Highly Enantioselective Ferrocenyl Palladacycle-Acetate Catalysed Arylation of Aldimines and Ketimines with Arylboroxines

Article abstract of DOI:10.1002/chem.201605244

Benzylic N-substituted stereocenters constitute a frequent structural motif in drugs. Their highly enantioselective generation is hence of technical importance. An attractive strategy is the arylation of imines with organoboron reagents. Chiral Rh complexes have reached a high level of productivity for this reaction type. In this article we describe that an electron rich Pd^IIcatalyst also performs well in the arylation of aldimines, comparable to the best Rh catalysts. The ferrocenyl palladacycle-acetate catalyst allows for a broad substrate scope and very high enantioselectivities. Commonly observed side reactions like aryl–aryl homocouplings and imine hydrolysis could be blocked. Mechanistic studies implicate that a) the acetate ligand is crucial for transmetallation, b) the active catalyst is most likely a palladacycle-OAc monomer, c) the rate limiting step is probably the product release. By added KOAc the arylation could also be applied to ketimines.

Full text of DOI:10.1002/chem.201605244

A Journal of
Accepted Article
Title: Highly Enantioselective Ferrocenyl Palladacycle-Acetate
Catalysed Arylation of Aldimines and Ketimines with
Arylboroxines
Authors: Carmen Schrapel, Wolfgang Frey, Delphine Garnier, and
René Peters
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201605244
Link to VoR: http://dx.doi.org/10.1002/chem.201605244
Supported by

10.1002/chem.201605244
Chemistry - A European Journal
FULL PAPER
Highly Enantioselective Ferrocenyl Palladacycle-Acetate
Catalysed Arylation of Aldimines and Ketimines with
Arylboroxines
Carmen Schrapel,[a] Wolfgang Frey,[a] Delphine Garnier,[a] and René Peters*[a]
Abstract: Benzylic N-substituted stereocenters constitute
a
reagents.[10] The latter include, e.g., aryl zinc,[11] titanium[12] and
tin[13] reagents.
frequent structural motif in drugs. Their highly enantioselective
generation is hence of technical importance. An attractive strategy
is the arylation of imines with organoboron reagents. Chiral Rh
complexes have reached a high level of productivity for this
reaction type. In this article we describe that an electron rich PdII
catalyst also performs well in the arylation of aldimines,
comparable to the best Rh catalysts. The ferrocenyl palladacycle-
acetate catalyst allows for a broad substrate scope and very high
enantioselectivities. Commonly observed side reactions like aryl-
aryl homocouplings and imine hydrolysis could be blocked.
Mechanistic studies implicate that (a) the acetate ligand is crucial
for transmetallation, (b) the active catalyst is most likely a
palladacycle-OAc monomer, (c) the rate limiting step is probably
the product release. By added KOAc the arylation could also be
applied to ketimines.
Figure 1. Examples for pharmacologically important chiral benzylic amines
containing a N-substituted stereocenter.
Introduction
Chiral enantiomerically pure amines often display a high
biological activity and are thus employed in a number of active
pharmaceutical ingredients (APIs) and agrochemicals.[1,2] An
important subclass form chiral benzylic amines containing a N-
substituted stereocenter,[3] because this structural motif can be
found in a number of drugs or drug candidates (Figure 1).
Prominent examples include Cetirizine[4] for the treatment of
hayfever, Sertraline[5] and Tianeptine[6] for the treatment of
depression, Rasagiline[7] (treatment of Parkinson’s disease),
Rivastigmine[8] (treatment of Alzheimer’s disease) or (+)-
BW373U86[9] (a non-peptidic δ-opioid agonist). Due to the
importance of chiral benzylic amines in medicine, efficient
methods suitable for their preparation on technical scale are
indispensable. For that reason a number of new synthetic
methods have been developed in the last few years, which
might provide advantages with regard to practicality,
Most frequently studied have arguably been arylboronic acid
derivatives.[3,10] The latter are synthetically particularly attractive
due to their high robustness, their high compatibility with many
functional groups and low toxicity.[10] In addition, their
preparation is operationally simple and they are often
commercially available at relatively low prices.[10] Rhodium
complexes are today well established as versatile catalysts for
the asymmetric arylation of imines with arylboronic acid
derivatives.[10,14,15] Moreover, several palladium(II)-catalysts
have recently been described as an attractive alternative for
the arylation of aldimines.[16-18]
Aryl-Pd species have been considered to be less nucleophilic
than aryl-Rh species,[19] possibly rendering the addition to the
C=N bond more difficult.[20] In contrast to the RhI-catalysed
reactions, the PdII-catalysed additions to aldimines are often
accompanied by a significant competing imine hydrolysis.
Molecular sieve was often necessary to minimise hydrolysis.[16]
Another problem using PdII catalysts is the formation of bisaryl
homocoupling side products (Ar-Ar).[16] The latter are most
likely generated by formation of off-cycle bisaryl-PdII
sustainability,
cost-efficiency,
enantioselectivity,
scope,
catalytic activity, catalyst robustness, etc.[10]
In this context the arylation of aldimines has been intensively
studied using various catalyst types and aryl transfer
[a]
Dr. Carmen Schrapel, Dr. Wolfgang Frey, Dr. Delphine Garnier,
Prof. Dr. René Peters
Institut für Organische Chemie, Universität Stuttgart,
Pfaffenwaldring 55, D-70569 Stuttgart, Germany
E-mail: rene.peters@oc.uni-stuttgart.de
intermediates, which are prone to undergo
a reductive
elimination to deliver the bisaryl product and at the same time
a catalyst at a non-productive Pd0 oxidation state. Both the
imine hydrolysis and the formation of the bisaryl-PdII
intermediate were probably facilitated by the fact that neutral
bidentate chiral ligands were employed thus resulting in a
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/chem.201605244
Chemistry - A European Journal
FULL PAPER
relatively high Lewis acidity of the catalytically active PdII
center.[16]
could initially solve this problem. The arylation product was
formed in
a
nearly quantitative yield and in almost
To address these issues we envisioned a monoanionic
C,N-ligand[21] to be beneficial. In particular the generation of an
undesired bisaryl-PdII intermediate seemed to be much less
favourable in that case, because a monoanionic Pd-center
would be formed. Additionally, to attain enantioselectivities
comparable to the best Rh-catalysts we studied planar chiral
palladacycles in order to facilitate a differentiation of the
enantiotopic imine faces at defined binding sites in the catalyst.
They were expected to be defined, because the position of
different ligand types in C,N-palladacycles is usually controlled
by their charges. Neutral ligands (substrates) usually prefer the
position trans to the N-donor, whereas anionic ligands bind
trans to the C-donor.[21]
In this article we describe that a readily accessible
ferrocene imidazoline palladacycle (abbreviated as FIP), which
was originally used by our group for allylic imidate
rearrangements,[22a,b] behaves as a highly enantioselective
catalyst for the arylation of imines[23] with arylboroxines.[24] By
that a large number of benzylic amines were obtained in
almost enantiomerically pure form with an efficiency that is
comparable to the best Rh catalysts. An important factor for
this performance is the use of acetate as anionic ligand.
enantiomerically pure form (entry 5). Also with just 0.5 mol% of
[FIP-Cl]2 excellent data were obtained (entry 6).[27]
Nevertheless, the reactivity strongly depended on the boronic
acid batch which was used. NMR analysis of various batches
of 2A freshly bought from different commercial suppliers
showed that they usually contain large quantities of the
corresponding boroxine. In one case, a commercial batch
labelled as boronic acid 2A was nearly pure boroxine 3A. This
batch resulted in no product formation applying the conditions
of entry 5. As a general rule product yields decreased with an
increasing percentage of the boroxine.[28] Our trials to
synthesise boronic acids of sufficient quality failed, because an
anhydrous material was essential in order to avoid the imine
hydrolysis. During removal of residual water the formation of
varying amounts of boroxine was always found.
In order to avoid the described reproducibility problems and
because anhydrous boronic acid 2A of reproducibly high
quality was not accessible, we considered the development of
a method which makes use of arylboroxine reagents to be
more attractive. The observation that phenylboroxine displays
a much lower reactivity indicated that the transmetallation step
initiated by KF was slower than with phenylboronic acid. We
were speculating that addition of an alcohol additive might
solve this problem, as the reaction of boroxines and alcohols
should create B-OH units in-situ, which might facilitate the
Results and Discussion
transmetallation step via
a
‘boronate-pathway’.[26] In the
presence of two equivalents of MeOH the reactivity was indeed
partly regained, yet at the expense of increased levels of imine
hydrolysis and a reduced enantioselectivity (entry 7).
Optimisation of the Arylation of Aldimines
The addition of commercial Ph-B(OH)2 (2A) to N-tosylimine 1a
was studied as a model reaction (Table 1). The resulting
product 4aA has previously been demonstrated to be a
precursor for Cetirizine.[14a] Two classes of metallocene derived
imidazoline metallacycles have recently been identified by our
research group to be efficient for various catalytic asymmetric
reactions: (i) metallocene bisimidazoline bismetallacycles[25]
and (ii) ferrocenyl monoimidazoline monopalladacycles.[22] The
ferrocen-2,2’-diyl bisimidazoline bispalladacycle precatalyst
[FBIP-Cl]2 (3.0 mol%) – activated by AgOTf (12 mol%) to
remove the four chloride bridges in order to facilitate a
coordination of the substrates[25] – provided the addition
product in only moderate yield when the reaction was
performed at room temperature in THF in the presence of
activated 4Å molecular sieves (MS). KF was utilised as
stoichiometric base to enable the transmetallation of the
boronic acid’s aryl residue to the Pd(II) center.[26] On the other
hand, 4aA was delivered in almost enantiomerically pure form
(Table 1, entry 1) and hydrolysis was not detected under these
conditions.
Also with monopalladacycle [FIP-Cl]2 (6 mol% to use the
same Pd loading as with the bispalladacycle) – again activated
by AgOTf (12 mol%)[22a,d] – 4aA was formed in almost
enantiopure form under otherwise unchanged conditions, but
with an improved yield (entry 2). The product yield was further
improved when the reaction was run in toluene as solvent
(entry 3). However, with a synthetically more useful catalyst
loading (1 mol% precatalyst), hydrolysis could not be
completely suppressed anymore (entry 4).
Alternatively, the transmetallation step might be facilitated
by an anionic ligand with an adequate Lewis-basicity for a
simultaneous coordination to the boroxine reagent, related to
the ‘oxo-palladium-pathway’ in Suzuki-Miyaura couplings.[26,29]
Acetate was found to be a suitable anionic ligand, with which
the transmetallation is apparently accelerated. During the
course of these studies, Fairlamb et al. described an
investigation of a stoichiometric transmetallation using a PdII-
OAc species, which confirmed this hypothesis.[30] Some
reactivity towards nearly enantiomerically pure 4aA was
already found at room temperature in the presence of
stoichiometric NaOAc (entry 8). This reactivity disappeared
again in the absence of NaOAc as additive at room
temperature (entry 9), whereas at a reaction temperature of
70 °C 4aA was formed in a quantitative yield (entry 10).
Different silver salts for chloride removal were then
investigated in the catalyst activation. These experiments
confirmed that the choice of the anionic ligand is central for a
satisfactory reactivity. Less Lewis-basic anions than acetate
like triflate and 2,4,6-triisopropylbenzensulfonate resulted in no
or trace amounts of product (entries 11 & 12). With
trifluoroacetate a yield of 24% of enantiomerically pure 4aA
was determined (entry 13). These results support the crucial
role of the acetate ligand in the transmetallation event.
To our surprise the direct use of the chloride-bridged [FIP-
Cl]2 as catalyst (1 mol%), without an activation by a silver salt,
This article is protected by copyright. All rights reserved.

10.1002/chem.201605244
Chemistry - A European Journal
FULL PAPER
Table 1. Development of the model reaction.
Precatalyst
(x mol%)
Y
2A or 3A
Solvent
T
imine
hydrolysis [%]a)
Yield
4aA
ee
4aA
#
Base
additive
(2 or 4x mol%)
(equiv.)
[°C]
(equiv.)
[%]a)
35
51
92
74
<1-99
98
75
25
2
[%]b)
>99
>99
>99
>99
>99
>99
87
1
[FBIP-Cl]2 (3)
[FIP-Cl]2 (6)
[FIP-Cl]2 (6)
[FIP-Cl]2 (1)
[FIP-Cl]2 (1)
[FIP-Cl]2 (0.5)
[FIP-Cl]2 (1)
[FIP-Cl]2 (1)
[FIP-Cl]2 (1)
[FIP-Cl]2 (1)
[FIP-Cl]2 (1)
[FIP-Cl]2 (1)
[FIP-Cl]2 (1)
[FIP-Cl]2 (1)
[FIP-Cl]2 (1)
[FIP-Cl]2 (1)
[FIP-Cl]2 (0.5)
[FIP-Cl]2 (1)
[FBIP-Cl]2 (0.5)
[FBIPP-Cl]2 (0.5)
[RuBIP-Cl]2 (0.5)
[PPFIP-Cl]2 (1)
[FIP(Tf)-Cl]2 (1)
OTf (4x)
OTf (2x)
OTf (2x)
OTf (2x)
-
'2A' (2)c)
'2A' (2)c)
'2A' (2)c)
'2A' (2)c)
2(3)A (2)
'2A' (2)c)
3A (0.8)
3A (1)
KF (1.0)
THF
THF
4Å MS
20
20
20
20
20
20
20
20
20
70
70
70
70
70
70
70
70
65
65
65
65
65
65
<1
<1
3
2
KF (1.0)
4Å MS
3
KF (1.0)
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
F3CC6H5
ClC6H5
1,2-Cl2C6H4
ClC6H5
ClC6H5
ClC6H5
ClC6H5
ClC6H5
ClC6H5
ClC6H5
4Å MS
4
KF (1.0)
4Å MS
13
<1-7
2
5
KF (1.0)
4Å MS
6
-
KF (1.0)
4Å MS
7
-
KF (1.0)
MeOHd)
8
8
OAc (2x)
OAc (2x)
OAc (2x)
OTf (2x)
O3S-2,4,6-iPr3C6H2 (2x)
O2CCF3 (2x)
OAc (2x)
OAc (2x)
OAc (2x)
OAc (2x)
OAc (2x)
OAc (4x)
OAc (4x)
OAc (4x)
OAc (2x)
OAc (2x)
NaOAc (1.0)
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
<1
<1
<1
11
2
>99
n.d.
>99
n.d.
n.d.
>99
>99
>99
>99
>99
>99
>99
n.d.
n.d.
n.d.
>99
9
3A (1)
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
3A (1)
99
<1
6
3A (1)
3A (1)
3A (1)
<1
<1
<1
<1
<1
<1
<1
1
24
95
98
76
82
99
35
3
3A (1)
3A (1)
3A (1)
3A (1)
3A (1)
3A (1)
3A (1)
3A (1)
<1
<1
<1
38
1
3A (1)
3A (1)
47
This article is protected by copyright. All rights reserved.

10.1002/chem.201605244
Chemistry - A European Journal
FULL PAPER
a)
b)
c)
Determined by 1H-NMR of the crude product using mesitylene as internal standard. Determined by HPLC. Commercial batch of 2a with a ratio boronic
acid/boroxine = 1.3:1, ⩯ 0.68 Äquiv. PhB(OH)2). d) 2 equiv. were used.
Because imine 1a and also product 4aA possess a poor
solubility in toluene, a precipitation of both compounds was
observed at the phase transitions solution/gas phase/glass wall.
Because the reactions were performed in a parallel synthesiser
(Heidolph Synthesis 1) which was constantly shaking during the
course of the reaction, a careful adjustment of the shaking
frequency was required (200-250 rpm) in order to avoid that
precipitated starting material at the vial’s wall would not react
anymore as it had not contact with the solution anymore. To
increase the reproducibility/yields, a screening of different
solvents was performed. In ethereal solvents (1,4-dioxane, THF,
diglyme) product formation was in general inferior (not shown)
compared to the use of toluene. In addition, several halogenated
aromatic solvents were tested. (α,α,α)-Trifluorotoluene and
chlorobenzene allowed for efficient arylation reactions (entries
14 & 15). In these solvents the formation of much less Pd-black
was observed indicating that there is less catalyst decomposition.
In particular chlorobenzene is a good choice as the solubility
was largely improved, whereas 1,2-dichlorobenzene led to
inferior results (entry 16). In chlorobenzene the precatalyst
amount could also be reduced to 0.5 mol% still providing a good
yield after 20 h at 70 °C (entry 17). A temperature screening
showed that the best data is obtained at 65 °C (entry 18).
= 18:1) the cyclopalladation was accomplished with Na2PdCl4 /
NaOAc in MeOH at room temperature.[22a] This older version
proceeded very rapidly as visible from the nearly instantaneous
product precipitation. In the modified procedure
a
MeOH/benzene mixture was used, in which the product has a
significantly higher solubility. This might favour the
diastereoselective formation of the thermodynamically more
stable diastereomer (93% yield).
The optimised conditions were also applied using different
bismetallacycles
and
a
pentaphenylferrocene
based
monopalladacycle. None of them performed with a similar
proficiency (entries 19-23). For the bismetallacycles a catalyst
loading of 0.5 mol% was used to have the same Pd loading as
with the monopalladacycle [FIP-Cl]2. With the mixed platina-
/palladacycle [FBIPP-Cl]2[31] only trace amounts of product were
Scheme 1. Improved synthesis of the precatalyst [FIP-Cl]2.
[25]
formed (entry 20), whereas the bispalladacycles [FBIP-Cl]2
[32]
and [RuBIP-Cl]2
provided moderate product yields (entries
19 & 21). With the sterically demanding pentaphenylferrocene
[22]
based monopalladacycle [PPFIP-Cl]2 almost no product was
detected (entry 22). The N-triflyl protected ferrocenyl
[22a]
monopalladacycle [FIP(Tf)-Cl]2
gave the desired product in
moderate yield (entry 23).
Improved Synthesis and X-Ray Crystal Structure Analysis of
[FIP-Cl]2
For the synthesis of the best precatalyst identified in this study –
[FIP-Cl]2 – our original procedure[22a] was slightly modified
resulting in an improved overall yield and a diastereomerically
pure catalyst (Scheme 1). Ferrocene carboxamide[33] was
converted into the imidazoline FI(H) by activation with
Meerwein’s salt and subsequent treatment with enantiomerically
pure (1R,2R)-1,2-diphenylethane-1,2-diamine. FI(H) was
characterised in the solid state by X-ray crystal structure
analysis (Figure 2).[34] The torsion angle between the C=N bond
and the substituted Cp ligand is relatively small in the solid state,
meaning that the basicity of the imidazoline unit is probably
increased by resonance with the electron-rich ferrocene core.
The less basic ligand FI(Ts) was obtained after N-tosylation in
an overall yield of 85% starting from ferrocene carboxamide.
In contrast to our previous studies, the direct cyclopalladation
could be achieved with full diastereoselectivity regarding the
introduced element of planar chirality. In the previous studies (dr
Figure 2. Representation of the X-ray crystal structure analysis of the
imidazoline FI(H). Only one of four molecules per unit cell, which slightly differ
in their conformations, is shown. Colour code: C (gray); N (blue); H (white); Fe
(orange).[34]
On the dimeric stage the dr determination by NMR is in principal
rendered difficult by the existence of geometrical isomers around
the Pd-Cl square-planes, which can in general result in up to 6
different species (Sp/Sp-cis, Sp/Sp-trans, Sp/Rp-cis, Sp/Rp-trans,
Rp/Rp-cis, Rp/Rp-trans). In our case only two of these isomers
This article is protected by copyright. All rights reserved.

10.1002/chem.201605244
Chemistry - A European Journal
FULL PAPER
were visible in a ratio of 2:1 indicating that Sp-configured
palladacycles were nearly exclusively formed.
boroxine is noteworthy (entry 4), because this type of substrate
is in general very challenging.[10] A similar result as with
phenylboroxine 3A (entry 1) was also achieved with para-
biphenylboroxine (3E, entry 5). Also less nucleophilic boroxines
3F-H carrying two chlorine or fluorine -acceptor atoms (entries
6 & 7) or an ester as -acceptor moiety (entry 8) still provided
useful yields of nearly enantiopure products. The heteroaromatic
extended -system of a 4-dibenzofuranyl moiety was also well
accommodated (entry 9). Not accepted were bromo atoms on
the boroxine, neither were they on the imine (not shown).
Ortho-, meta- and para-substituents were also well tolerated on
the aromatic imine component 1. Next to the model imine 1a
with a chloro substituent at the 4-position, chloro atoms at the 3-
and 2-positions also allowed for high yields of nearly
enantiopure products (imines 1b and 1c, entries 10-12 and 13-
15, respectively). In addition, reactivity differences caused by
electronic effects by substituents on R1 were found to be rather
small. A -acceptor substituent on the aromatic residue was thus
not necessary as shown for R1 = Ph (entries 16-19), but also a
-donor (entry 20), -donor (entries 21 & 22), -acceptor (entries
23-25) and a polycyclic aromatic hydrocarbon (entries 26-28)
were well accommodated.
To confirm the complete diastereoselectivity with regard to the
element of planar chirality, [FIP-Cl]2 was also transformed into
an isomerically pure monomeric complex with PPh3 or PMe3
(quantitative yields) according to a literature procedure.[22a]
For the Sp/Sp-trans-configured dimer crystals suitable for X-ray
structure analysis could be obtained (Figure 3).[35] Both Pd-
centers show the expected slightly distorted square-planar
coordination geometry in a nearly C2-symmetric solid state
molecular structure. Two Pd centers and the two bridging
chloride ligands form a puckered four-membered ring. The
dihedral angle in the Cl-Pd-Cl-Pd entity is around 25°. As a
result of the deviation from planarity, which is found for most
halide bridged ferrocenyl monopalladacycles,[36] both ferrocene
units are placed on the concave face of the four-membered ring.
The axes of both ferrocene moieties are thus tilted by ca. 60°
relative to each other. Both the Pd-C and Pd-N distances are
very similar to the bismetallacycle [FBIP-Cl]2.[25a] By the
puckering of the (PdCl)2-ring the Pd···Pd separation is about 4%
shorter than in the almost flat (PdCl)2-rings in [FBIP-Cl]2.
A more special behaviour was found for imine 1i carrying a 2-
thiophenyl residue R1. Whereas with phenylboroxine (3A) as
reaction partner a similarly high enantioselectivity was attained
(entry 29) as in all other cases described above, with the 4-
methoxylphenylboroxine (3B) and the para-biphenylboroxine
(3E) only moderate enantioselectivity was noticed (entries 30 &
31).
Noteworthy is also the fact that aliphatic imines 1j-l can be
employed. With α-branched alkyl residues slightly harsher
reaction conditions were required, but the products were still
formed in high yields and almost enantiopure form (entries
32&33). A particular challenge is the use of α-unbranched alkyl
residues due to the expected imine/enamine tautomerism. The
latter did not disturb the enantioselectivity, but side-reactions
and a product instability during isolation led to a moderate
product yield (entry 34).
Figure 3. Representation of the X-ray crystal structure analysis of the
precatalyst [FIP-Cl]2 possessing a trans geometry around the Pd-Cl-square-
planes. Solvent molecules and hydrogen atoms are omitted for clarity. Colour
code: C (gray); N (blue); O (red); S (yellow); Cl (green); Fe (orange); Pd
(magenta). H atoms and solvent molecules are omitted for clarity. [35]
From a practical point of view it is important that the arylation
reactions are also applicable to aldimines with N-sulfonyl
residues other than tosyl. For example, nearly enantiopure
products were formed with substrates carrying para-nosyl[37]
and dimethylaminosulfonyl groups[38] (entries 35 & 36). Both of
these groups are established as versatile N-protective groups
which can be readily removed under relatively mild reaction
conditions.
Due to the broad scope regarding the possible imine and
boroxine components, it is also possible to synthesise both
possible product enantiomers in almost enantiopure form, using
the same catalyst batch in an enantiodivergent manner by
simply exchanging the substitution patterns of imine and
boroxine. This is showcased by entries 16 and 21 (products 4dB
and 4fA), entries 18 and 26 (products 4dJ and 4hA) and entries
19 and 29 (products 4dK and 4iA)
Substrate Scope of the Arylation of Aldimines
The scope of the optimised reaction conditions was then
examined using different aldimines and boroxines (Table 2). All
reactions were conducted in chlorobenzene at 65 °C for 20 h
applying a precatalyst loading of 1 mol% if not mentioned
otherwise. They were performed at a shaking frequency of 250
rpm. In the large majority of examples nearly enantiomerically
pure products were formed with good to excellent yields. A
number of different boroxines 3A-I were investigated for the
reaction with imine 1a. Methoxy substituents at the boroxine
were well tolerated in the para (-donor), meta (-acceptor) and
ortho (-donor) positions (entries 2-4). In particular the high
productivity in the presence of the ortho-substituent on the
The absolute configurations of products 4aA and 4dB were
unambiguously determined by X-ray crystal structure analysis
confirming the attack at the Re-face of the imine.[39]
This article is protected by copyright. All rights reserved.

10.1002/chem.201605244
Chemistry - A European Journal
FULL PAPER
32
33
34
35
36
4jA
4kA
4lA
cyclo-Hex
i-Pr
Ph
Ph
Ph
Ph
Ph
pTs
pTs
pTs
pNs
DAS
95c,d)
92c,d)
32d,e)
91d)
>99
>99
99
Table 2. Application of the optimised reaction conditions to various substrates.
Ph(CH2)2
4-ClC6H4
4-ClC6H4
4aA’
4aA’’
>99
99
58d)
b)
c)
#
4
R1
R2
SO2R Yield
[%]a)
ee
a) Yield of isolated product. Determined by HPLC. Reaction temperature:
[%]b)
d)
e)
80 °C.
standard. (pTs
dimethylaminosulfonyl).
2 mol% of [FIP-Cl]2.
Determined by 1H-NMR using an internal
4-O2NC6H4SO2, DAS N,N-
=
4-MeC6H4SO2, pNs
=
=
1
4aA
4aB
4aC
4aD
4aE
4aF
4aG
4aH
4aI
4-ClC6H4
4-ClC6H4
4-ClC6H4
4-ClC6H4
4-ClC6H4
4-ClC6H4
4-ClC6H4
4-ClC6H4
4-ClC6H4
3-ClC6H4
3-ClC6H4
3-ClC6H4
2-ClC6H4
2-ClC6H4
2-ClC6H4
Ph
Ph
pTs
pTs
pTs
pTs
pTs
pTs
pTs
pTs
99
84
>99
>99
>99
>99
>99
99
2
4-MeOC6H4
3-MeOC6H4
2-MeOC6H4
4-PhC6H4
Mechanistic Studies and Considerations
a) The Activated Catalyst
3
95
4
96
As mentioned above, the precatalyst [FIP-Cl]2 is a chloride
bridged dimer and the best activity was achieved by catalyst
activation with silver acetate. Unfortunately, all attempts towards
direct cyclopalladation of the ligand FI(Ts) with Pd(OAc)2 failed.
In previous studies, the activation of our chloride bridged
palladacycle catalysts usually led to monomeric catalysts, when
the activation was performed in the presence of acetonitrile,
because the latter adopts the position trans to the imidazoline N-
donor.[22,25] In the present study, [FIP-Cl]2 was initially stirred for
4.5 h with AgOAc (2.0 equiv.) in acetonitrile at room temperature.
Later a more time-efficient activation procedure was used in
which the precatalyst/silver salt mixture was sonicated (160 W)
for few minutes. Inspection of the 1H-NMR spectra of the
activated catalyst batches revealed that they consisted of at
least two different species. The relative ratio of these two
species depended on how the activation was performed and had
an influence on the catalytic activity. For comparison, different
catalyst batches were investigated in the model arylation
reaction of 1a and 3A in chlorobenzene at 65 °C for 20 h using a
precatalyst loading of 0.5 mol%.
Under some catalyst activation conditions, a uniform catalyst
species could also be obtained. This was for instance the case,
when the activation was done in dry toluene in the absence of
MeCN (with and without sonication). However, this uniform
species, which could be assigned to the acetate bridged dimeric
structure [FIP-OAc]2 as described below, showed a relatively
low catalytic activity (34% yield). In contrast, when using the
standard protocol in MeCN, the second species – a monomer
(see below) – appeared and showed significantly broader
signals in the 1H-NMR spectra than the dimer.[40] Various
catalyst activation experiments at different temperatures (20 –
70 °C) confirmed that catalyst batches with a higher amount of
5
98
6
3,4-Cl2C6H3
3,4-F2C6H3
4-MeO2CC6H4
50
7
70c,d)
50d)
75c)
86
>99
>99
98
8
9
4-dibenzofuranyl pTs
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
4bA
4bB
4bE
4cA
4cB
4cE
4dB
4dE
4dJ
4dK
4eA
4fA
4fE
Ph
4-MeOC6H4
4-PhC6H4
Ph
pTs
pTs
pTs
pTs
pTs
pTs
pTs
pTs
pTs
pTs
pTs
pTs
pTs
pTs
pTs
pTs
pTs
pTs
pTs
pTs
pTs
pTs
99
99
>99
99
82
91
>99
96
4-MeOC6H4
4-PhC6H4
4-MeOC6H4
4-PhC6H4
2-naphthyl
thiophen-2-yl
Ph
85
85
99
98
99
Ph
99
>99
>99
90
Ph
86c),d)
99
Ph
4-MeC6H4
4-MeOC6H4
4-MeOC6H4
4-O2NC6H4
4-O2NC6H4
4-O2NC6H4
2-naphthyl
2-naphthyl
2-naphthyl
thiophen-2-yl
thiophen-2-yl
thiophen-2-yl
96
>99
>99
95
Ph
86
4-PhC6H4
Ph
82
4gA
4gB
4gE
4hA
4hB
4hE
4iA
63
99
4-MeOC6H4
4-PhC6H4
Ph
78
>99
>99
>99
99
85
the second species provided better catalytic activities.
A
relatively high amount of it could be obtained by ultrasound
activation at 35 °C (9 min, 160 W, Scheme 2). Under these
conditions a ratio dimeric/monomeric species = 1:1.6 was
determined and a yield of 73% was obtained (quantitative yield
with 1 mol% precatalyst). This activation procedure allowed for
high reproducibility under the standard conditions and was used
for the experiments described in Table 2.
91
4-MeOC6H4
4-PhC6H4
Ph
88
66
99
55c,d)
96
98
4iB
4-MeOC6H4
4-PhC6H4
81
4iE
57
77
This article is protected by copyright. All rights reserved.

10.1002/chem.201605244
Chemistry - A European Journal
FULL PAPER
acetonitrile complex must also be present, as it was detected by
ESI-MS, but it was not clearly visible in the standard 1H-NMR.
We hypothesised that the active second species might be the
monomeric complex FIP-OAc depicted in Figure 4. Catalyst
activation by sonication using higher temperatures might favour
the monomer entropically. Broader 1H-NMR signals as
compared to the dimer [FIP-OAc]2 might be due to dynamic
effects in which the acetate might behave as either monodentate
or bidentate ligand (coordination numbers 3 and 4 at Pd,
Scheme 2. Activation procedure of the precatalyst [FIP-Cl]2.
1
respectively, see Figure 6). For that reason, H-NMR spectra of
the same catalyst batch were recorded at temperatures between
–8 and +50 °C (500 MHz). The spectra at +40 and +50 °C led to
significantly sharper signals of the monomeric species. In
contrast, at 0 °C and below two different signals sets were
resolved for this species, displaying e.g. different signals for the
Cp-spectator ligand at 4.47 and 4.36 ppm. Moreover, at the low
temperature an additional minor catalyst species became visible
by a broad resonance at 3.84 ppm, which was usually too broad
to be detected in the temperature range of 20-50 °C. Addition of
acetonitrile to this sample led to an increase of the signal’s
intensity. 2D-NMR experiments showed that this {CH} moiety is
a spin system on its own. Therefore, it can be assigned to
another C5H5 spectator ligand. Our data suggest that this signal
belongs to the hidden monomeric acetonitrile complex FIP-
OAc/MeCN. This species is probably not the active catalyst,
because MeCN was found to inhibit the catalytic reaction, but
probably in rapid equilibrium with FIP-OAc.
Analysis of both the catalyst mixture and pure [FIP-OAc]2 by
ESI-MS did not detect the corresponding molecular ions, but two
major fragments. In these fragments the acetate ligands were
not bound anymore to the Pd-centers. One of the fragments is
the palladacycle which carries an additional acetonitrile ligand. It
thus seemed likely that the second monomeric catalyst species
1
that was detected above in the initial H-NMR studies, might be
the monomeric FIP-OAc/MeCN complex (Figure 4). In addition,
an ammonia adduct was found, which is formed under the
conditions of the ESI-MS experiments.
The presence of more than two catalyst species is also
supported by 13C-NMR, because three distinct C(2)-imidazoline
signals at 172.1, 171.7 ([FIP-OAc]2) and 171.5 ppm have been
resolved at room temperature. In addition three acetate
resonances were found at 177.8, 177.5 and 176.8 ppm.
Figure 4. Results and interpretation of ESI-MS experiments.
The presence of acetate in the activated catalyst could be
confirmed by negative mode ESI-MS experiments (found:
m/z = 59.01). In addition, HSQC-TOCSY NMR experiments of
the acetate bridged dimer showed cross-peaks for a 13C-NMR
1
signal at 23.0 ppm (CH3), a H-NMR signal at 0.88 ppm (CH3)
and a 13C-NMR signal at 177.5 ppm (C=O).
Figure 5. Varying temperature NMR experiments (500 MHz) activated
However, when the catalyst activation was repeated in CH2Cl2 in
the absence of any MeCN, the same two catalyst species as
described above (dimer and monomer) were surprisingly found
in the 1H-NMR. It thus can be excluded that the second species,
which apparently leads to a better catalytic activity, is an
acetonitrile complex. On the other hand, using the above
described standard activation protocol performed in MeCN, the
according to Scheme 2: 323 K (1), 313 K (2), 303 K (3), 273 K (4), 265 K (5).
In a recent study Richards et al. investigated the nature of
related ferrocenylphosphine palladacycles bearing acetate
ligands by IR spectroscopy.[41] A key conclusion was that these
complexes form the expected dimeric acetate bridged structure
in the solid state (IR-ATR, reported (CO) signals at 1571 and
This article is protected by copyright. All rights reserved.

10.1002/chem.201605244
Chemistry - A European Journal
FULL PAPER
1405 cm1), but in solution they exist as monomeric species
bearing bidentate acetate ligands (IR in CHCl3 solution, reported
(CO) signals at 1486 and 1448 cm1).[42,43] In our case we found
for the uniform [FIP-OAc]2 species in the solid state (CO)
signals at 1553 and 1451 cm1 (IR-ATR).[44] Nearly identical data
were obtained for the mixture of dimeric [FIP-OAc]2, monomeric
FIP-OAc and FIP-OAc/MeCN in the solid state after removal of
all solvent by carefully drying in high-vacuum (IR-ATR: (CO)
signals at 1552 and 1452 cm1). In contrast, the IR spectrum of a
solution in chlorobenzene showed, along these two (CO) bands
at 1555 and 1458 cm1, two additional bands at 1596 and 1476
cm1. The latter are both negligible for the solution of the uniform
[FIP-OAc]2 species in chlorobenzene. These data support the
presence of the monomeric species with bidentate and
monodentate acetate ligands next to the dimeric species in
solution for the active catalyst.[45]
To confirm that all present catalyst species possess after the
activation exclusively acetate as an anionic ligand, the mixture
was also treated with PPh3. By that monomeric FIP-OAc/PPh3
was obtained as the only product in a quantitative yield. A
hypothetical hydroxide containing species responsible for the
transmetallation of the arylboroxine is thus unlikely under
anhydrous conditions. (OH) signals were also not visible in the
recorded IR spectra. In contrast, the IR spectra of FIP-
OAc/PPh3 in chlorobenzene solution shows an intensive (C=O)
signal at 1603 cm1 confirming that the peak mentioned above at
1596 cm1 for the most active catalyst mixture belongs to the
monomeric species with a monodentate acetate ligand.
obtained from the ferrocenyl ligand. This is indicative that the
faster-than-NMR-time-scale association-dissociation process
([(C∩N)Pd(Cl)(PMe3)] ⇄ [[(C∩N)Pd(Cl)] + PMe3 does not occur
in solution (or at least is negligible). Such a rapid chemical
equilibrium would modify the apparent volume of the
corresponding phosphine-containing species, and thus prevent
comparison. Based on these data, it seems reasonable to
assume that the [FIP-OAc]2 species is indeed a dimer.
Since the active catalyst apparently consists of monomeric and
dimeric complexes, the catalytic model reaction of 1a and 3B
was also investigated for a possible non-linear effect under the
standard conditions of Table 2, entry 1.[48] The ee values of the
precatalyst batches used were adjusted in 10% steps by mixing
the corresponding [FIP-Cl]2 enantiomers with either (Sp)- or (Rp)-
configuration. All 11 catalyst samples investigated provided the
product in quantitative yields. In addition, an almost perfectly
linear correlation between the product ee values and the catalyst
ee values was found (for details see the Supporting Information).
This might indicate that homo- and heterochiral [FIP-OAc]2
dimers possess a similar stability/reactivity under the reaction
conditions with regard to the transformation to the catalytically
probably relevant monomeric FIP-OAc.
Scheme 3. Control experiment to confirm that all species of the activated
catalyst carry an acetate ligand.
The aggregation behaviour of the various catalyst species in
solution was further investigated by pulsed gradient spin echo
(PGSE) NMR experiments in CDCl3.[46] Using the Stokes-
Einstein equation and the DiffAtOnce software package,[47] the
self-diffusion constant D = 5.964·10-10 m2 s-1 grants access to
the hydrodynamic radius (rH = 6.35 Å) and the hydrodynamic
volume (VH = 1070.2 Å) of [FIP-OAc]2. By contrast, the
Figure 6. Composition of the activated catalyst and characteristic 1H-NMR
signals.
monomeric,
phosphine-containing
complex
FIP-Cl/PMe3
measured under the same conditions possesses a much higher
self-diffusion constant D = 7.458·10-10 m2 s-1, accounting for a
smaller hydrodynamic radius (rH = 5.07 Å) and a considerably
smaller hydrodynamic volume (VH = 547.2 Å) compared to [FIP-
OAc]2. It is noteworthy that in both cases the self-diffusion
coefficients and hydrodynamic radius / volume values obtained
from the doublet signal of the methyl of the PMe3 ligand (δ(1H) =
1.74 ppm, FIP-Cl/PMe3) show no deviation compared to those
b) The Catalytic Cycle
A mechanistic proposal for a possible catalytic cycle is shown in
Scheme 4. As MeCN has an inhibiting influence and the amount
of monomeric FIP-OAc has a significant impact on the catalytic
activity, we suggest that this is the catalytically active species.
This compound is supposed to undergo a transmetallation
reaction with the boroxine related to the so-called ’oxo-pathway’
in Suzuki-Miyaura reactions.[26] The Lewis-basic acetate is
This article is protected by copyright. All rights reserved.

10.1002/chem.201605244
Chemistry - A European Journal
FULL PAPER
expected in this ‘acetate-pathway’ to coordinate by its available
carbonyl group to a Lewis acidic boron center thus increasing
the nucleophilicity of the aromatic residue to be transferred to
the Pd-center.
Our spectroscopic data point to the product release as the rate
limiting step of this reaction. ESI-MS investigation of the reaction
mixture under standard reaction conditions and a reaction time
of 1 h identified the mass of the catalyst resting state
intermediate, which would fit to both II or III (found m/z:
1060.0725; calculated: 1060.0745). Differentiation between both
1
species was possible by H-NMR spectroscopy, which showed
the absence of the imino N=CH proton, whereas a new broad
signal is present at ca. 6.0 ppm which can be assigned to the N-
CH proton in intermediate III after the 1,2-addition has taken
place. After the reaction time of 20 h, nearly all catalyst has
been transformed into the product catalyst adduct (determined
by the use of an internal standard; next to the Cp signal also the
expected 5 characteristic catalyst protons of the substituted Cp
ring and the imidazoline core can be clearly resolved at chemical
shifts between 4 and 5.5 ppm), whereas after 1 h approximately
50% of catalyst is present in this state as judged from the C5H5
signals (for details see the Supporting Information). Additionally,
this data indicates that finally the dimeric [FIP-OAc]2 is also
transformed into the catalytically active form, but much slower
than the monomer explaining the differences in catalytic activity,
in particular using low loadings.
On the other hand this mechanistic picture suggests that the
transmetallation which was initially the bottleneck using
boroxines (see Table 1) is now an efficient elemental step in the
catalytic cycle.
The Arylation of Ketimines
The method development presented above was done with
aldimine substrates and hence for the generation of N-
substituted tertiary stereocenters. However, we were also
interested, if FIP-OAc is an efficient catalyst for ketimine
substrates to form N-substituted quaternary stereocenters as
well.
Scheme 4. Proposal of a simplified catalytic cycle including evidence by ESI-
MS data.
In 2010, Hayashi et al. reported
a Rh-catalysed highly
enantioselective arylation of N-sulfonylketimines with Ar4BNa.[49]
The reactions proceeded efficiently for arylmethyl ketimines as
well as indanone- and tetralone-derived imines, usually in the
presence of 5 mol% of Rh. Shortly thereafter, they reported a
modified version with improved atom-economy using potassium
organotrifluoroborates, for which also a high efficiency was
The success of the transmetallation of an aryl residue of a
boroxine to the activated catalyst bearing an acetate ligand in
the absence of an exogenous base could be demonstrated by
ESI-HRMS experiments. In these experiments a solution of the
activated catalyst (2 mol%) and boroxine 3A was stirred for 60
min in chlorobenzene at 65 °C and was then analysed mass-
spectrometrically. The prevailing Pd-containing species that was
detected was the Pd-Ph adduct FIP-Ph (found m/z: 742.0552;
calculated: 742.0579). The transmetallation suggests the
simultaneous formation of a B-OAc containing compound like 5
(the other Ar groups in 5 might later also be replaced).
Coordination of the imine is expected to occur preferentially
trans to the imidazoline N-donor based on the typical
coordination behaviour of C,N-palladacycles.[21,22b] The observed
almost exclusive Re-face attack at the imine could be explained
by a stereochemical model in which the imine’s sulfonyl group in
II points away from the ferrocenyl core in order to reduce
repulsive interactions. Product release might be achieved by
reaction of the catalyst/product-adduct III with the B-OAc
species 5. Thus the catalyst FIP-OAc would be regenerated.
demonstrated using
5
mol% Rh.[50] Triaryl-substituted
stereogenic centers could finally be created with high
enantioselectivity using boroxine reagents.[51]
The first Pd catalysed enantioselective arylation of ketimines
using an organoboron reagent was reported by Zhang et al. in
2013.[52] They utilised cyclic ketimines and arylboronic acids,
which very efficiently reacted in the presence of an in-situ
formed PdII catalyst (5 mol%) bearing
a neutral chiral
pyridine/oxazoline bidentate ligand. An initial issue of catalyst
decomposition by formation of Pd-black was addressed by
performing the reaction in the presence of O2. Moreover, by
portionwise addition of the boronic acid (10x) to minimise the
formation of Pd-black the catalyst loading could be reduced to 1
mol% still allowing for a high yield and enantioselectivity.
Soon after, Hayashi et al. reported a highly enantioselective
addition of arylboronic acids to less reactive six-membered
This article is protected by copyright. All rights reserved.

10.1002/chem.201605244
Chemistry - A European Journal
FULL PAPER
cyclic N-sulfonyl ketimines which provided synthetically versatile
sulfamidate products.[53,54] A neutral phosphinooxazoline ligand
was employed and usually 5-10 mol% of catalyst were used.
From a mechanistic standpoint there should be a couple of
differences in the reactions of either aldimines or ketimines. For
instance, in aldimines and cyclic ketimines the non-bonding
orbital at the Lewis-basic imino N atom can be located in a
different environment. In the former case it is preferentially
located cis to the aryl moiety of the imine, whereas in the latter
case it adopts the trans-position. Moreover, it appears that the
challenges using ketimines are somewhat shifted compared to
aldimines. Thus the nucleophilic addition step is naturally more
difficult as a result of larger steric hindrance in ketimines than in
aldimines and a usually lower electrophilicity. On the other side,
in case of ketimines it can be expected that the initial
catalyst/product adduct being generated in the 1,2-addition
elemental step is less prone to undergo a decomposition, since
a -hydride elimination cannot occur, as it is the case with
aldimines. In addition, the ketimine substrates are considerably
less sensitive towards hydrolysis than aldimines due to their
diminished electrophilicity. The use of aldimines and cyclic
ketimines in the arylation with organoboron reagents thus
creates a different scenario for the reaction development.
We initially chose the alkyl substituted ketimine 7a as model
substrate (Table 3, entry 1, R1 = nBu). Applying the reaction
conditions optimised for aldimines did not furnish any of the
desired product, using either 1 or 2 mol% of precatalyst both in
the absence and presence of NaOAc (1.0 equiv.) as additive.
Traces of product 8aA were found for a precatalyst amount of 5
mol% in the presence of NaOAc. An improvement of the
reactivity was found with KOAc. The product was obtained in a
yield of 23% and with an ee value of 96% under these conditions
(entry 2).
Table 3. Asymmetric arylation of cyclic ketimines 7.
#
8
R1
R2
x
T [°C] additive (equiv.) Yield ee
[%]a) [%]b)
1
2
3
4
5
6
7
8
9
8aA
8aA
8bB
8bB
8bB
8bB
8bB
8bB
nBu
nBu
Ph
Ph
5
5
5
5
5
5
1
65
65
50
40
50
40
50
NaOAc (1.0)
KOAc (1.0)
KOAc (1.0)
KOAc (1.0)
-
6
n.d.
96
Ph
23
99
99
86
51
4-MeOC6H4
4-MeOC6H4
4-MeOC6H4
4-MeOC6H4
4-MeOC6H4
95
Ph
94
Ph
34
51
Ph
-
Ph
KOAc (1.0)
KOAc (1.0)
KOAc (1.0)
KOAc (1.0)
KOAc (1.0)
>99 c) 95
80 92
Ph
4-MeOC6H4 0.5 50
8cB 3-MeOC6H4 4-MeOC6H4
1
1
1
50
50
50
75c) 91
98c) 94
45c),d) 82
10 8dB 4-MeC6H4 4-MeOC6H4
11 8eA
CO2Et
Ph
a)
Determined by 1H-NMR of the crude product using mesitylene as internal
b)
standard if not indicated otherwise. Determined by HPLC. A negative sign
indicates that the other enantiomer than the one depicted was formed (n.d.=
not determined). c) Yield of isolated product. d) 5 mol% of [FIP-Cl]2.
Since we assume that the C-H acidic alkyl chain R1 caused the
low yield, we examined substrate 7b bearing a Ph group instead
in the reaction with boroxine 3B. In that case the reactivity was
considerably higher. Both at 50 and 40 °C the product 8bB was
formed in nearly quantitative yields and with high
enantioselectivity (entries 3 & 4). A useful reactivity was even
noticed in the absence of the additional exogenous base (entries
5 & 6). Unexpectedly though, this resulted in a switch of the
preferentially formed enantiomer. Whereas in the presence of
KOAc a Si-face attack at imine 7b was largely dominating,[55] in
the absence of KOAc the Re-face of 7b was more reactive. In
that latter case the enantioselectivity also strongly depended on
the reaction temperature. At reaction temperatures of 40 and
50 °C the product was formed with ee values of 51% and 34%,
respectively. This enantiodivergency was not observed for
aldimine substrates.
The conditions of entry 7 were then applied to some other
substrates. Unfortunately, ketimine 7b did not undergo an
arylation with boroxines carrying electron-withdrawing groups
(not shown). In contrast, an electron withdrawing (entry 9), but
also an electron donating group (entry 10) on the ketimine
residue R1 was well tolerated. The use of the electron deficient
-iminoester 7e resulted in a lower enantioselectivity and
reactivity than in the other cases (ee = 82%, entry 12), maybe
for a different coordination behaviour of this substrate or the
formed product due to the possibility of chelate formation.[56]
The observation that the arylation of ketimines could not be
performed with electron-deficient boroxines indicates that either
the transmetallation or the 1,2-addition are the limiting steps in
the catalytic cycle. The fact that KOAc massively improves the
reactivity also seems to support a scenario, in which the
transmetallation is problematic. On the other hand, the
transmetallation already worked smoothly using the aldimine
substrates even in the absence of KOAc and using electron
deficient boroxines. As the same catalyst is used it thus seems
unlikely that with ketimines the transmetallation is rate-limiting.
We suggest that the 1,2-addition is the most difficult step in this
case. To explain the effect of KOAc on both the enantioface
differentiation and the reactivity we propose a mechanistic
scenario that is depicted in Scheme 5.
In the presence of KOAc, different catalyst loadings were
investigated at a reaction temperature of 50 °C (in analogy to
entry 3). Using 1 mol% of precatalyst, the product could be
isolated in a quantitative yield in highly enantioenriched form (ee
= 95%, entry 7). Also for just 0.5 mol% of precatalyst a good
yield and enantioselectivity were determined (entry 8).
This article is protected by copyright. All rights reserved.

10.1002/chem.201605244
Chemistry - A European Journal
FULL PAPER
commonly observed coordination modes in palladacycles (see
above).[21,22b] The Lewis-basic exogenous acetate is proposed to
promote this isomerisation via the formation of pentacoordinate
intermediate IV and acetate dissociation. The switch of the
coordination positions would result in
a switch of the
enantioselectivity. The isomerised complex is probably not only
thermodynamically preferred, but also more reactive explaining
the improved yields with KOAc. NaOAc is less efficient, as it is
less Lewis basic than KOAc due to its lower ionic character.[57]
Without exogenous acetate the enantioselectivity is moderate
due to a relatively slow isomerisation. Lower temperatures
further slow down the isomerisation explaining the considerably
higher preference for the unexpected enantiomer in the absence
of KOAc.
The question remains why no exogenous acetate is required for
high enantioselectivity with aldimines as the isomerisation would
also be necessary there. This might be rationalised by a higher
Lewis-basicity of the aldimines caused by a better steric
accessibility. The imine substrate might thus already play the
role reserved for acetate with ketimines.
Conclusions
In conclusion, we have reported a very efficient Pd catalyst for
the arylation of aldimines by an organoboron reagent. The
performance of this catalyst is on a comparable level with the
best Rh catalysts. Compared to cationic PdII catalysts bearing
neutral bidentate ligands the catalyst reported herein is
electronically different. It is a more electron-rich palladacycle
with an anionic C,N-ligand. By that the often found tendency to
form bisaryl homocoupling products could be successfully
eliminated. Using acetate as anionic ligand on the Pd-center is
of key importance for two reasons: a) in the presence of acetate
there was nearly no more imine hydrolysis observed and b) the
acetate apparently promotes the transmetallation from the
boroxine to the Pd-center, which was initially the limiting step
with boroxine reagents. Synthetically attractive results were
obtained for a very broad substrate scope. The element of
planar chirality of the ferrocenyl ligand was probably responsible
for the remarkably high level of enantioselectivity for most
examples. Mechanistic studies suggest that a monomeric
ferrocenyl palladacycle acetate complex is the catalytically
relevant species and that the product release is rate limiting. The
catalyst/product-adduct is thus the resting state in the catalytic
cycle. Using ketimine substrates, the presence of KOAc as
exogenous base was necessary for an optimal reactivity and
Scheme 5. Possible explanation for the observed enantiodivergency in the
absence/presence of KOAc.
In this scenario, KOAc is responsible for a smooth isomerisation
necessary to obtain high enantioselectivity. We suggest that the
acetate triggered transmetallation starting from FIP-OAc is a
concerted process which initially places the aryl moiety R2 in
trans-position to the imidazoline N-donor, whereas the cis-
position would be available for the ketimine coordination. Like for
aldimine substrates, we think that the imine coordination should
proceed in a way that the sterically demanding sulfonyl residue
points away from the ferrocene core to minimise repulsive
interactions. This kinetically preferred coordination mode would
result in the (R)-configured product by a Re-face attack. On the
other hand, it is expected that that a switch of the positions of
the anionic aryl residue and the neutral ketimine substrate would
be thermodynamically preferred in agreement with the
enantioselectivity.
A
surprising enantiodivergency in the
absence/presence of this additive might be explained by an
isomerisation of the catalytic intermediate ready to undergo the
1,2-addition. With ketimines, the arylation step itself is probably
the slowest step, explaining why electron-deficient boroxines
were not applicable. This is in contrast to the study with
aldimines. Whereas the electron rich nature of the catalyst thus
limits the scope with ketimines, it is crucial for the high efficiency
with aldimines.
This article is protected by copyright. All rights reserved.

10.1002/chem.201605244
Chemistry - A European Journal
FULL PAPER
[2]
[3]
C. A. Busacca, D. R. Fandrick, J. J. Song, C. H. Senanayake, Adv.
Synth. Catal. 2011, 353, 1825.
Experimental Section
A. J. Burke, C. S. Marques, Catalytic Arylation Methods - From the
Academic Lab to Industrial Process, WILEY-VCH GmbH & Co. KGaA,
Weinheim, 2015.
General Procedure for the Activation of [FIP-Cl]2
Method A: [FIP-Cl]2 (1 equiv., 11.58 µmol, 16.2 mg) and silver acetate (2
equiv., 23.16 µmol, 3.8 mg) were suspended in acetonitrile (2 mL) under
N2 atmosphere and stirred in the absence of light for 4.5 h at room
temperature. The mixture was then filtered through a syringe filter (PTFE)
and the solvent was removed under a permanent nitrogen flow. A stock
solution was prepared (c = 11.58 µmol/2 mL chlorobenzene) and used
directly for catalysis.
[4]
[5]
D. A. Pflum, D. Krishnamurthy, Z. Han, S. A. Wald, C. H. Senanayake,
Tetrahedron Lett. 2002, 43, 923.
W. M. Welch, A. R. Kraska, R. Sarges, K. B. Koe, J. Med. Chem. 1984,
27, 1508.
[6]
[7]
A. J. Wagstaff, D. Ormrod, C. M. Spencer, CNS Drugs 2001, 15, 231.
F. Stocchi, C. Fossati, M. Torti, Expert Opin. Pharmacother. 2015, 16,
2231.
[8]
[9]
J. Corey-Bloom, R. Anand, J. Veach, Int. J. Geriatr. Psychopharmacol.
1998, 1, 55.
Method B: [FIP-Cl]2 (1 equiv., 11.58 µmol, 16.2 mg) and silver acetate (2
equiv., 23.16 µmol, 3.8 mg) were suspended in acetonitrile (2 mL) under
N2 atmosphere and treated with ultra-sound (160 W) for 9 min at 35 °C.
The mixture was then filtered through a syringe filter (PTFE) and the
solvent was removed under a permanent nitrogen flow. A stock solution
was prepared (c = 11.58 µmol/2 mL chlorobenzene) and used directly for
catalysis.
K.-J. Chang, G. C. Rigdon, J. L. Howard, R. W. McNutt, J. Pharmacol.
Exp. Ther. 1993, 267, 852.
[10] C. S. Marques, A. J. Burke, ChemCatChem 2011, 3, 635.
[11] a) N. Hermanns, S. Dahmen, C. Bolm, S. Bräse, Angew. Chem. Int. Ed.
2002, 41, 3692; Angew. Chem. 2002, 114, 3844; b) M. Shi, W. Zhang,
Tetrahedron: Asymmetry 2003, 14, 3407; c) T. Nishimura, Y. Yasuhara,
T. Hayashi, Org. Lett. 2006, 8, 979.
General Procedure for the Catalytic Asymmetric Arylation of N-
Sulfonyl-Aldimines with Arylboroxines
[12] T. Hayashi, M. Kawai, N. Tokunaga, Angew. Chem. Int. Ed. 2004, 43,
6125; Angew. Chem. 2004, 116, 6251.
[13] M. Hayashi, M. Ishigedani, J. Am. Chem. Soc. 2000, 122, 976.
[14] Selected examples with aldimines: a) N. Kuriyama, T. Soeta, X. Hao, Q.
Chen, K. Tomioka, J. Am. Chem. Soc. 2004, 126, 8128; b) N.
Tokunaga, Y. Otomaru, K. Okamoto, K. Ueyama, R. Shintani, T.
Hayashi, J. Am. Chem. Soc. 2004, 126, 13584; c) R. B. C. Jagt, P. Y.
Toullec, D. Geerdink, J. G. de Vries, B. L. Feringa, A. J. Minnaard,
Angew. Chem. Int. Ed. 2006, 45, 2789; Angew. Chem. 2006, 118,
2855; d) Z.-Q. Wang, C.-G. Feng, M.-H. Xu, G.-Q. Lin, J. Am. Chem.
Soc. 2007, 129, 5336; e) H. Nakagawa, J. C. Rech, R. W. Sindelar, J. A.
Ellman, Org. Lett. 2007, 9, 5155; f) R. Shintani, R. Narui, Y. Tsutsumi,
S. Hayashi, T. Hayashi, Chem. Commun. 2011, 47, 6123; g) T.
Korenaga, A. Ko, K. Uotani, Y. Tanaka, T. Sakai, Angew. Chem. Int. Ed.
2011, 50, 10703; Angew. Chem. 2011, 123, 10891; h) R. H. Crampton,
M. Fox, S. Woodward, Tetrahedron: Asymmetry 2013, 24, 599; i) G.-Z.
Zhao, G. Sipos, A. Salvador, A. Ou, P. Gao, B. W. Skelton, R. Dorta,
Adv. Synth. Catal. 2016, 358, 1759; j) T. Yasukawa, T. Kuremoto, H.
Miyamura, S. Kobayashi, Org. Lett. 2016, 18, 2716; k) quinolinium
salts: Y. Wang, Y. Liu, D. Zhang, H. Wei, M. Shi, F. Wang, Angew.
Chem. Int. Ed. 2016, 55, 3776.
To the corresponding N-tosylaldimine 1 (1 equiv., 0.058 mmol) and the
corresponding boroxine
3 (1 equiv., 0.058 mmol) was added the
activated catalyst (prepared from 1 mol% [FIP-Cl]2, see above) in
chlorobenzene (c = 0.29 mol imine/L) under N2 atmosphere. The mixture
was heated to 65 °C in a synthesizer shaking at 250 rpm. After 20 h the
reaction mixture was cooled to room temperature and applied to silica gel
column chromatography (petrol ether/ethyl acetate (10:1 + 1% NEt3 to
4:1 + 1% NEt3)).
General Procedure for the Catalytic Asymmetric Arylation of Cyclic
Ketimines with Arylboroxines
To the corresponding N-sulfonyl ketimine 7 (1 equiv., 0.058 mmol), the
corresponding boroxine 3 (1 equiv., 0.058 mmol) and potassium acetate
(1 equiv., 0.058 mmol) was added the activated catalyst (prepared from
1 mol% [FIP-Cl]2, see GP1) in chlorobenzene (c = 0.29 mol imine/L). The
mixture was heated to 50 °C at 250 rpm. After 20 h the reaction mixture
was cooled to room temperature, washed with aqueous HCl (1 equiv.)
and dried over MgSO4. The solvent was removed under reduced
pressure and the residue was purified by silica gel column
chromatography (petrol ether/ethyl acetate (10:1 to 5:1)).
[15] For an early study of a non-enantioselective reaction, see: M. Ueda, N.
Miyaura, J. Organomet. Chem. 2000, 595, 31.
[16] a) H. Dai, M. Yang, X. Lu, Adv. Synth. Catal. 2008, 350, 249; b) G.-N.
Ma, T. Zhang, M. Shi, Org. Lett. 2009, 11, 875; c) H. Dai, X. Lu,
Tetrahedron Lett. 2009, 50, 3478; d) Z. Liu, M. Shi, Tetrahedron 2010,
66, 2619; e) C. S. Marques, A. J. Burke, Eur. J. Org. Chem. 2010,
1639; f) J. Chen, X. Lu, W. Lou, Y. Ye, H. Jiang, W. Zeng, J. Org.
Chem. 2012, 77, 8541; g) X. Gao, B. Wu, Z. Yan, Y.-G. Zhou, Org.
Biomol. Chem. 2016, 14, 55; h) T. Johnson, B. Luo, M. Lautens, J. Org.
Chem. 2016, 81, 4923; i) T. Beisel, A. M. Diehl, G. Manolikakes, Org.
Lett. 2016, 18, 4116.
Acknowledgements
This work was financially supported by the Deutsche
Forschungsgemeinschaft (DFG, PE 818/4-1 and PE 818/4-2).
We thank Mr. Andreas Schneider for initial experiments with
ketimine substrates and M.Sc. Che-Hung Lin (Plietker group,
Universität Stuttgart) for recording IR spectra of activated
catalyst in solution.
[17] For non-enantioselective methods, see: a) P. He, Y. Lu, Q.-S. Hu,
Tetrahedron Lett. 2007, 48, 5283; b) F. J. Williams, E. R. Jarvo, Angew.
Chem. Int. Ed. 2011, 50, 4459; Angew. Chem. 2011, 123, 4551.
[18] There is also one report using a Ru-catalyst: C. S. Marques, A. J. Burke,
Eur. J. Org. Chem. 2012, 4232.
[19] a) M. Miura, M. Nomura, Top. Curr. Chem. 2002, 219, 211; b) D. A.
Culkin, J. F. Hartwig, Acc. Chem. Res. 2003, 36, 234; c) R. Jana, T. P.
Pathak, M. S. Sigman, Chem. Rev. 2011, 111, 1417.
Keywords: 1,2-additions • boronic acids • ferrocene • imine •
transmetallation
[20] Q. Zhang, J. Chen, M. Liu, H. Wu, J. Cheng, C. Qin, W. Su, J. Ding,
Synlett 2008, 935.
[1]
D. Ghislieri, N. J. Turner, Top. Catal. 2014, 57, 284.
[21] a) J. Dupont, M. Pfeffer, Palladacycles; Wiley-VCH: Weinheim,
Germany, 2008; b) C. J. Richards, in Chiral Ferrocenes in Asymmetric
This article is protected by copyright. All rights reserved.

10.1002/chem.201605244
Chemistry - A European Journal
FULL PAPER
Catalysis, (Eds: L.-X. Dai and X.-L. Hou,), Wiley-VCH, Weinheim, 2010,
pp. 337-368; c) H. Nomura, C. J. Richards, Chem. Asian J. 2010, 5,
1726; d) R. Peters, D. F. Fischer, S. Jautze, Top. Organomet. Chem.
2011, 33, 139.
Angew. Chem. Int. Ed. 2013, 52, 13223; Angew. Chem. 2013, 125,
13465.
[26] Excellent review about the transmetallation mechanism of boronic acids
or derivatives in Suzuki-Miyaura-reactions: A. J. J. Lennox, G. C. Lloyd-
Jones, Angew. Chem. Int. Ed. 2013, 52, 7362; Angew. Chem. 2013,
125, 7506.
[22] Selected studies using ferrocenyl imidazoline monopalladacycles: a) M.
E. Weiss, D. F. Fischer, Z.-q. Xin, S. Jautze, W. B. Schweizer, R.
Peters, Angew. Chem. Int. Ed. 2006, 45, 5694; Angew. Chem. 2006,
118, 5823; b) D. F. Fischer, A. Barakat, Z.-q. Xin, M. E. Weiss, R.
Peters, Chem. Eur. J. 2009, 15, 8722; c) R. Peters, Z.-q. Xin, F. Maier,
Chem. Asian J. 2010, 5, 1770; d) S. H. Eitel, M. Bauer, D. Schweinfurth,
N. Deibel, B. Sarkar, H. Kelm, H.-J. Krüger, W. Frey, R. Peters, J. Am.
Chem. Soc. 2012, 134, 4683; e) M. Weber, R. Peters, J. Org. Chem.
2012, 77, 10846; f) S. H. Eitel, S. Jautze, W. Frey, R. Peters, Chem.
Sci. 2013, 4, 2218; g) J. M. Bauer, W. Frey, R. Peters, Angew. Chem.
Int. Ed. 2014, 53, 7634; Angew. Chem. 2014, 126, 7764; h) T. Hellmuth,
W. Frey, R. Peters, Angew. Chem. Int. Ed. 2015, 54, 2788; Angew.
Chem. 2015, 127, 2829; selected studies on imidazoline pincer
complexes: i) K. Hyodo, S. Nakamura, N. Shibata, Angew. Chem. Int.
Ed. 2012, 51, 10337; j) H. Huang, R. Peters, Angew. Chem. Int. Ed.
2009, 48, 604; k) M. Ohara, S. Nakamura, N. Shibata, Adv. Synth.
Catal. 2011, 353, 3285; l) K. Hyodo, S. Nakamura, K. Tsuji, T. Ogawa,
Y. Funahashi, N. Shibata, Adv. Synth. Catal. 2011, 353, 3385; m) S.
Nakamura, K. Hyodo, M. Nakamura, D. Nakane, H. Masuda, Chem.
Eur. J. 2013, 19, 7304; n) K. Hyodo, M. Kondo, Y. Funahashi, S.
Nakamura, Chem. Eur. J. 2013, 19, 4128; o) M. Kondo, T. Nishi, T.
Hatanaka, Y. Funahashi, S. Nakamura, Angew. Chem. Int. Ed. 2015,
54, 8198; p) M. Kondo, N. Kobayashi, T. Hatanaka, Y. Funahashi, S.
Nakamura, Chem. Eur. J. 2015, 21, 9066; q) M. Kondo, T. Nishi, T.
Hatanaka, Y. Funahashi, S. Nakamura, Angew. Chem. Int. Ed. 2015,
54, 8198; r) M. Kondo, M. Sugimoto, S. Nakamura Chem. Commun.
2016, 52, 13604.
[27] The use of lower catalyst amounts was also studied. With 0.1 mol%
[FIP-Cl]2 a yield of 93% of nearly enantiopure (>99% ee) product 3aA
was obtained, but reproducibility was problematic with such very low
loadings.
[28] The different commercial suppliers were not able to overcome this
purity problem after we informed them about this issue.
[29] During the course of these studies a transmetallation product has been
characterised, which was generated from a Pd-OAc species: A. R.
Kapdi, G. Dhangar, J. L. Serrano, J. Pérez, L. Garcia, I. J. S. Fairlamb,
Chem. Commun. 2014, 50, 9859.
[30] A. R. Kapdi, G. Dhangar, J. L. Serrano, J. Pérez, L. Garcia, I. J. S.
Fairlamb, Chem. Commun. 2014, 50, 9859.
[31] M. Weiss, W. Frey, R. Peters, Organometallics 2012, 31, 6365.
[32] T. Hellmuth, S. Rieckhoff, M. Weiss, K. Dorst, W. Frey, R. Peters, ACS
Catal. 2014, 4, 1850.
[33] For ferrocenecarboxamide the solid state structure was also analysed.
CCDC-1463864 contains the supplementary crystallographic data.
These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
[34] CCDC-1463866 contains the supplementary crystallographic data for
compound FI(H). These data can be obtained free of charge from The
Cambridge
Crystallographic
Data
Centre
via
www.ccdc.cam.ac.uk/data_request/cif.
[35] CCDC-1463867 contains the supplementary crystallographic data for
compound [FIP-Cl]2. In addition CCDC-1463863 contains the
supplementary crystallographic data for the enantiomeric compound
ent-[FIP-Cl]2 These data can be obtained free of charge from The
[23] Selected alternative addition reactions of arylboronic acids (or
derivatives) to non-imine electrophiles: a) J. Wong, K. Gan, H. J. Chen,
S. A. Pullarkat, Adv. Synth. Catal. 2014, 356, 3391; b) A. L.
Gottumukkala, J. Suljagic, K. Matcha, J. G. de Vries, A. J. Minnaard,
ChemSusChem 2013, 6, 1636; c) A. L. Gottumukkala, K. Matcha, M.
Lutz, J. G. de Vries, A. J. Minnaard, Chem. Eur. J. 2012, 18, 6907; d)
Y. Suzuma, S. Hayashi, T. Yamamoto, Y. Oe, T. Ohta, Y. Ito,
Tetrahedron: Asymmetry 2009, 20, 2751; e) Q. Zhang, J. Chen, M. Liu,
H. Wu, J. Cheng, C. Qin, W. Su, J. Ding, Synlett 2008, 935; f) T.-K.
Zhang, D.-L. Mo, L.-X. Dai, X.-L. Hou, Org. Lett. 2008, 10, 3689; g) Y.-
X. Liao, C.-H. Xing, P. He, Q.-S. Hu, Org. Lett. 2008, 10, 2509; h) P. He,
Y. Lu, C.-G. Dong, Q.-S. Hu, Org. Lett. 2007, 9, 343; i) T. Yamamoto,
T. Ohta, Y. Ito, Org. Lett. 2005, 7, 4153; j) P. He, Y. Lu, Q.-S. Hu,
Tetrahedron Lett. 2007, 48, 5283; k) Y. Suzuma, T. Yamamoto, T. Ohta,
Y. Ito, Chem. Lett. 2007, 36, 470; l) F. Gini, B. Hessen, A. J. Minnaard,
Org. Lett. 2005, 7, 5309; m) T. Nishikata, Y. Yamamoto, N. Miyaura,
Angew. Chem. Int. Ed. 2003, 42, 2768; Angew. Chem. 2003, 115,
2874.
Cambridge
Crystallographic
Data
Centre
via
www.ccdc.cam.ac.uk/data_request/cif.
[36] See e.g. C. E. Anderson, Y. Donde, C. J. Douglas, L. E. Overman, J.
Org. Chem. 2005, 70, 648 and cited references.
[37] P. G. M. Wuts, T. W. Greene, Greene's Protective Groups in Organic
Synthesis, Wiley-Interscience; 4th ed., 2006.
[38] a) R. B. C. Jagt, P. Y. Toullec, D. Geerdink, J. G. deVries, B. L.
Feringa, A. J. Minnaard, Angew. Chem. Int. Ed. 2006, 45, 2789;
Angew. Chem. 2006, 118, 2855; b) J. M. Bauer, W. Frey, R. Peters,
Chem. Eur. J. 2016, 22, 5767.
[39] CCDC-1050456
(4aA)
and
CCDC-1050457
(4dB)
contain
supplementary crystallographic data. These data can be obtained free
of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
[40] Catalyst activation with TlOAc (2 equiv.) instead resulted in the same
catalyst species with comparable activity. This indicated that Ag(I) ions
should not play a significant role in the activated catalyst and that redox
processes like formation of Pd(III) or Fe(III) complexes as catalytically
active species can be ruled out.
[24] Preliminary results were published as a communication: C. Schrapel, R.
Peters, Angew. Chem. Int. Ed. 2015, 54, 10289; Angew. Chem. 2015,
127, 10428.
[25] Selected studies using ferrocenyl bisimidazoline bispalladacycles: a) S.
Jautze, P. Seiler, R. Peters, Angew. Chem. Int. Ed. 2007, 46, 1260;
Angew. Chem. 2007, 119, 1282; b) S. Jautze, R. Peters, Angew. Chem.
Int. Ed. 2008, 47, 9284; Angew. Chem. 2008, 120, 9424; c) S. Jautze,
S. Diethelm, W. Frey, R. Peters, Organometallics 2009, 28, 2001; d) M.
Weber, S. Jautze, W. Frey, R. Peters, J. Am. Chem. Soc. 2010, 132,
12222; e) M. Weber, W. Frey, R. Peters, Adv. Synth. Catal. 2012, 354,
1443; f) M. Weber, S. Jautze, W. Frey, R. Peters, Chem. Eur. J. 2012,
18, 14792; g) S. H. Eitel, S. Jautze, W. Frey, R. Peters, Chem. Sci.
2013, 4, 2218; h) M. Weber, W. Frey, R. Peters, Chem. Eur. J. 2013,
19, 8342; i) M. Weber, J. E. M. N. Klein, B. Miehlich, W. Frey, R. Peters,
Organometallics 2013, 32, 5810; j) M. Weber, W. Frey, R. Peters,
[41] K. Panchal, J. Amin, F. X. Roca, M. Motevalli, P. N. Horton, S. J. Coles,
C. J. Richards, J. Organomet. Chem. 2015, 775, 12.
[42] A dynamic behaviour which was assigned to an interconversion of
dimeric and monomeric C,P-palladacycle complexes bearing weakly
coordinating acetate ligands has previously been reported: a) A. J.
Cheney, B. L. Shaw, J. Chem. Soc. Dalton Trans. 1972, 860; b) W. A.
Herrmann, C. Brossmer, C.-P. Reisinger, T. H. Riermeier, K. Öfele, M.
Beller, Chem. Eur. J. 1997, 3, 1357; c) B. L. Shaw, S. D. Perera, E. A.
Staley, Chem. Commun. 1998, 1361.
[43] K. Nakamoto, Infrared and Raman Spectra of Inorganic and
Coordination Compounds, 3rd ed., Wiley-Interscience, New York, 1978.
1
[44] For [FIP-Cl]2 a signal at 1542 cm is also detected (IR-ATR), but the
signal intensity is much lower than for the (CO) signals. This signal
This article is protected by copyright. All rights reserved.

10.1002/chem.201605244
Chemistry - A European Journal
FULL PAPER
can be assigned to the C=N bond in the imidazoline similar to the free
ligand (ref. 22a). The C=N bands in the activated catalyst species are
probably hidden by the much more intensive acetate signals.
[50] R. Shintani, M. Takeda, Y. T. Soh, T. Ito, T. Hayashi, Org. Lett. 2011,
13, 2977.
[51] a) T. Nishimura, A. Noishiki, G. Chit Tsui, T. Hayashi, J. Am. Chem.
Soc. 2012, 134, 5056; b) see also: T. Nishimura, A. Noishiki, Y. Ebe, T.
Hayashi, Angew. Chem. Int. Ed. 2013, 52, 1777; Angew. Chem. 2013,
125, 1821.
[45] Unfortunately, all trials to get single crystal of the activated catalyst
suitable for X-ray analysis have failed so far, often as a result of a slow
catalyst decomposition. As one decomposition product the compound
.
[Fe3O(OAc)7 H2O] was identified, which consist of [Fe3(µ3-O)(µ-
[52] a) G. Yang, W. Zhang, Angew. Chem. Int. Ed. 2013, 52, 7540; Angew.
Chem. 2013, 125, 7688; see also: b) Q. He, L. Wu, X. Kou, N. Butt, G.
Yang, W. Zhang, Org. Lett. 2016, 18, 288; c) M. Quan, G. Yang, F. Xie,
I. D. Gridnev, W. Zhang, Org. Chem. Front. 2015, 2, 398.
[53] C. Jiang, Y. Lu, T. Hayashi, Angew. Chem. Int. Ed. 2014, 53, 9936;
Angew. Chem. 2014, 126, 10094.
OAc)6]clusters that are linearly connected by acetate ions (see the
Supporting Information). CCDC-1463868 contains the supplementary
crystallographic data for this compound. These data can be obtained
free of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif. A related material forming 3D-
networks was recently reported: J.-P. Zhao, S.-D. Han, X. Jiang, J. Xu,
Z. Chang, X.-H. Bu, Chem. Commun. 2015, 51, 4627. For the
molecular structure of the cluster see also: A. R. Burgoyne, R.
Meijboom, A. Muller, B. O. Omondi, Acta Cryst. 2011, 67, 1092.
[46] The PGSE experiments were run as three sets of 12 PGSE
experiments as described in T. Augenstein, F. Dorner, K. Reiter, H. E.
Wagner, D. Garnier, W. Klopper, F. Breher, Chem. Eur. J. 2016, 22,
7935. For details see the supporting information.
[54] Very recently,
a
Pd-catalyzed enantioselective arylation of
trifluoromethylated/perfluoroalkylated 2-quinazolinones could also be
achieved: B. Zhou, C. Jiang, V. R. Gandi, Y. Lu, T. Hayashi, Chem. Eur.
J. 2016, 22, 13068.
[55] The absolute configurations were assigned by comparison to literature
data reported for 8aA (see ref. 52) and for 8bB (see ref. 51a).
[56] Very recently, a Pd-catalyzed enantioselective arylation of cyclic α-
ketiminophosphonates with arylboronic acids could also be achieved: Z.
Yan, B. Wu, X. Gao, Y.-G. Zhou, Chem. Commun. 2016, 52, 10882.
[57] R. S. Sah, B. Sinha, M. N. Roy, Fluid Phase Equilib. 2011, 307, 216.
[47] DiffAtOnce Molecular Diffusion, available at www.diffatonce.com.
[48] T. Satyanarayana, S. Abraham, H. B. Kagan, Angew. Chem. Int. Ed.
2009, 48, 456; Angew. Chem. 2009, 121, 464.
[49] R. Shintani, M. Takeda, T. Tsuji, T. Hayashi, J. Am. Chem. Soc. 2010,
132, 13168.
This article is protected by copyright. All rights reserved.

10.1002/chem.201605244
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents
FULL PAPER
Carmen Schrapel, Wolfgang Frey,
Delphine Garnier, and René Peters*
Page No. – Page No.
Highly Enantioselective Ferrocenyl
Palladacycle-Acetate Catalysed
Arylation of Aldimines and Ketimines
with Arylboroxines
A planar chiral ferrocenyl palladacycle acetate complex catalyzes the 1,2-addition
of arylboroxines to N-sulfonyl aldimines and cyclic ketimines with high
enantioselectivity. Spectroscopic studies suggest that (1) a monomeric species is
the active catalyst, (2) the acetate ligand promotes the transmetallation, (3) the
product release is rate-limiting for aldimines. With ketimines additional acetate is
necessary for high enantioselectivity causing a switch of the preferred enantiomer.
This article is protected by copyright. All rights reserved.