
9
Me allyl complex
Me TSC1
Me OCC1
1
2.46
3.13
3.28
2.24
2.22
2.28
2.27
2.20
2.25
2.19
2.21
2.15
2.16
2.86
3.16
120.5
110.4
104.7
ꢀ108.1
8.1
ꢀ1.8
ꢀ7.2
14.6
20.8
ꢀ61.8
ꢀ102.8
ꢀ117.7
56.5
1.96
1.59
1.97
1.56
Me TSC3
ꢀ99.2
ꢀ91.7
88.8
111.0
10 Me OCC3
11 Ph allyl complex
12 Ph TSC1
13 Ph OCC1
14 Ph TSC3
1
2.63
3.18
3.27
2.29
2.25
2.31
2.25
2.20
2.24
2.18
2.19
2.16
2.16
2.85
3.18
121.0
109.7
105.6
ꢀ108.8
10.3
ꢀ1.4
ꢀ5.1
18.8
23.6
ꢀ67.6
ꢀ107.5
ꢀ117.2
58.2
1.85
1.62
2.01
1.56
ꢀ96.6
ꢀ90.3
90.4
114.2
15 Ph OCC3
[a] The model system was generated from the X-ray crystal structure of K4 f by replacing ligands L2 by L3 and cod by cot.
DFT calculations: unsimplified allyl complexes K4 and their
reactions
With several crystal structures of cod complexes at hand,
there was an incentive to carry out calculations on the non-
simplified complexes D-P-endo-K4, R1 =H, Me or Ph, R2 =
H.
Complexes K4 with a D-P-endo configuration: Representa-
tive bond lengths and angles are listed in Table 7. The bond
Figure 13. Comparison of the energies along the favoured reaction path-
way (left) and two reactions yielding the minor enantiomer (middle and
right); R1 or R2 =Me (black) and Ph (grey).
ꢀ
length Ir C1A is overestimated computationally by 0.11 ꢂ
with R=Me and by 0.21 ꢂ with R=Ph. On the other hand,
the difference between the calculated systems K4 and K4’ is
ꢀ
lieved by conformational adaptation of the fragment {Ir-
negligible in both cases. The distance Ir C2A is overestimat-
ed computationally by 0.09 and 0.08 ꢂ for K4 and K4’, re-
(C,P-L3)}. The energies of all conformers according to
Figure 7 were calculated and they exceeded 8.3 kcalmolꢀ1.
Therefore, the reaction via D-P-endo-K4’, R1 =H, R2 =Ph,
was not considered viable, and the transition state of its ami-
nation was not assessed. Hence, in case of the cinnamyl
system the minor enantiomer is likely formed from the D-P-
exo isomer. The calculated activation barrier of
7.1 kcalmolꢀ1 is comparable to
spectively. The bond length Ir C3A is well reproduced.
ꢀ
Hence B3LYP appears to overestimate the distortion of al-
lylic complexes towards the h1-coordination mode. Bond
angles C1A-C2A-C3A and P-Ir-C8 are well reproduced by the
calculations, whereas for dihedral angles deviations are
large. Given that the torsional potentials are generally soft,
the 7.8 kcalmolꢀ1 calculated for
Table 7. Comparison of selected bond lengths [ꢂ] and angles [8] from X-ray crystal structures and DFT calcu-
lations (D-P-endo configuration).[a]
the reaction of the correspond-
ing D-P-endo isomer at C1A.
We conclude that minor
R1 =Me, R2 =H
R1 =Ph, R2 =H
X-ray (K4d) DFT (K4d)[b] DFT (K4’)[b] X-ray (K4b) DFT (K4b)[c] DFT (K4’)[c]
enantiomers
are
possibly
ꢀ
Ir C1A
2.33
2.18
2.20
1.36
1.42
124.8
73.1
153.3
2.44
2.27
2.21
1.41
1.43
2.46
2.28
2.21
1.40
1.43
2.43
2.23
2.17
1.38
1.42
2.64
2.31
2.19
1.40
1.43
2.63
2.31
2.18
1.40
1.44
formed by addition of the nu-
cleophiles at C1A of the D-P-
exo complexes. However, com-
putational assessment of enan-
tioselectivities is not possible at
the level of sophistication used
here.
ꢀ
Ir C2A
ꢀ
Ir C3A
ꢀ
C1A C2A
ꢀ
C2A C3A
C1A-C2A-C3A
P-Ir-C8
t(C8’-C7’-N-C7)
t(Ir-C8-C7-C1)
122.8
123.0
123.5
123.5
124.5
74.2
154.3
75.6
160.4
74.6
165.2
73.6
153.4
75.5
160.9
ꢀ169.7
ꢀ165.4
ꢀ166.1
ꢀ164.9
ꢀ168.2
ꢀ165.8
[a] For the crystal structures, averaged values from the non-equivalent complexes are listed. [b] The starting
geometry was derived from the crystal structure of the complex K4d. [c] The starting geometry was derived
from the crystal structure of the complex K4b.
6610
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 6601 – 6615

Asymmetric Catalysis
FULL PAPER
crystal packing effects could exert a considerable influence
here.
most stable conformer of [K3’Hb]+ (0.4 kcalmolꢀ1) are de-
picted in Figure 16. Note that for once the conformers with
a pseudoaxial a phenyl are preferred. The agreement be-
tween the calculated relative stabilities of [K3’Ha]+ and
[K3’Hb]+ and the experimentally observed ratio of the cat-
ionic hydrides [K3H]+ of 3:1 (cf Figure 3) is fortuitous.
Reactions at complexes D-P-endo-K4: Results are presented
in Figure 14. For all allyl complexes the activation barriers
were calculated to be higher by about 1–2 kcalmolꢀ1 for the
cod than the cot complexes. The complexes of the branched
products (OCC1) were found to be less stable than those of
the linear products (OCC3). This is a further indication that
the regioselectivity of the Ir-catalysed allylic substitution is
not controlled by the stability of the product olefin complex.
Figure 14. Energy profiles of the reactions of the complexes D-P-endo-K4
with NH3. Energies were obtained from gas-phase geometry optimisa-
tions and subsequent gas phase zero-point and THF single-point (scrf-
cpcm) corrections. For a colour figure see the Supporting Information.
Figure 16. Partial structures of the energetically low conformers of the
(hydrido)Ir complexes [K3’Ha]+ (both 0.0 kcalmolꢀ1
)
and [K3’Hb]+
(0.4 kcalmolꢀ1) obtained by DFT calculations. For a colour figure see the
Supporting Information.
DFT calculations: hydride complexes [K3’H]+
Conclusion
It was desirable to shed light on the structures of the cation-
ic (hydrido)Ir complexes [K3H]+ (cf. Scheme 3). Consider-
ing only square pyramidal configurations, there are ten iso-
mers possible (see Figure 15).
In this article a number of mechanistic aspects of allylic sub-
stitutions with iridium catalysts derived from phosphorami-
dites by C H activation were investigated. The following
ꢀ
main results were obtained:
1) Determination of the resting states by 31P NMR spectros-
ꢀ
copy revealed that the C H activation process is reversi-
ble upon use of in situ prepared catalysts.
2) A novel one-pot procedure for the preparation of (p-al-
lyl)Ir complexes from simple starting materials was de-
veloped.
3) The new complexes were characterised by X-ray crystal
structure analyses and spectral data.
4) The (p-allyl)Ir complexes were found to be fully active
catalysts of the allylic substitution reaction and resting
states under conditions, which excluded reversion of the
Figure 15. Ten configurationally isomeric (hydrido)Ir complexes [K3’H]+.
The chelate ligands are as defined in Figure 9.
In addition, there is conformational variability with re-
spect to the iridacycle and the phenethyl group analogous to
that described in Figure 7. To assess the viability of these hy-
dride complexes, the topographically distinct model com-
plexes [K3’H]+ derived from cot and ligand L3 were investi-
gated, which should be particularly stable according to con-
sideration of the trans influences of the ligands. These iso-
mers were generated by protonation of iridium at the two
diastereotopic half-spaces of K3’.
ꢀ
C H activation.
5) Finally, DFT calculations on the allyl complexes, transi-
tion states of the allylic substitution and product olefin
complexes were carried out on both slightly simplified
and unsimplified complexes.
Upon consideration of the conformational parameters
mentioned above, five conformers of [K3’Ha]+ and four
conformers of [K3’Hb]+ were generated. The two most
stable conformers of [K3’Ha]+ (both 0.0 kcalmolꢀ1) and the
Experimental Section
Computational methods: All computations were performed using the
hybrid Becke functional (B3)[23] for electron exchange and the correlation
Chem. Eur. J. 2010, 16, 6601 – 6615
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6611

G. Helmchen et al.
functional of Lee, Yang and Parr (LYP),[24] as implemented in the
GAUSSIAN 03 software package.[25] For iridium, the SDD basis set with
the associated effective core potential was employed.[26] All other atoms
were modelled at the 6-31G(d) level of theory.[27] Geometry optimisations
were performed without any symmetry constraints (C1 symmetry).
ed at RT on the following specACTHNGUETRNNUtG rometers: Bruker DRX-250
(101.26 MHz), Bruker AC-300 (121.50 MHz), Bruker Avance 500
(202.46 MHz). The spectra were recorded in CDCl3 or [D8]THF and
chemical shifts are reported in d units relative to TMS in 1H NMR spec-
tra[32] or phosphoric acid or triphenylphosphine as external standard. The
assignments of signals for compounds K4a–f were confirmed by H,H-
COSY, H,C-COSY, TOCSY, HMBC, NOESY and DEPT spectra. Atom
numbers are specified at the formulas. 1H NMR spectra of all new com-
pounds are presented in the Supporting Information. High-resolution
mass spectra were recorded on a Bruker ApexQe FT-ICR (ESI+) mass
spectrometer. X-ray data were collected on a Bruker APEX diffractome-
ter equipped with a MoKa radiation source (l=0.71073 ꢂ) and a graphite
monochromator. Intensities were corrected for Lorentz and polarisation
effects; an empirical absorption correction was applied using
SADABS.[33] All structures were solved by direct methods and refined
against F2 using the SHELXTL software package.[34] The corresponding
crystallographic data are displayed in Table 8. CCDC-756243 (K4b),
-756244 (K4d), -738902 (K4e) and -756245 (K4 f) contain the supplemen-
tary 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. Flash-column chromatography was
carried out with silica gel (0.032–0.062) of Macherey, Nagel and Co. Tet-
rahydrofuran was dried over benzophenone ketyl, and the water content
Starting geometries for the transition state search were located by per-
forming restricted geometry optimisations. The distance between the re-
acting carbon terminus and the nitrogen nucleophile was fixed at 2.0 ꢂ
for the nucleophilic attack at C1A or C3A and 1.9 ꢂ for reactions at C2A,
respectively. The thus obtained geometries were subsequently fully opti-
mised as saddle points of first-order, employing the Berny algorythm.[28]
Frequency calculations were undertaken to confirm the nature of the sta-
tionary points, yielding zero imaginary frequencies for all allyl and olefin
complexes. The amination transition states possessed one imaginary fre-
ꢀ
quency, which represented the vector of the C N bond formation in all
cases. Zero-point energy corrections were carried out for all computed
energies. The solvent influence was additionally taken into account by
conducting single-point polarised continuum (CPCM-SCRF) calcula-
tions,[29] with THF (e=7.6) as the solvent.[30]
General: Unless stated otherwise, reactions were carried out under argon
1
using anhydrous conditions. H NMR spectra were recorded at RT on the
following spectrometers: Bruker AC-300 (300 MHz), Bruker Avance 500
(500 MHz). The spectra were recorded in CDCl3 or [D8]THF and chemi-
cal shifts are reported in d units relative to TMS or to the residual undeu-
terated solvent (CHCl3 in CDCl3 at dH =7.26 ppm).[31] The following ab-
breviations are used to indicate the signal multiplicity: s (singlet), d (dou-
blet), t (triplet), q (quartet), m (multiplet), dd (doublet of doublets), dt
(doublet of triplets), ddd (doublet of doublets of doublets), brs (broad
signal). 13C NMR spectra were recorded at RT on the following spec-
was determined by Karl Fischer titration. 1,5,7-TriazabicycloACTHNUTRGNEUGN[4.4.0]dec-5-
ene (TBD) was stored in a desiccator over KOH. IUPAC names were
generated by the ACD/Labs 6.0 programm from Advanced Chemistry
Development Inc.
Syntheses, spectroscopic data and determinations of enantiomeric excess-
es of the following compounds were already reported: 1c,[35] 2a, Nu=
CH
NPhth,[5] 2c, Nu=CH
AHCTUNGTRENNUNG
(CO2CH3)2,[9] a Krapcho reaction was carried out (NaCl (1.2 equiv), H2O
(CO2CH3)2,[36] 2a, Nu=N(Boc)2,[5] 2a, Nu=NPhth,[5] 2b, Nu=
ACHUTGTNRENNUG CAHTUNGTRENNUGN
ACHTUNGTRENNUNGtrometers: Bruker AC-300 (75.48 MHz), Bruker Avance 500
AHCTUNGTRENNUNG
(CO2CH3)2.[36] In the case of 2b, Nu=CH-
(125.76 MHz). The spectra were recorded in CDCl3 or [D8]THF and
chemical shifts are reported in d units relative to the residual undeuterat-
ed solvent CHCl3 (dC =77.16 {central line of the triplet})[31] The following
abbreviations were used: s (singlet, quarternary C-atom), d (doublet,
CH-group), t (triplet, CH2-group), q (quartet, CH3-group). The multiplic-
ity stated refers to {1H}-decoupled spectra. 31P NMR spectra were record-
(6 equiv), DMSO, 1608C, 14 h) and the product was analysed by GC:
Hewlett Packard Gas Chromatograph 5890A, column Varian CP-Chirasil-
Dex (b-CD), 25 m
ꢃ
0.25 mmꢃ0.25 m, isothermal, 708C, 24.7 min,
26.1 min.
Table 8. Data of the crystal structure analysis of the (p-allyl)Ir complexes.
K4b
K4d
K4e
C51H52F3IrNO7PS·3C6H6
K4 f
formula
Mr
C55H54F6IrNO4PSb·1= C6H6
1290.97
C50H52F6IrNO4PSb·3C6H6
1424.17
C49H48F3IrNO5PS·2C6H6
1199.33
2
1337.49
crystal size [mm3]
crystal system
space group
a [ꢂ]
b [ꢂ]
c [ꢂ]
0.14ꢃ0.12ꢃ0.10
monoclinic
P21
13.6182(8)
20.5836(12)
18.6485(11)
90
93.604(1)
90
5217.1(5)
4
0.30ꢃ0.17ꢃ0.05
orthorhombic
P212121
12.184(2)
17.401(3)
29.009(5)
90
0.31ꢃ0.26ꢃ0.12
triclinic
P1
0.27ꢃ0.24ꢃ0.16
monoclinic
P21
11.6538(15)
20.556(3)
12.4857(16)
90
116.433(2)
90
2678.3(6)
2
12.4002(16)
15.0182(19)
17.062(2)
89.086(3)
89.860(3)
76.275(3)
3086.4(7)
2
a [8]
b [8]
90
90
g [8]
V [ꢂ3]
Z
6150.7(18)
4
1.54
1calcd [gcmꢀ3
]
1.64
3.17
1.44
2.29
1.49
2.62
m [mmꢀ1
]
2.69
max/min tranmission
index ranges
0.74/0.67
ꢀ18ꢂhꢂ18
ꢀ27ꢂkꢂ27
ꢀ24ꢂlꢂ24
1.9–28.4
200(2)
0.88/0.50
ꢀ15ꢂhꢂ15
ꢀ21ꢂkꢂ21
ꢀ36ꢂlꢂ36
1.4–26.4
0.77/0.54
ꢀ15ꢂhꢂ15
ꢀ18ꢂkꢂ18
0ꢂlꢂ21
1.9–26.4
200(2)
0.68/0.54
ꢀ15ꢂhꢂ15
ꢀ27ꢂkꢂ27
ꢀ16ꢂlꢂ16
2.0–28.3
q range [8]
T [K]
200(2)
200(2)
reflns collected
independent reflns [Rint
observed reflns [I>2s(I)]
data/restraints/parameters
GOF on F2
55479
25676 [0.0505]
22210
25676/13/1285
1.09
55907
12566 [0.0604]
10780
12566/435/767
1.16
12548
12552 [0.0000]
11621
12552/210/1426
1.13
28505
13025 [0.0204]
12519
13025/25/675
1.03
]
final R1/wR2 indices [I>2s(I)]
absolute structure parameter
0.057/0.113
0.020(4)
2.35/ꢀ1.80
0.077/0.158
0.019(10)
5.43/ꢀ1.32
0.046/0.109
0.091(10)
1.52/ꢀ0.72
0.024/0.056
ꢀ0.014(3)
1.07/ꢀ0.41
max/min residual [eꢂꢀ3
]
6612
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 6601 – 6615

Asymmetric Catalysis
FULL PAPER
General procedure for the preparation of (p-allyl)Ir complexes (GP1): In
a flame-dried Schlenk tube under an argon atmosphere, complex K1 was
prepared by stirring a solution of [{IrACHTUNRGTNEUNG(cod)Cl}2] (6.7 mg, 10 mmol) and L*
9.0 Hz, 1H; BINOL), 7.81–7.86 (m, 3H; BINOL, 2ꢃPhAllyl), 7.53–7.58
(m, 3H; Ph, BINOL), 7.43–7.49 (m, 4H; Ph, 3ꢃPhAllyl), 7.34–7.40 (m,
4H; 4ꢃBINOL), 7.25–7.33 (m, 2H; 2ꢃPh), 7.09–7.13 (m, 2H; 2ꢃPh),
6.98 (dd, J=7.3, 7.3 Hz, 1H; Ph), 6.95 (d, J=7.9 Hz, 1H; Ph), 5.74 (d,
J=12.4 Hz, 1H; 1A-H), 5.26–5.31 (m, 1H; cod-CH), 5.05 (ddd, J=11.6,
11.6, 9.7 Hz, 1H; 2A-H), 4.73 (dd, J=11.2, 4.5 Hz, 1H; 7-H), 4.51–4.57
(m, 1H; 7’-H), 4.12 (s, 3H; OCH3), 4.04–4.09 (m, 1H; cod-CH), 3.57–
3.63 (m, 4H; cod-CH, OCH3), 2.77–2.93 (m, 3H; 3A,syn-H, cod-CH2),
2.45–2.53 (m, 1H; cod-CH2), 2.12–2.35 (m, 3H; 8b-H, 2ꢃcod-CH2), 2.09
(d, J=11.3 Hz, 1H; 3A,anti-H), 1.79–1.86 (m, 1H; cod-CH2), 1.57–1.64 (m,
1H; cod-CH2), 1.44–1.52 (m, 1H; cod-CH2), 0.97 (dd, J=12.2, 12.2 Hz,
(20 mmol) in dry THF (0.5 mL, <30 mg H2OmLꢀ1 THF, Karl–Fischer ti-
tration) at room temperature for the stated time. The allylic carbonate
(30 or 40 mmol) and AgX (20 mmol) were added to the orange solution
causing the formation of a white precipitate. The mixture was stirred for
20 min and kept at room temperature for the stated time. The white pre-
cipitate was removed by centrifugation and/or filtration, and the excess
of carbonate was removed with the solvent under reduced pressure or by
flash column chromatography (silica gel, dichloromethane, then dichloro-
methane/isopropanol 97:3–95:5).
ACTHNUTRGNEUNG
1H; 8a-H), 0.49 ppm (d, J=7.3 Hz, 3H; 8’-H); 13C{1H,31P} NMR
(125.76 MHz, [D8]THF): d=157.90, 157.77, 149.77, 148.51 (s, 4ꢃPh),
135.52 (s; PhAllyl), 133.91, 133.49, 133.00, 132.97 (s; 4ꢃPh), 132.86, 132.29
(d; 2ꢃBINOL), 131.43 (s; Ph), 130.14, 130.11, 129.98 (d; 2ꢃPh, PhAllyl),
129.87 (s; Ph), 129.57, 129.56, 129.50, 129.47 (d; 2ꢃBINOL, 2ꢃPhAllyl),
128.93 (d; Ph), 128.04, 127.95, 127.75, 127.64, 126.98, 126.92 (d; 6ꢃ
BINOL), 124.86 (d; Ph), 123.59 (s; Ph), 122.33, 122.32, 122.20 (d; 2ꢃ
BINOL, s; Ph), 121.54, 121.20, 111.30 (d; 4ꢃPh), 101.77 (d; CH-cod),
98.15 (d; C-2A), 92.57, 91.22 (d; 2ꢃCH-cod), 86.26 (d; C-1A), 84.25 (d;
CH-cod), 55.75, 55.25 (q; 2ꢃOCH3), 53.29 (d; C-7’), 40.63 (t; C-3A),
35.71, 33.93, 28.36, 27.00 (t; 4ꢃCH2-cod), 19.16 (q; C-8’), 15.11 ppm (t;
C-8); 31P NMR (121.50 MHz, [D8]THF): d=118.96 ppm; HRMS (ESI+):
m/z calcd for C55H54IrNO4P+: 1016.3414; found: 1016.3452 [M]+.
General procedure for the allylic substitutions according to Scheme 1
(GP2): Allylic substitution: a) NuH as pronucleophile (“salt-free condi-
tions“): In a dry Schlenk flask under argon, a solution of [{IrACHTNUTRGNEUNG(cod)Cl}2]
(2 mol%), L2 (4 mol%) and TBD (8 mol%) (in situ procedure) or K4c
(4 mol%) in absolute THF (0.5 mL) was prepared; then carbonate 1
(0.5 mmol), a pronucleophile (0.6 mmol) and possibly TBD (8 mol%)
were added and conversion monitored by TLC. b) NaNu as pronucleo-
phile: As in a), however, a solution of NaNu in THF (0.5 mL), which was
freshly prepared from NuH (0.6 mmol) and NaH (0.6 mmol), was used.
K4a: According to GP1 [{IrACHTNUTRGNE(NUG cod)Cl}2] (40.3 mg, 60.0 mmol) and (R,R,aR)-
L2 (72.0 mg, 120.0 mmol) were stirred in dry THF (3 mL) at room tem-
perature for 30 min. Cinnamyl methyl carbonate (49.9 mg, 260.0 mmol)
and AgClO4 (24.6 mg, 120.0 mmol) were added. After 20 h the formed
white precipitate was removed by filtration, and the crude product was
subjected to flash column chromatography (silica gel, dichloromethane,
then dichloromethane/isopropanol 97:3) to yield K4a (115.2 mg, 86%) as
a light yellow powder. 1H NMR (500.13 MHz, [D8]THF): d=8.41 (d, J=
8.6 Hz, 1H; BINOL), 8.31 (d, J=8.9 Hz, 1H; BINOL), 8.12 (d, J=
8.5 Hz, 1H; BINOL), 8.10 (d, J=8.5 Hz, 1H; BINOL), 8.03 (d, J=
8.8 Hz, 1H; BINOL), 7.93 (d, J=9.0 Hz, 1H; BINOL), 7.86–7.88 (m,
2H; 2ꢃPhAllyl), 7.24–7.57 (m, 13H; 4ꢃPh, 6ꢃBINOL, 3ꢃPhAllyl), 7.08–
7.13 (m, 2H; 2ꢃPh), 6.99 (dd, J=7.5, 7.5 Hz, 1H; 4-H), 6.93 (d, J=
8.3 Hz, 1H; 6-H), 5.91 (dd, JP, H =12.0 Hz, J=12.0 Hz, 1H; 1A-H), 5.34–
5.38 (m, 1H; cod-CH), 4.93–4.99 (m, 1H; 2A-H), 4.67 (dd, J=11.6,
4.9 Hz, 1H; 7-H), 4.48–4.57 (m, 1H; 7’-H), 4.11 (s, 3H; OCH3), 3.95–4.01
(m, 1H; cod-CH), 3.89–3.94 (m, 1H; cod-CH), 3.58 (s, 3H; OCH3), 2.76–
2.90 (m, 4H; cod-CH, 2ꢃcod-CH2, 3A,syn-H), 2.56–2.64 (m, 1H; cod-
CH2), 2.35 (d, J=10.8 Hz, 1H; 3A,anti-H), 2.19–2.31 (m, 2H; cod-CH2, 8b-
H), 2.10–2.13 (dd, J=14.7, 8.2 Hz, 1H; cod-CH2), 1.76–1.79 (m, 1H; cod-
CH2), 1.55–1.62 (m, 1H; cod-CH2), 1.41–1.51 (m, 1H; cod-CH2), 0.90
(dd, J=12.0, 12.0 Hz, 1H; 8a-H), 0.48 ppm (d, J=7.4 Hz, 3H; 8’-H); 13C-
K4c: According to GP1 [{IrACHTNUTRGNE(NUG cod)Cl}2] (40.4 mg, 60.0 mmol) and (S,S,aS)-
L2 (71.9 mg, 120.0 mmol) were stirred in dry THF (3 mL) at room tem-
perature for 30 min. Crotyl methyl carbonate (33.1 mg, 250.0 mmol) and
AgClO4 (26.2 mg, 130.0 mmol) were added. After 6 h the formed white
precipitate was removed by filtration, and the crude product was subject-
ed to flash column chromatography (silica gel, dichloromethane, then di-
chloromethane/isopropanol 95:5) to yield K4c (129.3 mg, 99%) as a light
yellow powder containing 4% of an isomeric by-product. All analytical
data given refers to a 96:4 mixture of the isomers (ratio determined by
31P NMR spectroscopy). 1H NMR (500.13 MHz, [D8]THF, major isomer):
d=8.39 (d, J=9.0 Hz, 1H; BINOL), 8.26 (d, J=8.5 Hz, 1H; BINOL),
8.11 (d, J=7.9 Hz, 1H; BINOL), 8.08 (d, J=8.5 Hz, 1H; BINOL), 7.93
(d, J=8.5 Hz, 1H; BINOL), 7.91 (d, J=8.5 Hz, 1H; BINOL), 7.49–7.56
(m, 3H; Ph, 2ꢃBINOL), 7.43 (d, J=7.3 Hz, 1H; Ph), 7.30–7.39 (m, 4H;
4ꢃBINOL), 7.19–7.27 (m, 2H; 2ꢃPh), 7.06–7.11 (m, 1H; Ph), 6.99 (d,
J=7.9 Hz, 1H; Ph), 6.89–6.96 (m, 2H; 2ꢃPh), 5.17–5.24 (m, 1H; cod-
CH), 4.85–4.94 (m, 1H; 1A-H), 4.42–4.52 (m, 2H; 7-H, 7’-H), 4.22 (ddd,
J=10.7, 10.7, 8.5 Hz, 1H; 2A-H), 3.98–4.06 (m, 1H; cod-CH), 3.88 (s,
3H; OCH3), 3.54 (s, 3H; OCH3), 3.32–3.39 (m, 1H; cod-CH), 3.21–3.30
(m, 1H; cod-CH2), 3.11–3.18 (m, 1H; cod-CH), 2.84–2.96 (m, 2H; 3A,syn
-
ACHTUNGTRENNUNG
{1H,31P} NMR (125.76 MHz, [D8]THF): d=157.85, 157.76, 149.83, 148.69
H, cod-CH2), 2.33–2.42 (m, 1H; cod-CH2), 2.24–2.32 (m, 1H; cod-CH2),
2.20 (d, J=11.3 Hz, 1H; 3A,anti-H), 2.03–2.10 (m, 2H; cod-CH2, 8b-H),
1.85 (dd, J=8.8, 6.0 Hz, 3H; CH3), 1.59–1.67 (m, 1H; cod-CH2), 1.40–
1.56 (m, 2H; 2ꢃcod-CH2), 0.81 (dd, J=12.1, 12.1 Hz, 1H; 8a-H),
(s; 4ꢃPh), 135.92 (s; PhAllyl), 133.95, 133.42 (s; 2ꢃPh), 133.10 (d;
BINOL), 133.03, 132.29 (s; 2ꢃPh), 132.19 (d; BINOL), 131.70 (s; Ph),
130.09, 129.96, 129.96, 129.87, 129.67, 129.60, 129.46, 129.29, 128.77 (d;
3ꢃPh, 2ꢃBINOL, 2ꢃPhAllyl, s; Ph), 127.95, 127.74, 127.69, 127.69, 126.82,
126.81 (d; 6ꢃBINOL), 124.84 (d; Ph), 123.71 (s; Ph), 123.02, 122.31 (d;
2ꢃBINOL), 122.13 (s; Ph), 121.51, 121.33, 111.28, 111.20 (d; 4ꢃPh),
102.42 (d; CH-cod), 98.21 (d; C-2A), 92.72, 90.44 (d; 2ꢃCH-cod), 86.51
(d; C-1A), 83.65 (d; CH-cod), 60.44 (d; C-7), 55.82, 55.24 (q; 2ꢃOCH3),
53.21 (d; C-7’), 41.17 (t C-3A), 35.62, 34.03, 28.41, 26.93 (t; 4ꢃCH2-cod),
19.24 (q; C-8’), 14.84 ppm (t; C-8); 31P NMR (202.46 MHz, [D8]THF):
d=120.01 ppm; HRMS (ESI+): m/z calcd for C55H54IrNO4P+:
1016.3414; found: 1016.3409 [M]+.
ACTHNUTRGNEUNG
0.43 ppm (d, J=7.3 Hz, 3H; 8’-H); 13C{1H,31P} NMR (125.76 MHz,
[D8]THF, major isomer): d=157.89, 157.80, 149.87, 148.73, 133.95, 133.43
(s; 6ꢃPh), 132.96 (d; BINOL), 132.88, 132.88 (s; 2ꢃPh), 132.09 (d;
BINOL), 131.92, 130.08 (s; 2ꢃPh), 129.89, 129.82 (d; 2ꢃPh), 129.54,
129.44 (d; 2ꢃBINOL), 128.63 (d; Ph), 127.88, 127.68, 127.68, 127.65,
126.74, 126.73 (d; 6ꢃBINOL), 124.97 (d; Ph), 123.70 (s; Ph), 122.93,
122.43 (d; 2ꢃBINOL), 122.18 (s; Ph), 121.36, 121.15, 111.35, 111.12 (d;
4ꢃPh), 107.72 (d; C-2A), 102.93, 91.71, 88.22, 83.09 (d; 4ꢃCH-cod), 76.76
(d; C-1A), 60.57 (d; C-7 or C-7’), 55.63, 55.17 (q; 2ꢃOCH3), 53.11 (d; C-7
or C-7’), 44.71 (t; C-3A), 37.27, 35.64, 28.48, 26.19 (t; 4ꢃCH2-cod), 19.22
(q; C-8’), 15.31 (q; CH3), 12.13 ppm (t; C-8); 31P NMR (202.46 MHz,
[D8]THF, major isomer): d=123.14 ppm; HRMS (ESI+): m/z calcd for
C50H52IrNO4P+: 954.3258; found: 954.3248 [M]+.
K4b: According to GP1 [{IrACTHNUTRGNE(NUG cod)Cl}2] (16.8 mg, 25.0 mmol) and
(R,R,aR)-L2 (30.0 mg, 50.0 mmol) were stirred in dry THF (1.3 mL) at
room temperature for 30 min. Cinnamyl methyl carbonate (19.2 mg,
100 mmol) and AgSbF6 (>20 mg (17.2 mgꢃ50.0 mmol)) were added.
After 20 h the formed white precipitate was removed by filtration, and
the crude product was subjected to flash column chromatography (silica
gel, dichloromethane/isopropanol 97:3) to yield K4b (43.6 mg, 70%) as
an orange powder. A saturated solution of K4b in THF, treated with
benzene, deposited crystals suitable for X-ray diffraction. These were iso-
lated by filtration. 1H NMR (500.13 MHz, [D8]THF): d=8.40 (d, J=
9.0 Hz, 1H; BINOL), 8.27 (d, J=9.0 Hz, 1H; BINOL), 8.12 (d, J=
8.5 Hz, 1H; BINOL), 8.10 (d, J=9.0 Hz, 1H; BINOL), 7.93 (d, J=
K4d: According to GP1 [{IrACTHNUTRGNE(NUG cod)Cl}2] (33.6 mg, 50.0 mmol) and
(R,R,aR)-L2 (60.0 mg, 100.0 mmol) were stirred in dry THF (2.5 mL) at
room temperature for 30 min. Crotyl methyl carbonate (26.0 mg,
200.0 mmol) and an excess of AgSbF6 (40–50 mg (34.4 mgꢃ100.0 mmol))
were added. After 20 h the formed white precipitate was removed by fil-
tration, and the crude product was subjected to flash column chroma-
AHCTUNGTRENNUNG
Chem. Eur. J. 2010, 16, 6601 – 6615
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6613

G. Helmchen et al.
an isomeric by-product. All analytical data given refers to a 97:3 mixture
of the isomers (ratio determined by 31P NMR spectroscopy). A saturated
solution of K4d in THF, treated with benzene, deposited crystals suitable
for X-ray diffraction. These were isolated by filtration. 1H NMR
(300.13 MHz, [D8]THF, major isomer): d=8.39 (d, J=8.9 Hz, 1H;
BINOL), 8.23 (d, J=8.7 Hz, 1H; BINOL), 8.12 (d, J=8.4 Hz, 1H;
BINOL), 8.09 (d, J=8.4 Hz, 1H; BINOL), 7.92 (d, J=8.7 Hz, 1H;
BINOL), 7.75 (d, J=8.9 Hz, 1H; BINOL), 7.50–7.59 (m, 3H; Ph, 2ꢃ
BINOL), 7.24–7.43 (m, 7H; 3ꢃPh, 4ꢃBINOL), 7.09 (dd, J=7.6, 7.6 Hz,
1H; Ph), 7.01 (d, J=8.1 Hz, 1H; Ph), 6.91–6.96 (m, 2H; 2ꢃPh), 5.11–
5.19 (m, 1H; cod-CH), 4.68–4.82 (m, 1H; 1A-H), 4.41–4.55 (m, 2H; 7-H,
7’-H), 4.22–4.32 (m, 1H; 2A-H), 4.02–4.13 (m, 1H; cod-CH), 3.90 (s, 3H;
OCH3), 3.58 (s, 3H; OCH3), 3.19–3.35 (m, 2H; cod-CH, cod-CH2), 3.07–
3.15 (m, 1H; cod-CH), 2.79–2.98 (m, 2H; 3A,syn-H, cod-CH2), 2.25–2.38
(m, 2H; 2ꢃcod-CH2), 2.08–2.20 (m, 2H; 8b-H, cod-CH2), 1.94 (d, J=
11.2 Hz, 1H; 3A,anti-H), 1.88 (dd, J=8.9, 6.1 Hz, 3H; CH3), 1.40–1.77 (m,
3H, 3ꢃcod-CH2), 0.86 (dd, J=11.5, 11.5 Hz, 1H; 8a-H), 0.44 ppm (d, J=
precipitate was removed by filtration, and the crude product was subject-
ed to flash column chromatography (silica gel, dichloromethane, then di-
chloromethane/isopropanol 97:3) to yield K4 f (91.2 mg, 88%) as
a
yellow powder containing 2% of an isomeric by-product. All analytical
data given refers to a 98:2 mixture of the isomers (ratio determined by
31P NMR spectroscopy). A saturated solution of K4 f in THF, treated
with benzene, deposited crystals suitable for X-ray diffraction. These
were isolated by filtration. 1H NMR (500.13 MHz, [D8]THF, major
isomer): d=8.36 (d, J=8.8 Hz, 1H; BINOL), 8.28 (d, J=8.8 Hz, 1H;
BINOL), 8.09 (d, J=7.7 Hz, 1H; BINOL), 8.07 (d, J=7.1 Hz, 1H;
BINOL), 8.02 (d, J=8.8 Hz, 1H; BINOL), 7.97 (d, J=8.8 Hz, 1H;
BINOL), 7.50–7.54 (m, 3H; Ph, 2ꢃBINOL), 7.13–7.39 (m, 13H; 9ꢃPh,
4ꢃBINOL), 5.13–5.17 (m, 1H; cod-CH), 4.79–4.86 (m, 1H; 1A-H), 4.53
(ddd, J=11.5, 11.5, 7.7 Hz, 1H; 2A-H), 4.13–4.19 (m, 1H; cod-CH), 3.99
(dd, J=11.5, 5.8 Hz, 1H; 7-H), 3.86–3.92 (m, 1H; 7’-H), 3.40 (dt, J=8.2,
7.5 Hz, 1H; cod-CH), 3.25 (dd, J=14.5, 8.5 Hz, 1H; cod-CH2), 3.15 (dd,
J=7.8, 5.9 Hz, 1H; cod-CH), 2.86–2.94 (m, 2H; 3A,syn-H, cod-CH2), 2.32–
2.41 (m, 3H; 3A,anti-H, 2ꢃcod-CH2), 2.05 (dd, J=14.8, 8.0 Hz, 1H; cod-
CH2), 1.92 (dd, J=12.3, 5.5 Hz, 1H; 8b-H), 1.84 (brs, 3H; CH3), 1.49–
1.70 (m, 3H; 3ꢃcod-CH2), 1.04–1.09 (m, 1H; 8a-H), 0.57 ppm (d, J=
ACTHNUTRGNEUNG
7.5 Hz, 3H; 8’-H); 13C{1H,31P} NMR (125.76 MHz, [D8]THF, major
isomer): d=157.90, 157.86, 149.81, 148.55, 133.91, 133.51, 132.92, 132.92
(s, 8ꢃPh), 132.60, 132.20 (d, 2ꢃBINOL), 131.64, 130.05 (s, 2ꢃPh),
129.96, 129.82, 129.48, 129.48, 128.78 (d, 3ꢃPh, 2ꢃBINOL), 127.98,
127.92, 127.76, 127.59, 126.91, 126.84 (d, 6ꢃBINOL), 124.91 (d; Ph),
123.58 (s; Ph), 122.38, 122.34, 122.27 (d, 2ꢃBINOL, s; Ph), 121.35,
120.97, 111.38, 111.21 (d; 4ꢃPh), 107.70 (d; C-2A), 102.33, 91.50, 88.88,
83.51 (d; 4ꢃCH-cod), 76.51 (d; C-1A), 60.62 (d; C-7 or C-7’), 55.59, 55.16
(q; 2ꢃOCH3), 53.13 (d; C-7 or C-7’), 44.07 (t; C-3A), 37.19, 35.73, 28.42,
26.20 (t; 4ꢃCH2-cod), 19.12 (q; C-8’), 15.18 (q; CH3), 12.29 ppm (t; C-8);
31P NMR (202.46 MHz, [D8]THF, major isomer): d=122.33 ppm; HRMS
(ESI+): m/z calcd for C50H52IrNO4P+: 954.3258; found: 954.3279 [M]+.
ACTHNUTRGNEUNG
7.4 Hz, 3H; 8’-H); 13C{1H,31P} NMR (125.76 MHz, [D8]THF, major
isomer): d=149.89, 148.76, 144.07, 141.41, 133.88, 133.48, 132.98 (s, 7ꢃ
Ph), 132.92 (d; BINOL), 132.83 (s; Ph), 132.33 (d; BINOL), 129.56,
129.40, 129.07, 128.60, 128.14, 127.84, 127.76, 127.71, 127.67, 127.51,
126.77 (d; 5ꢃPh, 6ꢃBINOL), 123.44 (s; Ph), 123.04, 122.37 (d; 2ꢃ
BINOL), 122.29 (s; Ph), 107.64 (d; C-2A), 103.66, 92.43, 87.87, 82.98 (d;
4ꢃCH-cod), 77.03 (d; C-1A), 66.38 (d; C-7), 60.77 (d; C-7’), 45.20 (t; C-
3A), 37.11, 35.34, 28.60, 26.19 (t; 4ꢃCH2-cod), 19.03 (q; C-8’), 15.14 (t; C-
8), 14.95 ppm (q; CH3); 31P NMR (121.50 MHz, [D8]THF, major isomer):
d=122.80 ppm; HRMS (ESI+): m/z calcd for C48H48IrNO2P+: 894.3046;
found: 894.3071 [M]+.
K4e: According to GP1 [{IrACHTNUTRGNE(NUG cod)Cl}2] (33.6 mg, 50.0 mmol) and (S,S,aS)-
L2 (60.0 mg, 100.0 mmol) were stirred in dry THF (2.5 mL) at room tem-
perature for 30 min. Crotyl methyl carbonate (26.0 mg, 200.0 mmol) and
AgOTf (26.0 mg, 100.0 mmol) were added. After 20 h the formed white
precipitate was removed by filtration, and the crude product was subject-
ed to flash column chromatography (silica gel, dichloromethane/isopro-
panol 97:3) to yield K4e (98.2 mg, 88%) as a light yellow powder. A sa-
turated solution of K4e in THF, treated with benzene, deposited crystals
suitable for X-ray diffraction. These were isolated by filtration. 1H NMR
(500.13 MHz, [D8]THF): d=8.38 (d, J=9.0 Hz, 1H; BINOL), 8.25 (d,
J=9.0 Hz, 1H; BINOL), 8.11 (d, J=8.5 Hz, 1H; BINOL), 8.08 (d, J=
8.5 Hz, 1H; BINOL), 7.89–7.93 (m, 2H; 2ꢃBINOL), 7.49–7.55 (m, 3H;
Ph, 2ꢃBINOL), 7.43 (d, J=7.3 Hz, 1H; Ph), 7.31–7.38 (m, 4H; 4ꢃ
BINOL), 7.21–7.27 (m, 2H; 2ꢃPh), 7.08 (dd, J=7.1, 7.1 Hz, 1H; Ph),
6.99 (d, J=8.5 Hz, 1H; Ph), 6.94 (dd, J=7.6, 7.6 Hz, 1H; Ph), 6.91 (d,
J=8.5 Hz, 1H; Ph), 5.18–5.23 (m, 1H; cod-CH), 4.88–4.95 (m, 1H; 1A-
H), 4.44–4.50 (m, 2H; 7-H, 7’-H), 4.18–4.24 (m, 1H; 2A-H), 4.01 (brs,
1H; cod-CH), 3.87 (s, 3H; OCH3), 3.54 (s, 3H; OCH3), 3.40–3.44 (m,
1H; cod-CH), 3.22–3.27 (m, 1H; cod-CH2), 3.12–3.16 (m, 1H; cod-CH),
2.84–2.94 (m, 2H; 3A,syn-H, cod-CH2), 2.22–2.41 (m, 3H; 3A,anti-H, 2ꢃcod-
CH2), 2.04–2.12 (m, 2H; 8b-H, cod-CH2), 1.85 (d, J=6.2 Hz, 3H; CH3),
1.60–1.68 (m, 1H; cod-CH2), 1.41–1.57 (m, 2H; 2ꢃcod-CH2), 0.81 (dd,
J=12.2, 12.1 Hz, 1H; 8a-H), 0.43 ppm (d, J=7.4 Hz, 3H; 8’-H); 13C-
K4h: According to GP1 [{IrACHTNUTRGNE(NUG cod)Cl}2] (33.6 mg, 50.0 mmol) and (S,S,aS)-
L2 (60.0 mg, 100.0 mmol) were stirred in dry THF (2.5 mL) at room tem-
perature for 30 min. Cinnamyl methyl carbonate (38.0 mg, 200.0 mmol)
and AgOTf (26.0 mg, 100.0 mmol) were added. After 20 h the formed
white precipitate was removed by filtration, and the crude product was
subjected to flash column chromatography (silica gel, dichloromethane/
isopropanol 97:3) to yield K4h (105.0 mg, 89%) as a light yellow powder.
1H NMR (500.13 MHz, [D8]THF): d=8.39 (d, J=9.2 Hz, 1H; BINOL),
8.31 (d, J=9.2 Hz, 1H; BINOL), 8.12 (d, J=8.1 Hz, 1H; BINOL), 8.10
(d, J=8.1 Hz, 1H; BINOL), 8.02 (d, J=8.9 Hz, 1H; BINOL), 7.88–7.94
(m, 3H; BINOL, 2ꢃPhAllyl), 7.24–7.58 (m, 13H; 4ꢃPh, 6ꢃBINOL, 3ꢃ
PhAllyl), 7.08–7.13 (m, 2H; 2ꢃPh), 6.99 (dd, J=7.6 Hz, J=7.6 Hz, 1H;
Ph), 6.94 (d, J=8.4 Hz, 1H; Ph), 5.98 (dd, JP, H =12.1 Hz, J=12.1 Hz, 1H;
1A-H), 5.33–5.41 (m, 1H; cod-CH), 4.98 (ddd, J=11.7 Hz, J=11.6 Hz,
J=7.8 Hz, 1H; 2A-H), 4.69 (dd, J=11.7 Hz, J=4.8 Hz, 1H; 7-H), 4.53
(dq, JP, H =21.8 Hz, J=7.3 Hz, 1H; 7’-H), 4.11 (s, 3H; OCH3), 3.95–4.06
(m, 2H; 2 cod-CH), 3.58 (s, 3H; OCH3), 2.74–2.91 (m, 4H; 3A,syn-H, cod-
CH, cod-CH2), 2.56–2.67 (m, 1H; cod-CH2), 2.43 (d, J=10.9 Hz, 1H;
3A,anti-H), 2.05–2.35 (m, 3H; 8b-H, 2 cod-CH2), 1.36–1.79 (m, 3H; 3 cod-
CH2), 0.91 (dd, J=12.4 Hz, J=11.7 Hz, 1H; 8a-H), 0.49 ppm (d, J=
7.5 Hz, 3H; 8’-H); 13C {1 H, 31P} NMR (125.76 MHz, [D8]THF): d=
157.85, 157.77, 149.84, 148.69 (4s; Ph), 136.00 (s; PhAllyl), 133.97, 133.42,
133.08, 133.04 (4s; Ph), 132.92, 132.14 (d; 2ꢃBINOL), 131.76 (s; Ph),
130.09, 130.01, 129.91, 129.91, 129.69, 129.59, 129.45, 129.24, 128.77 (8d;
3ꢃPh, 2ꢃBINOL, 3ꢃPhAllyl, s; Ph), 127.98, 127.76, 127.71, 127.69, 126.85,
126.81 (d; 6ꢃBINOL), 124.86 (d; Ph), 123.75 (s; Ph), 123.01, 122.28 (2d;
BINOL), 122.14 (s; Ph), 121.49, 121.30, 111.26, 111.21 (d; 4ꢃPh), 102.43
(d; CH-cod), 98.26 (d; C-2A), 92.72, 90.39 (d; 2ꢃCH-cod), 86.52 (d; C-
1A), 83.60 (d; CH-cod), 60.48 (d; C-7), 55.74, 55.21 (q; 2ꢃOCH3), 53.22
(d; C-7’), 41.16 (t; C-3A), 35.60, 34.01, 28.41, 26.85 (t; 4ꢃCH2-cod), 19.22
(q; C-8’), 14.78 ppm (t; C-8); 31P NMR (121.50 MHz, [D8]THF): d=
120.15 ppm; HRMS (ESI+): m/z calcd for C48H48IrNO2P+: 1016.3414;
found: 1016.3420 [M]+.
ACHTUNGTRENNUNG C atom of
{1H,31P} NMR (125.76 MHz, [D8]THF): the quaternary
CF3SO3 could not be assigned to a 13C signal; d=157.88, 157.81, 149.86,
148.71, 133.95, 133.44, 132.96, 132.88 (s; 8ꢃPh), 132.83, 132.11 (d; 2ꢃ
BINOL), 131.91, 130.09 (s; 2ꢃPh), 129.89, 129.82 (d; 2ꢃPh), 129.53,
129.45 (d; 2ꢃBINOL), 128.66 (d; Ph), 127.91, 127.74, 127.69, 127.64,
126.77, 126.77 (d; 6ꢃBINOL), 124.94 (d; Ph), 123.70 (s; Ph), 122.87,
122.40 (d; 2ꢃBINOL), 122.21 (s; Ph), 121.36, 121.11, 111.36, 111.12 (d;
4ꢃPh), 107.72 (d; C-2A), 102.90, 91.71, 88.23, 83.07 (d; 4ꢃCH-cod), 76.68
(d; C-1A), 60.57 (d; C-7 or C-7’), 55.61, 55.15 (q; 2ꢃOCH3), 53.10 (d; C-7
or C-7’), 44.64 (t; C-3A), 37.24, 35.61, 28.47, 26.14 (t; 4ꢃCH2-cod), 19.22
(q; C-8’), 15.26 (q; CH3), 12.10 ppm (t; C-8); 31P NMR (202.46 MHz,
[D8]THF): d=123.19 ppm; HRMS (ESI+): m/z calcd for C50H52IrNO4P+:
954.3258; found: 954.3294 [M]+.
K4 f: According to GP1 [{IrACHTNUTRGNE(NUG cod)Cl}2] (33.6 mg, 50.0 mmol) and (S,S,aS)-
L1 (54.0 mg, 100.0 mmol) were stirred in dry THF (2.5 mL) at room tem-
perature for 30 min. Crotyl methyl carbonate (26.0 mg, 200.0 mmol) and
AgOTf (25.7 mg, 100.0 mmol) were added. After 20 h the formed white
6614
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 6601 – 6615

Asymmetric Catalysis
FULL PAPER
NMR tube at ꢀ808C, and the solution was allowed to warm up to
RT. Excepting line broadening due to the low temperature,
31P NMR spectra did not change.
Acknowledgements
This work was supported by the Fonds der Chemischen Industrie, the
Deutsche Forschungsgemeinschaft (SFB 623), the Studienstiftung des
Deutschen Volkes (J.A.R.) and the Landesgraduiertenfçrderung (S.S.).
We thank Prof. B. F. Straub and Dr. M. Hofmann for helpful discussions.
[19] a) B. Bartels, C. Garcia-Yebra, G. Helmchen, Eur. J. Org. Chem.
2003, 1097; b) A. Leitner, C. Shu, J. F. Hartwig, Proc. Natl. Acad.
Sci. USA 2004, 101, 5830.
[20] M. J. Pouy, L. M. Stanley, J. F. Hartwig, J. Am. Chem. Soc. 2009, 131,
11312.
[21] The first two allylic complexes were obtained through replacement
of the PMe3 ligand by an unsubstituted allylic moiety, yielding the
D-P-endo and D-P-exo diastereomers. Geometric alterations within
these two complexes were conducted by 1) altering the helicity (L
or D) of the complex through the appropriate modifications of the
coordination for both cod and the allylic fragment and 2) rotating
[1] Reviews: a) G. Helmchen, A. Dahnz, P. Dꢄbon, M. Schelwies, R.
Weihofen, Chem. Commun. 2007, 675; b) G. Helmchen in Iridium
Complexes in Organic Synthesis (Eds.: L. A. Oro, C. Claver), Wiley-
VCH, Weinheim, 2009, pp. 211–250; c) R. Takeuchi, S. Kezuka,
Synthesis 2006, 3349; d) H. Miyabe, Y. Takemoto, Synlett 2005, 1641.
[2] C. Welter, O. Koch, G. Lipowsky, G. Helmchen, Chem. Commun.
2004, 896.
[3] C. A. Kiener, C. Shu, C. Incarvito, J. F. Hartwig, J. Am. Chem. Soc.
2003, 125, 14272.
[4] S. Spiess, C. Welter, G. Franck, J.-P. Taquet, G. Helmchen, Angew.
Chem. 2008, 120, 7764; Angew. Chem. Int. Ed. 2008, 47, 7652.
[5] R. Weihofen, O. Tverskoy, G. Helmchen, Angew. Chem. 2006, 118,
5673; Angew. Chem. Int. Ed. 2006, 45, 5546.
[6] J. Tsuji, I. Minami, I. Shimizu, Tetrahedron Lett. 1984, 25, 5157.
[7] DFT calculations have been carried out on chiral intermediates of
Ir-catalysed allylic substitutions that have not been substantiated ex-
perimentally: a) N. Kinoshita, K. H. Marx, K. Tanaka, K. Tsubaki,
T. Kawabata, N. Yoshikai, E. Nakamura, K. Fuji, J. Org. Chem.
2004, 69, 7960; b) M. Kimura, Y. Uozumi, J. Org. Chem. 2007, 72,
707.
[8] S. Spiess, J. A. Raskatov, C. Gnamm, K. Brçdner, G. Helmchen,
Chem. Eur. J. 2009, 15, 11087.
[9] a) D. Polet, A. Alexakis, K. Tissot-Croset, C. Corminboeuf, K. Di-
trich, Chem. Eur. J. 2006, 12, 3596; b) K. Tissot-Croset, D. Polet, A.
Alexakis, Angew. Chem. 2004, 116, 2480; Angew. Chem. Int. Ed.
2004, 43, 2426.
the IrACTHNUGRTENUNG(C,P-L1) fragment so that one of the allylic termini occupied
the coordination site trans to the CH2 group of the chiral ligand.
The conformation of the chiral ligand backbone and the five-mem-
bered iridacycle were kept within the crystallographically deter-
mined conformation.
[22] a) J. B. Wakefield, J. M. Stryker, J. Am. Chem. Soc. 1991, 113, 7057;
b) E. B. Tjaden, K. E. Schwiebert, J. M. Stryker, J. Am. Chem. Soc.
1992, 114, 1100; c) E. B. Tjaden, J. M. Stryker, J. Am. Chem. Soc.
1990, 112, 6420.
[23] A. D. Becke, J. Chem. Phys. 1993, 98, 1372; A. D. Becke, J. Chem.
Phys. 1993, 98, 5648.
[24] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785.
[25] Gaussian 03, Revision B.03, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgom-
AHCTUNGERTGeNNUN ry, Jr., T. Vreven, N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyen-
gar, J. Tomasi, V. Barone, B Mennucci, M. Cossi, G. Scalmani, N.
Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
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, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.
Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C.
Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cio-
slowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaro-
mi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng,
A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wall-
ingford CT, 2004.
´
[10] D. Markovic, J. F. Hartwig, J. Am. Chem. Soc. 2007, 129, 11680.
´
[11] S. T. Madrahimov, D. Markovic, J. F. Hartwig, J. Am. Chem. Soc.
2009, 131, 7228.
[12] The complex K1H does not seem to affect further transformations.
In the following the mixture of K1 and K1H as depicted in Fig-
ure 1A is designated K1.
[13] K5 is, of course, the first product of the substitution reaction. The
fact that it was not found in the reaction mixture to a significant
extent during the reaction suggested the possibility that it is formed
by a reaction between K1 or K2 and the decomplexed reaction
product 2. To test this possibility, K1 was generated by mixing [{Ir-
[26] a) R. Ditchfield, W. J. Hehre, J. A. Pople, J. Chem. Phys. 1971, 54,
724; b) W. J. Hehre, R. Ditchfield, J. A. Pople, J. Chem. Phys. 1972,
56, 2257; c) P. C. Hariharan, J. A. Pople, Theo Chim. Acta 1973, 28,
213; d) P. C. Hariharan, J. A. Pople, Mol. Phys. 1974, 27, 209;
e) M. S. Gordon, Chem. Phys. Lett. 1980, 76, 163.
[27] D. Andrae, U. Haeussermann, M. Dolg, H. Stoll, H. Preuss, Theor.
Chim. Acta 1990, 77, 123.
[28] C. Y. Peng, H. B. Schlegel, Isr. J. Chem. 1994, 34, 449.
[29] C. Benzi, M. Cossi, V. Barone, R. Tarroni, C. Zannoni, J. Phys.
Chem. B 2005, 109, 2584.
[30] J. Tomasi, B. Mennucci, R. Cammi, Chem. Rev. 2005, 105, 2999.
[31] H. E. Gottlieb, V. Kotylar, A. Nudelman, J. Org. Chem. 1997, 62,
7512.
[32] S. Berger, Angew. Chem. 2004, 116, 2070.
[33] SADABS 2008/1, G. M. Sheldrick, Bruker Analytical X-ray Divi-
sion, Madison, Wisconsin 2008.
ACHTUNGTRENNUNG(cod)Cl}2] and L2 (2 equiv) in THF; then an excess of the branched
amination product 2, Nu=HNBn or HNPh, and TBD (3–4 equiv),
were added. Indeed, K5 was found to be the main species of the
mixture after 1–2 h (31P NMR spectroscopy; see the Supporting In-
formation). For further experiments concerning this issue see the
Supporting Information of reference [10].
[14] Treatment of a complex K1 containing the ligand (aR,S,S)-L1 with
ꢀ
AgSbF6 in CH2Cl2 does not effect C H activation; a complex [Ir-
(cod)
R
is stabilised by a Ph!Ir coordination. See: T. Osswald, H. Rꢄegger,
A. Mezzetti, Chem. Eur. J. 2010, 16, 1388.
[15] We have earlier prepared and structurally characterised a (p-allyl)Ir
complex with a PHOX ligand. Dimethyl malonate reacted with this
complex at the central rather than the terminal allylic carbon, see C.
Garcia-Yebra, J. P. Janssen, F. Rominger, G. Helmchen, Organome-
tallics 2004, 23, 5459; for a (p-allyl)Ir complex containing a pybox
ligand see P. Paredes, J. Dꢅez, M. P. Gamusa, Organometallics 2008,
27, 2597.
[16] Remarkably, crystals were only obtained if benzene was contained
in the solvent, see reference [11].
[17] T. Ohmura, J. F. Hartwig, J. Am. Chem. Soc. 2002, 124, 15164.
[18] To test for identity of the crystal structure and the structure in solu-
tion, a crystal of complex K4a was dissolved in [D8]THF in an
[34] SHELXTL 2008/4, G. M. Sheldrick, Acta Crystallogr. Sect. A 2008,
64, 112.
[35] C. Gnamm, G. Franck, N. Miller, T. Stork, K. Brçdner, G. Helm-
chen, Synthesis 2008, 3331.
[36] C. Gnamm, S. Fçrster, N. Miller, K. Brçdner, G. Helmchen, Synlett
2007, 790.
Received: December 17, 2009
Published online: April 23, 2010
Chem. Eur. J. 2010, 16, 6601 – 6615
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6615