(N,N) vs. (N,S) chelation of palladium in asymmetric allylic substitution using bis(thiazoline) ligands: A theoretical and experimental study

Article abstract of DOI:10.1016/j.jorganchem.2008.04.044

New thiazoline-containing ligands including non-symmetric bis(thiazolines) and oxazoline-thiazolines were synthesized and then compared to C₂-symmetric bis(thiazolines) in the palladium-catalyzed allylic substitution. The experimental results obtained in this study support the hypothesis of a competition between the (N,N) and the (N,S) palladium chelation, when sterically hindered bis(thiazolines) are used as ligands. A quantum chemical study performed on the Pd-complexes derived from three selected ligands, two C₂-symmetric bis(thiazolines) and one oxazoline-thiazoline, also supports this hypothesis.

Full text of DOI:10.1016/j.jorganchem.2008.04.044

Journal of Organometallic Chemistry 693 (2008) 2499–2508
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
(N,N) vs. (N,S) chelation of palladium in asymmetric allylic substitution using
bis(thiazoline) ligands: A theoretical and experimental study
a
c
a,
*
Alexander Betz a, Lian Yu b, Markus Reiher b, , Annie-Claude Gaumont , Paul-Alain Jaffrès , Mihaela Gulea
*
a Laboratoire de Chimie Moléculaire et Thio-Organique (LCMT), ENSICAEN, Université de Caen Basse-Normandie, CNRS 6 Boulevard du Marechal Juin, 14050 Caen, France
b Laboratorium für Physikalische Chemie, ETH Zurich, Wolfgang-Pauli-Str. 10, CH-8093 Zurich, Switzerland
c CEMCA, UMR CNRS 6521, Faculté des Sciences et Techniques, Université de Bretagne Occidentale, 6 Av. Victor Le Gorgeu, 29238 Brest, France
a r t i c l e i n f o
a b s t r a c t
Article history:
New thiazoline-containing ligands including non-symmetric bis(thiazolines) and oxazoline–thiazolines
were synthesized and then compared to C2-symmetric bis(thiazolines) in the palladium-catalyzed allylic
substitution. The experimental results obtained in this study support the hypothesis of a competition
between the (N,N) and the (N,S) palladium chelation, when sterically hindered bis(thiazolines) are used
as ligands. A quantum chemical study performed on the Pd-complexes derived from three selected
ligands, two C2-symmetric bis(thiazolines) and one oxazoline–thiazoline, also supports this hypothesis.
Ó 2008 Elsevier B.V. All rights reserved.
Received 21 March 2008
Received in revised form 18 April 2008
Accepted 18 April 2008
Available online 6 May 2008
Keywords:
Non-symmetric bis(thiazoline) ligands
Oxazoline–thiazoline ligands
Asymmetric allylic substitution
Palladium complexes
Quantum chemical calculations
1. Introduction
catalysis to test the efficiency of new ligands. The results were
compared with those obtained with the oxazoline counterparts
Compared to the widely applied oxazoline ligands in asymmet-
ric catalysis [1], their sulfur analogues, the thiazolines, have been
rarely used for this purpose. The first study dealing with chiral
thiazoline ligands was reported by Helmchen in 1991 and found
a different behaviour compared to oxazolines in the rhodium cat-
alyzed asymmetric hydrosilylation of acetophenone [2]. In fact,
the real interest in thiazoline ligands only appeared ten years later.
Since then, their structures and applications have been regularly
renewed [3,4]. In some studies, thiazoline ligands have been eval-
uated and compared to the corresponding oxazolines in well
known metal-catalyzed asymmetric reactions [2,3b,3c,4]. As ex-
pected, different results have been noticed for the two types of li-
gands, due to electronic and steric effects resulting from the
substitution of the oxygen atom by a sulfur in the heterocycle
[5]. Thus, further investigations appear necessary for a favourable
development of thiazoline ligands, shedding light on the behaviour
of the thiazoline heterocycle towards the metal chelation.
and important differences in the behaviour of the two types of li-
gands were found in some cases. Moreover, unexpected results
were obtained in the bis(thiazoline) series (Scheme 1): (i) the
enantiomeric excess decreases with the increase of the steric hin-
drance of the substituent in 4-position of the heterocycle [ee: 88%
(Et) > 42% (iPr) > 31% (tBu)]; (ii) while bis(thiazolines) (R,R)-1a and
(R,R)-1b led to the major enantiomer (R) of the product, bis(thiaz-
oline) (S,S)-1c, which was expected to give major (S)-enantiomer,
led to the opposite enantiomer (R).
To explain these results, a competition between nitrogen and
sulfur in the palladium chelation was proposed in our previous pa-
per [3b] (Scheme 2). For the (R,R)-bis(thiazoline) 1a, the reaction
should proceed as for a C2-symmetric bis(oxazoline) [7], involving
a complex in which Pd is chelated by two nitrogen atoms. The
nucleophilic attack takes place on the allylic carbon with the lon-
gest Pd–C distance, leading to a major (R) product. An important
increase of the steric hindrance of the substituent (R = tBu), could
cause the rotation of one of the thiazoline cycles to release the con-
straint, leading to a (N,S)-chelation of the palladium by a bis(thiaz-
oline). Therefore, the C2-symmetric bis(thiazoline) would become a
C1-symmetric sulfanyl–thiazoline ligand, which would lead to two
diastereomeric Pd-complexes in equilibrium. These diastereomers
would lead, according to a nucleophilic attack on the allylic carbon
trans to the sulfur atom, to opposite enantiomers of the product
[8,9]. The possible formation and co-existence of three palladium
In a previous paper [3b], some of us reported the synthesis of
various thiazolines, analogous to the well known oxazoline ligands,
and their evaluation in the enantioselective Pd-catalyzed allylic
substitution [6], one of the benchmark reactions in asymmetric
* Corresponding authors. Tel.: +33 231452898.
E-mail addresses: markus.reiher@phys.chem.ethz.ch (M. Reiher), mihaela.
gulea@ensicaen.fr (M. Gulea).
0022-328X/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2008.04.044

2500
A. Betz et al. / Journal of Organometallic Chemistry 693 (2008) 2499–2508
Ligand (6 mol%)
and of a thiazoline were first compared in terms of geometric
and electronic properties, which can explain their different com-
plexing behaviours toward a metal. Then, a quantum chemical
MeO2C
Ph
CO2Me
Ph
OAc
[Pd(allyl)(Cl)]2 (2.5 mol%),
+
CH2(CO2Me)2
(3 equiv.)
Ph
Ph
(1 equiv.)
*
KOAc (0.05 equiv.), BSA (3 equiv.)
CH2Cl2 , rt
study of three
p-allylic palladium complexes was carried out
involving as representative ligands the two C2-symmetric
bis(thiazolines) 1a (R = Et) and 1c (R = tBu) and the oxazoline–
thiazoline 4a. The experimental and theoretical results of the study
are discussed in this paper.
S
S
S
S
S
S
Ligand:
N
N
N
N
N
N
Et
Et
iPr
iPr
tBu
(R,R )-1a
(R,R )-1b
42% ee (R )
Scheme 1. Pd-catalyzed allylic substitution using bis(thiazolines) 1.
tBu (S,S )-1c
Product: 88% ee (R)
31% ee (R)
2. Results and discussion
2.1. Synthesis of non-symmetric bis(thiazolines) and oxazoline–
thiazoline ligands
R
R
S
N
S
N
Four non-symmetric bis(thiazolines) 3a–3d and three oxazo-
line–thiazolines 4a–c were designed as target structures for this
study (see Scheme 3, Table 1). Derivatives 3a and 3b have different
substituents on the thiazoline heterocycles (Et and tBu): 3a has the
same configuration (S) on the two heterocycles, whereas 3b has (R)
configuration on the ethyl substituted thiazoline, and (S) on the
tert-butyl substituted cycle. Bis(thiazolines) 3c and 3d have one
non-substituted thiazoline cycle, and the other cycle substituted
by an ethyl group having (R) configuration for 3c, and by a tert-bu-
tyl having (S) configuration for 3d. Mixed oxazoline–thiazoline li-
gands 4a and 4b are substituted by the same substituent (ethyl)
on each heterocycle, but 4a has (R,R)-configuration, while 4b has
(R) configuration on the oxazoline and (S) on the thiazoline ring.
Oxazoline–thiazoline 4c has substituent on both heterocycles and
(S,S)-configuration.
The general synthesis of bis(thiazolines) 1 from commercially
available aminoalcohols that we developed previously [3a], has
the advantage of a stepwise construction of the two thiazoline
rings, allowing the introduction of two different 2-aminoalcohols
in the same chiral backbone, and, therefore, the synthesis of non-
symmetric bis(thiazolines) such as 3 (Scheme 3). In the first step,
thioamides 7 were prepared by thioacylation of the commercially
available aminoalcohols [(R)- and (S)-2-aminobutanol, (S)-tert-leu-
R
R
N
S
N
S
Pd+
Pd+
Ph
Ph
1
3
Ph
Ph
Nu-
1
3
exo
B2-(N,S)
endo
B1-(N,S)
Nu-
MeO2C
Ph
CO2Me
Ph
MeO2C
CO2Me
Ph
+
Ph
(R )
(S )
R
N
N
S
S
N
R
R
S
S
N
Pd+
Pd+
R
1
3
1
Ph
3
Ph
Ph
Ph
C-(S,S)
A-(N,N)
Nu-
Scheme 2. Proposed palladium complexes in the reaction mixture involving bis-
(thiazolines) 1: complex A-(N,N), two diastereomeric complexes B1-(N,S)-endo and
B2-(N,S)-exo, and C-(S,S).
complexes as intermediates in the reaction mixture (see com-
plexes A, B1, and B2 in Scheme 2), each one leading preferentially
to one enantiomer of the product, could explain the decrease of
the enantiomeric excess, as well as the inversion of the enantiose-
lection, depending on the steric hindrance of the R substituent. It is
worth noting that the possibility of a S,S-chelation of the metal
(complex C) cannot be excluded. In that case the chirality would
certainly be placed too far from the chelation site to induce enanti-
oselectivity, however, the consequence could be the decrease of
the enantiomeric excess.
To support this hypothesis we undertook a new study. First, to
determine the catalytic effect induced by different modifications
around the metal site, we synthesized new thiazoline ligands
(Fig. 1): non-symmetric bis(thiazolines), in which the two hetero-
cycles bear substituents (R1 ¼ R2) of different steric hindrance [10],
and oxazoline–thiazolines, in which only one of the heterocycles,
the thiazoline, can coordinate the Pd by either the nitrogen or
the sulfur atom. Then, these ligands were tested in the Pd-cata-
lyzed allylic substitution. The second part of the study consisted
in a theoretical approach. Simplified structures of an oxazoline
R1
OH
H2N
OH
R1
for X = S: (ii)
for X = O: (iii)
X
X
X
(i)
Y
N
N
H
R1
7: X = S, R1 = Et (95%), tBu (71%)
8: X = O, R1 = Et (95%), tBu (86%)
9: X = S, R1 = Et (87%), tBu (77%)
10: X = O, R1 = Et (59%), tBu (55%)
5: X = S, Y = SMe
6: X = O, Y = Cl
(iv)
R2
OH
H2N
SMe
S
S
(i)
X
(ii)
X
X
OH
N
S
HN
N
N
N
> 95%
R1
R2
R1
R2
R1
11: X = S, R1 = Et (63%), tBu (90%)
12: X = O, R1 = Et (63%), tBu (66%)
3a-d: X = S (61-76%)
4a-c: X = O (47-65%)
13: X = S (not isolated)
14: X = O (not isolated)
(i) NEt3 (1 equiv), THF, rt, 12 to 48h;
(ii) MsCl (1.5 equiv), NEt3 (3 equiv), CH2Cl2, 0°C to rt, 10 min;
(iii) PPh3/DEAD (1.5 equiv), THF, rt, 48h;
(iv) tBuLi (1.1 equiv), THF, -78°C, 1h, then CS2 (1.5 equiv), -10°C to rt, 2h, then MeI (1.5 equiv), -10°C to rt, 2h.
S
S
S
S
S
S
S
S
S
N
N
N
N
N
N
N
N
tBu
tBu
N
tBu
3c
3a
3b
Et
N
3d
Et
Et
S
S
O
O
O
S
S
O
S
N
N
N
N
4b
4c
4a
Et
Et
tBu
Et
Et
N
N
N
N
tBu
R1
R2
R
R
Scheme 3. Synthesis of non-symmetric bis(thiazolines) 3, and oxazoline–thiazo-
lines 4.
Fig. 1. Non-symmetric bis(thiazolines) and oxazoline–thiazoline ligands.

A. Betz et al. / Journal of Organometallic Chemistry 693 (2008) 2499–2508
2501
Table 1
100
88
Synthesis of thiazoline ligands 3a–3d and 4a–4c
80
Ligand
R1
R2
Config
(SEt,StBu
(REt,StBu
(R)
Overall yield (%)
63
a
60
40
3a
3b
3c
3d
4a
4b
4c
Et
Et
H
tBu
tBu
Et
tBu
Et
)
)
35
34
40
30
17
23
1
2
3
4
5
6
7
8
9
a
a
31
a
H
(S)
20
16
b
20
Et
Et
tBu
(Roxa,Rthia
(Roxa,Sthia
(Soxa,Sthia
)
)
)
b
Et
tBu
b
0
20
a
From dithioester 5.
From 2-methyl-propanoyl chloride 6.
-20
-40
-60
-80
-16
b
-25
-32
cinol] with dithioester 5. The non-substituted thiazoline ring (R1 or
R2 = H) obtained from the appropriate dithioester (5 or 11) and 2-
bromoethanolamine hydrobromide, by a one-step thioacylation/
cyclisation [11]. The intramolecular cyclisation of 7 via a mesylate
led to monothiazolines 9, which were deprotonated by t-BuLi to
introduce a new dithioester function, by reaction with carbon
disulfide and trapping of the anion with methyl iodide. From the
resulting thiazoline-dithioester 11 the same subsequent two-step
process thioacylation/cyclisation led to non-symmetric bis(thiazo-
lines) 3. Based on 5, the overall yields are given in Table 1. The
same methodology was used to prepare oxazoline–thiazolines 4
(Scheme 3). The sequence started in this case with the synthesis
of the oxazoline heterocycle. Oxazolines 10 were obtained in two
steps starting from 2-methyl-propanoyl chloride 6 and enantio-
pure 2-aminoalcohols [12]. Mitsunobu conditions [13] were used
for the cyclisation of b-hydroxy amide 8 to the corresponding oxaz-
oline. After introduction of the dithioester group and formation of
the thiazoline ring, oxazoline–thiazolines 4 were obtained, in over-
all yields based on 6, which are given in Table 1.
-66
Ligands (Table 2)
Graph 1. Comparative graph: Ee (%) with bis(thiazolines) of C2-symmetry 1a and 1c
(1–2) and of C1-symmetry 3a–3d (3–6), and with oxazoline–thiazolines 4a–4c (7–
9). Ees (%) are given in positive values for the (R) enantiomer and in negative ones
for the (S) enantiomer.
would be very complex, a qualitative analysis is accessible and
should help to understand the origin of the asymmetric induction.
Results obtained with non-symmetric bis(thiazolines) 3a and
3b show that the sense of the induction is determined by the bulk-
ier substituent. Thus, ligands 3a and 3b, which are diastereomers
having on the thiazoline bearing the tert-butyl substituent the
same configuration, and on the thiazoline bearing the ethyl substi-
tuent opposite configurations, gave quite similar enantiomeric ex-
cess (25% and 32%, respectively) in favour of the (S) product (Table
2, entries 3 and 4). This suggests that the configuration on the
thiazoline bearing the smaller ethyl group did not determine the
asymmetric induction. However, it is worth noting that the de-
crease of the steric hindrance on one of the thiazolines restores
the ‘‘normal” sense of the induction (Table 2, entries 2 and 3). Thus,
(S) thiazoline ligand bearing a tert-butyl group afforded (S)-enan-
tiomer as the major enantiomer. This can be considered as an
experimental evidence to support the hypothesis that the unex-
pected sense of the induction obtained in the case of ligand 1c is
due to the release of the steric strain by rotation of the thiazoline
cycle and coordination of the sulfur atom to the palladium, through
a (N,S) chelation. Equally, changing from 1c to 3d (suppression of
one tert-butyl substituent) the sense of the induction was reversed
from (R) to (S) (Table 2, entries 2 and 6). Going from 1a to 3c (sup-
pression of one ethyl substituent) the sense of the enantioselectiv-
ity was kept affording the same major (R) product, but with a
strong decrease of the enantiomeric excess from 88% to 20%, due
very probably to the loss of the C2-symmetry, but also to the low
steric hindrance of this ligand (Table 2, entries 1 and 5). This first
set of results shows the great influence of steric parameters on
the stereoselectivity of this reaction.
2.2. Pd-catalyzed allylic substitution using thiazoline-containing
ligands
The newly synthesized ligands 3 and 4 were tested in the Pd-
catalyzed allylic substitution in the conditions used before for
bis(thiazolines) 1a and 1c (see conditions in Scheme 1). The reac-
tion which used 1,3-diphenyl-2-propenyl acetate and dimethyl-
malonate, in the presence of 2 equiv. of bis(trimethylsilyl)-
acetamide (BSA) and 5 mol% of potassium acetate (AcOK) [14],
was carried out in dichloromethane, at room temperature, using
2.5 mol% of [Pd(C3H5)Cl]2, 6 mol% of ligand, and afforded (E)-2-
methoxycarbonyl-3,5-diphenylpent-4-enoate as product. The re-
sults are given in Table 2 and the enantiomeric excesses are also
represented in Graph 1 to facilitate comparison. In all cases the
conversion was almost complete (>95%) after about 24 h, except
for ligand 4c [15]. Although a rationalization of the obtained re-
sults by considering all the possible transition states in the reaction
Table 2
Pd-catalyzed asymmetric allylic substitution using thiazoline ligands 1a, 1c, 3a–3d,
and 4a–4c
Then, the oxazoline–thiazolines series was examined. First, li-
gand 4a was compared to bis(thiazoline) 1a, both having (R,R) con-
figuration and ethyl substituents on both heterocycles (Table 2,
entries 1 and 7). (R)-enantiomer was obtained as the major enan-
tiomer in both cases, but the enantiomeric excess decreased from
1a (88%) to 4a (63%). Oxazoline–thiazoline 4b, which is the diaste-
reomer of 4a having the same configuration on the oxazoline ring
and opposite configuration on the thiazoline one, gave (R)-product
in only 16% ee (Table 2, entry 8). Finally, ligand 4c was compared to
bis(thiazoline) 1c, both having (S,S) configuration and tert-butyl
substituents on both heterocycles (Table 2, entries 2 and 9). From
1c to 4c, the sense of the enantioselectivity was reversed and the
enantiomeric excess increased from 31% to 66% [16]. In the case
Entry
Liganda
Conv. (%)b (Time)
Ee%c
Product configurationb
1
2
3
4
5
6
7
8
9
1a
1c
3a
3b
3c
3d
4a
4b
4c
>95 (24 h)
>90 (30 h)
100 (24 h)
100 (24 h)
100 (24 h)
100 (24 h)
100 (24 h)
>95 (24 h)
25 (168 h)
88
31
25
32
20
16
63
16
66
(R)
(R)
(S)
(S)
(R)
(S)
(R)
(R)
(S)
a
For comparison, data for ligands 1a and 1c already published [3b] are included.
Determined by HPLC.
Each experiment was carried out at least two times.
b
c

2502
A. Betz et al. / Journal of Organometallic Chemistry 693 (2008) 2499–2508
of an oxazoline–thiazoline, both supposed (N,N) and (N,S) Pd-com-
plexes should have two diastereomers, A1/A2 and B1 /B2, respec-
tively. The four complexes from a (R,R) oxazoline–thiazoline
ligand of type 4a are represented in Scheme 4. For both complexes
A1 and A2, the steric hindrance between the R substituent and the
phenyl group of the allylic part should direct the sense of the
asymmetric induction toward the (R)-product. The N,S-complexes
B1 and B2 should lead to (S) and (R) enantiomer respectively,
according to the nucleophilic attack in trans position to the sulfur
atom. In the case of B1, the carbon atom trans to the sulfur is also
the most sterically congested. In the case of (R)-oxazoline-(S)-
thiazoline 4b, the low ee obtained (16%) can be explained by the
lack of a good enantiodiscrimination between the two allylic car-
bons in the considered diastereomeric N,N-complexes A1/A2
(Scheme 5).
1
S
1
O
5
4
5
4
2
2
N 3
I
N3
II
Fig. 2.
Table 3
Optimized bond lengths in Å, bond angles in degrees and selected atomic charges (q)
obtained with the natural population analysis (BP86/TZVP) for structures I and II
Heterocycle
Thiazoline I
Oxazoline II
Length (Å) C–Xa
Length (Å) C@N
Angle (°) C2–X–C5
Angle (°) X–C2–N
Dihedral angle X–C5–C4-N
q(N)
1.821
1.275
89.12
117.43
26.06
1.381
1.279
105.02
118.54
11.30
2.3. Oxazoline/thiazoline comparison
The unlike behaviour of thiazoline and oxazoline ligands could
result from several factors such as the heterocycle geometry, and
the modified features of the electronic structure of both ligands
(as deduced, e.g., by inspection of the change of the frontier orbi-
tals). Thiazoline I and oxazoline II have been used as simple models
for theoretical calculations (Fig. 2). Their geometries were opti-
mized, and bond lengths, angles, and electronic densities were cal-
culated for both heterocycles (Table 3). Due to the difference
between the carbon-heteroatom bond lengths and angles, the
two heterocycles have different geometries. Compared to the qua-
ꢀ0.392
ꢀ0.437
q(X)
+0.162
ꢀ0.445
a
The sulfur and oxygen atoms in I and II, respectively, are denoted by an ‘‘X” in
the 1st column.
si-planar geometry of oxazolines, thiazolines have a distorted and
larger ring. This may cause, especially when the heterocycles are
substituted, steric interactions in the metal–ligand complex, which
can influence the enantioselectivity. Moreover, the important dif-
ference of electronegativity between oxygen and sulfur can explain
the different electronic repartition in the two heterocycles.
As shown in Table 3, according to the results obtained from nat-
ural population analysis, the negative atomic charge of nitrogen in
oxazoline II is slightly higher than it is in thiazoline I, and the
atomic charge is positive for sulfur and negative for oxygen. This
is in accord with the fact that sulfur’s electronegativity is similar
to the one of carbon and lower than oxygen and nitrogen electro-
negativities [C (2.5) = S (2.5) < N (3.0) < O (3.5)] [17].
R
R
O
O
N
N
R
R
N
S
N
S
Pd+
Pd+
Ph
Ph
1
3
Ph
Ph
Nu-
1
3
B2-(N,S
)
B1-(N,S
)
Nu-
For a qualitative picture of the complexation of the metal with
an oxazoline or a thiazoline ligand, the HOMO should also be con-
sidered. As shown in Fig. 3, for oxazoline II, the largest part of the
HOMO is located on the nitrogen atom, however, the oxygen atom
contributes also significantly. In contrast, the HOMO of thiazoline I
is located on the sulfur atom (which may be taken as an indication
for its affinity to soft metals), while the contributions on the nitro-
gen atom are relatively small. This may be seen as an indication of
differences in the type of chelation and in the stability of the com-
plexes (qualitatively speaking, the HOMOs of these sigma-donor li-
gands will govern the donation of electron density to the transition
metal atom).
MeO2C
CO2Me
Ph
MeO2C
Ph
CO2Me
Ph
+
Ph
(
S
)
(R)
S
S
N
O
O
R
R
N
N
N
Pd+
Pd+
R
R
Ph
Ph
Ph
Ph
1
3
3
1
Nu-
Nu-
A2-(N,N
)
A1-(N,N
)
Scheme 4. Proposed palladium complexes involving a (R)-oxazoline-(R)-thiazoline
in the reaction mixture: two N,N-complexes A1 and A2 and two N,S-complexes B1
and B2.
S
S
O
O
N
N
N
N
Pd+
Pd+
R
R
R
Ph
R
Ph
Ph
1
3
Ph
1
3
A2-(N,N)
A1-(N,N)
Scheme 5. (N,N) complexes A1 and A2 involving a (R)-oxazoline-(S)-thiazoline
Fig. 3. HOMO of thiazoline I and of oxazoline II (both have been plotted with the
ligand.
same isosurface value).

A. Betz et al. / Journal of Organometallic Chemistry 693 (2008) 2499–2508
2503
These data show important differences between a oxazoline
and its thiazoline analogue and let us anticipate a different behav-
iour in the metal coordination (and therefore their different cata-
lytic activities and enantioselectivities).
[6] occurs faster than the nucleophilic attack. Therefore, interpreta-
tion of the theoretical results should be considered carefully, since
the rate-limiting and selectivity-determining steps do not have to
be the same.
For the theoretical investigation performed in this work we as-
sume that all diastereomers can interconvert. Thus, barriers are ex-
pected to be small, an equilibrium is established and the reaction
products are therefore determined by the relative thermodynamic
stability of the different structures. Consequently, we deem it suf-
ficient to analyse the stable minimum structures of all relevant dia-
stereomers. In order to do so, we only consider the difference in
electronic energy at 0 K. This is sufficient because temperature ef-
fects on the enthalpy and on the entropy as well as the zero-point
vibrational energy can safely be neglected because of the high sim-
ilarity of all structures under consideration. The electronic energy
differences allow us to discuss purely electronic effects.
As a first step, we carried out a theoretical investigation of pal-
ladium complexes involving bis(thiazolines) 1a and 1c. In Figs. 4–6
the optimized geometries (BP86/RI/TZVP) are shown and the corre-
sponding relative energies of the intermediate complexes with li-
gands 1a (Fig. 4), 1c (Fig. 5), and 4a (Fig. 6) are given. For
comparison, single-point energies from B3LYP/TZVP//BP86/RI/
TZVP calculations are given in parentheses. In addition, selected
carbon–palladium (Pd–C1 and Pd–C2) bond distances are given in
angstroms. The S,S-complex C was taken into consideration only
2.4. Quantum chemical study of the p-allylic palladium complexes
involving bis(thiazolines) 1a, 1c, and 4a
The reaction mechanism and the possible modes of stereoselec-
tion in palladium-catalyzed allylation are well studied [6]. With
symmetrically substituted allyl moiety (such as in the 1,3-diphe-
nyl-2-propenyl acetate), the stereoselection results from the differ-
ence in reactivity of the allyl termini in the nucleophilic addition
step. It was shown that in a meso-(p-allyl)palladium complex,
the carbon with the longest distance to Pd is preferentially at-
tacked by the nucleophile [6]. The Pd–C bond elongation can come
from a steric congestion, or, like in the case of C1-symmetric li-
gands with different donor heteroatoms, from the strong donating
effect of one atom (provoking the bond elongation trans to this
atom). Some theoretical studies have been done to gain a better
understanding of the enantioselectivity of this reaction [18].
According to one model, an elongation of the Pd–C bond by
D
l = 0.01 Å will approximately double the reactivity at this carbon
[18h]. It was also established that in the case of a C1-symmetric li-
gand, the equilibrium between the two diastereomeric complexes
Fig. 4. BP86/RI/TZVP-optimized structures and energies of Pd-complexes with ligand 1a.

2504
A. Betz et al. / Journal of Organometallic Chemistry 693 (2008) 2499–2508
Fig. 5. BP86/RI/TZVP-optimized structures and energies of Pd-complexes with ligand 1c.
Fig. 6. BP86/RI/TZVP-optimized structures and energies of Pd-complexes with ligand 4a.
in one example (for ligand 1a), since steric or electronic factors
could not influence the enantioselectivity in this case, and they
probably not vary from one ligand to another. However, to com-
pare its stability with those of the N,N and N,S complexes could
give an indication of its responsibility for a low ee.
Fig. 4 shows that the stability of the complexes decreases in the
order 1a-A > 1a-B1 > 1a-B2 > 1a-C. The N,N-complex 1a-A is 6.0
and 8.3 kJ/mol more stable than the two N,S-complexes 1a-B1/B2,
respectively. Complex 1a-C was found to be the least stable
(22.4 kJ/mol less stable than 1a-A), therefore, its contribution in
the context of this study can be neglected. Complex 1a-B2 was cal-
culated to be less stable than its diastereomer 1a-B1, however, the
difference of energy is small (2.3 kJ/mol) and hence almost negligi-
ble. According to the calculated Pd–allylic carbon bond lengths (see
Table 4), complexes 1a-A and 1a-B1 may lead to the main product
having (R) configuration, while 1a-B2 may lead to the opposite
enantiomer (S), since nucleophilic attack occurs on the carbon with
the longest Pd–C bond. These theoretical data suggest that the
nucleophile attacks preferentially the carbon C1 of the N,N-com-
plex 1a-A, leading to (R) enantiomer as the major product. These
theoretical results are thus in agreement with the experimental re-
sults obtained in the case of bis(thiazoline) ligand 1a (88% ee in fa-
vour of the R enantiomer). Fig. 5 shows that the stability of the
complexes decreases in the order 1c-B1 > 1c-B2 > 1c-A. It could

A. Betz et al. / Journal of Organometallic Chemistry 693 (2008) 2499–2508
2505
Table 4
Calculated elongation
D
l = j(Pd–C1)–(Pd–C3)j and expected configuration of the product, for allyl palladium complexes derived from ligands 1a, 1c, and 4a
Ligand
1a
1c
4a
Complex
1a-A
1a-B1
1a-B2
1a-C
1c-A
1c-B1
1c-B2
4a-A1
4a-A2
4a-B1
4a-B2
D
l (Å)
0.118a
(R)
0.031
(R)
0.063
(S)
0.017
(S)
0.189a
(S)
0.018
(S)
0.110
(R)
0.098
(R)
0.118a
(R)
0.014
(R)
0.052
(S)
Predicted configurationb
a
Most important elongation for each case of ligand.
The predicted configuration of the product is that resulting from the nucleophilic attack on the Pd-complexes A, B, or C corresponding to ligands 1a and 1c (see Scheme 2)
b
and to ligand 4a (see Scheme 4).
be immediately noticed that in contrast with the previous case of
1a (with Et substituents), in the case of ligand 1c (with bulky tBu
substituents) the N,N-complex 1c-A is the least stable. Moreover,
the difference of energies between complexes 1c-B2/B1 and 1c-A
becomes more important (18 and 21 kJ/mol, respectively). The dia-
stereomeric complexes 1c-B2 and 1c-B1, which are very close in
energy (3 kJ/mol in favour of 1c-B1), would lead to products with
opposite configuration. Therefore the low ee observed is not
surprising.
which the choice of the metal will direct the selection of the che-
lating heteroatom. Thus, thiazoline ligands should not be consid-
ered as simple oxazoline analogues, and the specific properties of
sulfur, which is able to coordinate soft metals unlike oxygen, need
to be taken into account.
4. Computational methods
All-electron calculations were performed with the quantum
chemical program package Turbomole [19] employing either the
BP86 functional [20,21] in combination with the resolution-of-
the-identity (‘RI’) density fitting technique or the B3LYP functional
[22,23]. For the palladium atom, an effective core potential of the
Stuttgart group was employed (which also accounts for the scalar
relativistic effects) [24]. This potential includes 28 electrons of
the inner shells, i.e. the 18 electrons belonging to the 4s, 4p, 4d
and 5s shell are treated explicitly. While BP86/RI was used for all
structure optimizations, B3LYP calculations were only performed
for single points i.e. for the BP86/RI optimized structures. We used
the TZVP basis by Schäfer et al. [25] throughout, which features a
valence triple-zeta basis set with polarization functions on all
atoms. For the calculations of charges, the natural population anal-
yses from GAUSSIAN03 [26] was used. Solvent effects on the relative
energies of the Pd-complexes investigated in this work may not
be negligible. In order to assess the role of solvent effects we
employed a continuum model – namely the COSMO model – to
account for a CH2Cl2 environment. These calculations have been
carried out for the 1a series with the default COSMO parameters
of the TURBOMOLE/COSMO implementation. The optimized atom-
ic COSMO radii of ref. [27] have been used in combinations with
non-optimized radii of 2.28 Å for palladium atoms [28]. As a result
of the COSMO solvation, the relative energies of the complexes 1a-
B1, B2, C were reduced by 3–4 kJ/mol compared to 1a-A. Hence, the
energetical ordering of the different structures is unchanged and
the qualitative picture is the same as in the case of the isolated
structures (in vacuum). Therefore, all energies reported here are
given for the generic complex structures in vacuum. Molecular
structures were visualized with the program Molden [29].
For the palladium complexes involving the mixed thiazoline–
oxazoline ligand 4a (Fig. 6) four complexes are considered, and
their stability decreases as follows: 4a-A1 > 4a-A2 > 4a-B1 > 4a-B2
(DE = 6.2 kJ/mol between the most stable and the less stable com-
plex). As with bis(thiazolines) 1a and 1c, the calculated energy dif-
ference between the two diastereomeric complexes 4a-B1 and 4a-
B2 is negligible (2.8 kJ/mol). The calculated energy difference be-
tween the diastereomeric complexes 4a-A1 and 4a-A2 is also small
(2.1 kJ/mol). According to the nucleophilic attack on the carbon
with the longest Pd–C bond, complexes A1, A2 and B1 may lead
to the main product of (R) configuration, while complex B2 may
lead to the opposite enantiomer (S). These theoretical data can
again be correlated to the experimental result obtained with 4a
(63% ee and R configuration of the product).
In Table 4 are given for each complex derived from ligands 1a,
1c, and 4a, the calculated elongation
Dl of one palladium–carbon
bond in the 1,3-diphenylallyl palladium complexes and the ex-
pected configuration of the product. In the three studied cases,
an elongation between 0.01 and 0.19 Å was found. For all N,S-com-
plexes (B), the longest Pd–C distance was found to be trans to the
sulfur atom. These theoretical results are in agreement with the ex-
pected influence of the steric hindrance of the substituents and
with the strong trans effect of sulfur in this type of ligands.
3. Conclusion
New thiazoline-containing ligands including non-symmetric
bis(thiazolines) and oxazoline–thiazolines were synthesized and
then compared to C2-symmetric bis(thiazolines) in the palla-
dium-catalyzed allylic substitution. The experimental results ob-
tained in this study support the hypothesis of a competition
between the (N,N) and the (N,S) palladium chelation when
bis(thiazolines) are used as ligands, and reveal the importance in
this reaction of the steric hindrance of the R substituent in alpha
position to the nitrogen atom. A quantum chemical study on the
Pd-complexes derived from three selected ligands, two C2-sym-
metric bis(thiazolines) (substituted by R = Et and tBu, respectively)
and one oxazoline–thiazoline, was also performed. Although the
obtained theoretical results should be considered carefully since
only the relative thermodynamic stability of the different struc-
tures was considered, they are in good agreement with the exper-
imental results and also support the competition between the
(N,N) and the (N,S) palladium chelation.
5. Experimental
5.1. General
The reactions were carried out under nitrogen atmosphere un-
less stated otherwise and monitored by TLC using silica plates 60
F254 (Merck). Solvents were used in RPE grade (Carlo Erba) with-
out further purification except for the reactions involving t-BuLi
and Grignard reagents. In these cases, solvents were purified with
a PURESOLV SPS400 apparatus developed by Innovative Technol-
ogy Inc. Column chromatography was performed on silica gel Si
60 (40–63 lm) from Merck. tert-Butyllithium was titrated prior
to use according to a described procedure [30]. NMR spectra were
recorded on a Bruker DRX400 spectrometer. Chemical shifts are re-
ported in ppm downfield from tetramethylsilane with residual sol-
vent signal used as an internal standard, coupling constants are
Therefore, a bis(thiazoline), which can be either a C2-symmetric
N,N- or S,S-ligand or a C1-symmetric N,S-ligand, is a particular type
of ligand, which can open the way to new catalytic reactions in

2506
A. Betz et al. / Journal of Organometallic Chemistry 693 (2008) 2499–2508
indicated in Hz. Splitting patterns are designated as: s = singlet,
d = doublet, t = triplet, q = quartet, sept, dd = doublet of doublets,
m = multiplet, and br = broad. IR spectra were measured on a Per-
kin–Elmer 16 PC FT-IR instrument. Mass spectra were recorded on
as aminoalcohols (R)-(ꢀ)-2-amino-1-butanol and (S)-(+)-tert-leu-
cinol, respectively. In these cases (formation of unsubstituted
thiazoline), the thiazoline-dithioester 11 was reacted with 2-bro-
moethylamine hydrobromide and thioacylation and cyclisation
take place in one-step (this step is detailed below for 3c).
a Waters Q-TOF micro by electrospray ionization (ES-MS). Optical
20
rotation values [
a]
were measured on a Perkin–Elmer 241 polar-
D
imeter and are given in 10ꢀ1° cm2 gꢀ1, c in g 100 mlꢀ1
.
5.3.1. (R)-4-Ethyl-2-[1-methyl-1-(4,5-dihydro)thiazolyl]ethyl-2-
thiazoline (3c)
5.2. Synthesis of non-symmetric bis(thiazolines) (3a and 3b)
2-Bromoethylamine hydrobromide (209 mg, 1.01 mmol) was
suspended in THF (2 mL) and deprotonated with triethylamine
(0.14 mL, 102 mg, 1.01 mmol) at 0 °C. The ice bath was removed
They were prepared according to the general procedure previ-
ously described for C2-symmetric bis(thiazolines) [3a], in five steps
starting from isopropyl methyldithioester 5, and using two differ-
ent aminoalcohols in the two thioacylation steps.
and
a
solution of (R)-methyl-2-methyl-2-(4-ethyl-4,5-dihy-
dro)thiazolyl-propanedithioate (250 mg, 1.01 mmol) in THF
(1 ml) was slowly added. The addition of two further equivalents
of NEt3 was performed after 30 min. The mixture was stirred over
night, filtered and solvents evaporated. Purification by chromatog-
raphy on silica gel (Et2O/pentane: 50/50, Rf = 0.30 in UV) afforded
the product as a pale yellow oil (185 mg, 0.76 mmol). Overall yield:
5.2.1. (SEt,St-Bu)-4-tert-Butyl-2-[1-methyl-1-(4-ethyl-4,5-
dihydro)thiazolyl]ethyl-2-thiazoline (3a)
Complex 3a was prepared according to the general procedure,
using as aminoalcohols (S)-(+)-2-amino-1-butanol and (S)-(+)-
tert-leucinol. Purification by chromatography on silica gel (Et2O/
40%. [
a]
20 +76 (c 1.2, CH2Cl2). 1H NMR (CDCl3, 400 MHz): d = 1.00 (t,
D
3JHH = 7.4 Hz, 3H, CH2CH3), 1.55 (s, 3H, C(CH3)2), 1.56 (s, 3H, C(CH3)2),
2
3
pentane: 20/80, Rf = 0.48 in UV) afforded the product as an orange
1.58–1.85 (m, 2H, CH2CH3), 2.97 (dd, JHH = 10.8 Hz, JHH = 7.1 Hz,
1H, SCHHCHEt), 3.27 (t, JHH = 8.4 Hz, 2H, SCH2CH2), 3.34 (dd,
20
3
oil (496 mg, 1.57 mmol). Overall yield: 35%. [
a
]
ꢀ128 (c 1.3,
D
CHCl3). 1H NMR (CDCl3, 400 MHz): d = 0.98 (s, 9H, C(CH3)3), 1.00 (t,
2JHH = 10.8 Hz, JHH = 8.6 Hz, 1H, SCHHCHEt), 4.25 (t, JHH = 8.4 Hz,
2H, SCH2CH2), 4.41–4.48 (m, 1H, CHEt). 13C NMR (CDCl3,
101 MHz): d = 10.8 (CH2CH3), 26.7 (C(CH3)2), 26.8 (C(CH3)2), 27.7
(CH2CH3), 34.0 (SCH2CH2), 37.7 (SCH2CHEt), 47.4 (C(CH3)2), 64.7
(SCH2CH2), 78.6 (CHEt), 173.5 (CNCHEt), 176.0 (CNCH2). IR (cmꢀ1):
2962, 2931, 2854, 1611, 1461, 1383, 1363, 1223, 1048, 988, 880,
713. TOF MS ES+ [M+H]+ Calc. for C11H19N2S2: 243.0990; Found:
243.0991. ES-MS m/z (%): 243 (M+H, 36), 183 (15), 155 (100), 128
(3).
3
3
3JHH = 7.4 Hz, 3H, CH2CH3), 1.53 (s, 3H, C(CH3)2), 1.54 (s, 3H,
2
C(CH3)2), 1.59–1.84 (m, 2H, CH2CH3), 2.96 (dd, JHH = 10.8 Hz,
3JHH = 7.0 Hz, 1H, SCHHCHEt), 3.08 (dd, JHH = 11.0 Hz, JHH = 9.8 Hz,
2
3
2
3
1H, SCHHCHt-Bu), 3.19 (dd, JHH = 10.9 Hz, JHH = 9.1 Hz, 1H,
SCHHCHt-Bu), 3.32 (dd, 2JHH = 10.8 Hz, 3JHH = 8.5 Hz, 1H, SCHHCHEt),
3
3
4.18 (dd, JHH = 9.1 Hz, JHH = 9.8 Hz, 1H, CHt-Bu), 4.37–4.47 (m, 1H,
CHEt). 13C NMR (CDCl3, 101 MHz): d = 10.8 (CH2CH3), 26.7 (C(CH3)2),
26.85 (C(CH3)3), 26.90 (C(CH3)2), 27.6 (CH2CH3), 34.4 (SCH2CHt-Bu),
35.5 (C(CH3)3), 37.7 (SCH2CHEt), 47.5 (C(CH3)2), 78.4 (CHEt), 87.0
(CHt-Bu), 172.6 (CNCHt-Bu), 173.9 (CNCHEt). IR (cmꢀ1): 2958,
2933, 2868, 1614, 1463, 1363, 1220, 1057, 994, 880. TOF MS ES+
[M+H]+ Calc. for C15H27N2S2: 299.1616; Found: 299.1621. ES-MS
m/z (%): 299 (M+H, 32), 211 (100), 183 (78), 129 (5), 89 (21), 83
(23). Elemental Anal. Calc. for C15H26N2S2: C, 60.35; H, 8.78; N,
9.38. Found: C, 60.20; H, 8.96; N, 9.12%.
5.3.2. (S)-4-tert-Butyl-2-[1-methyl-1-(4,5-dihydro)thiazolyl]-ethyl-2-
thiazoline (3d)
Was prepared according to the general procedure. Purification
by chromatography on silica gel (Et2O/pentane: 30/70, Rf = 0.30
in UV) afforded the product as
a pale yellow oil (254 mg,
20
0.94 mmol). Overall yield: 30%. [
a
]
D
ꢀ71 (c 0.95, CHCl3). 1H NMR
(CDCl3, 400 MHz): d = 0.99 (s, 9H, C(CH3)3), 1.54 (s, 3H, C(CH3)2),
2
3
5.2.2. (REt,St-Bu)-4-tert-Butyl-2-[1-methyl-1-(4- ethyl-4,5-
dihydro)thiazolyl]ethyl-2-thiazoline (3b)
1.56 (s, 3H, C(CH3)2), 3.09 (dd, JHH = 10.9 Hz, JHH = 9.8 Hz, 1H,
2
3
SCHHCHt-Bu), 3.21 (dd, JHH = 11.0 Hz, JHH = 9.1 Hz, 1H, SCHHCHt-
3
3
Complex 3b was prepared according to the general procedure,
using as aminoalcohols (R)-(ꢀ)-2-amino-1-butanol and (S)-(+)-
tert-leucinol. Purification by chromatography on silica gel (Et2O/
pentane: 20/80, Rf = 0.60 in UV) afforded the product as an orange
Bu), 3.26 (t, JHH = 8.3 Hz, 2H, SCH2CH2), 4.19 (dd, JHH = 9.1 Hz,
3
3JHH = 9.8 Hz, 1H, CHt-Bu), 4.25 (t, JHH = 8.3 Hz, 2H, SCH2CH2). 13C
NMR (CDCl3, 101 MHz): d = 26.6 (C(CH3)2), 26.8 (C(CH3)3), 26.9
(C(CH3)2), 34.1 (SCH2CH2), 34.4 (SCH2CHt-Bu), 35.5 (C(CH3)3), 47.5
(C(CH3)2), 64.6 (SCH2CH2), 87.1 (CHt-Bu), 172.4 (CNCHt-Bu), 176.2
(CNCH2). IR (cmꢀ1): 2954, 2866, 1614, 1464, 1363, 1222, 1058,
1022, 992, 929, 880, 720. TOF MS ES+ [M+H]+ Calc.: for
oil (591 mg, 1.98 mmol). Overall yield: 34%. [a]
20+9 (c 1.3, CHCl3).
D
1H NMR (CDCl3, 400 MHz): d = 0.98 (s, 9H, C(CH3)3), 0.99 (t,
3JHH = 7.4 Hz, 3H, CH2CH3), 1.53 (s, 3H, C(CH3)2), 1.55 (s, 3H,
2
C(CH3)2), 1.59–1.85 (m, 2H, CH2CH3), 2.96 (dd, JHH = 10.8 Hz,
C13H23N2S2: 271.1303; Found: 271.1316. ES-MS m/z (%): 271
2
3
3JHH = 7.2 Hz, 1H, SCHHCHEt), 3.08 (dd, JHH = 11.0 Hz, JHH = 9.9 Hz,
(M+H, 18), 211 (25), 155 (100), 128 (13), 83 (26). Elemental Anal.
Calc. for C13H22N2S2: C, 57.73; H, 8.20; N, 10.36; Found: C, 58.12;
H, 8.48; N, 10.62%.
2
3
1H, SCHHCHt-Bu), 3.20 (dd, JHH = 11.0 Hz, JHH = 9.0 Hz, 1H,
SCHHCHt-Bu), 3.33 (dd, 2JHH = 10.8 Hz, 3JHH = 8.5 Hz, 1H, SCHHCHEt),
3
3
4.18 (dd, JHH = 9.1 Hz, JHH = 9.8 Hz, 1H, CHt-Bu), 4.40–4.47 (m, 1H,
CHEt). 13C NMR (CDCl3, 101 MHz): d = 10.8 (CH2CH3), 26.6 (C(CH3)2),
26.9 (C(CH3)3), 27.0 (C(CH3)2), 27.7 (CH2CH3), 34.4 (SCH2CHt-Bu),
35.5 (C(CH3)3), 37.8 (SCH2CHEt), 47.5 (C(CH3)2), 78.5 (CHEt), 87.1
(CHt-Bu), 172.5 (CNCHt-Bu), 173.9 (CNCHEt). IR (cmꢀ1): 2959,
2933, 2869, 1615, 1463, 1363, 1219, 1057, 994, 880. TOF MS ES+
[M+H]+ Calc. for C15H27N2S2: 299.1616; Found: 299.1626. ES-MS
m/z (%): 299 (M+H, 18), 211 (100), 183 (58), 89 (11), 83 (12).
5.4. Synthesis of oxazoline–thiazolines (4a–4c)
They were prepared in five steps starting from isobutyric acid
chloride 6, and using two different aminoalcohols in the acylation,
then in the thioacylation step. Starting from oxazolines 10 (see
their general synthesis below) instead of thiazolines 9, the same
general procedure previously described for bis(thiazolines) was
used.
5.3. Synthesis of non-symmetric bis(thiazolines) (3c and 3d)
5.4.1. (R,R)-4-Ethyl-2-[1-methyl-1-(4-ethyl-4,5-dihydro)-
oxazolyl]ethyl-2-thiazoline (4a)
Complex 4a was prepared according to the general procedure
from oxazoline 10a using in the thioacylation step (R)-(ꢀ)-2-ami-
They were prepared according to the general procedure previ-
ously described for C2-symmetric bis(thiazolines) [3a], starting
from isopropyl methyldithioester 5, using for the first thioacylation

A. Betz et al. / Journal of Organometallic Chemistry 693 (2008) 2499–2508
2507
no-1-butanol. Purification by chromatography on silica gel (Et2O/
5.5.1. (R)-N-(1-Hydroxymethyl)propyl-2-methyl-propaneamide (8a)
pentane: 30/70, Rf = 0.10 in UV) afforded the product as yellow
Was prepared using (R)-(ꢀ)-2-amino-1-butanol as aminoalco-
oil (70 mg, 0.28 mmol). Overall yield: 17%. [
a
]
20 +108(c 1.0, CHCl3).
hol. Yield: 4.54 g (28.5 mmol, 95%). Rf = 0.30 (EtOAc, blue in
D
1H NMR (CDCl3, 400 MHz): d = 0.91 (t, 3JHH = 7.4 Hz, 3H,
PMA). M.p. 70 °C. [
a]
+44 (c 0.95, CHCl3). 1H NMR (CDCl3,
20
D
3
3
OCNCHCH2CH3), 0.97 (t, JHH = 7.4 Hz, 3 H, SCNCHCH2CH3), 1.49–
400 MHz): d = 0.95 (t, JHH = 7.5 Hz, 3H, CH2CH3), 1.16 (d,
3
1.81 (m, 4H, 2 ꢁ CH2CH3), 1.51 (s, 6H, C(CH3)2), 2,95 (dd,
3JHH = 6.9 Hz, 3H, CH(CH3)2), 1.17 (d, JHH = 6.9 Hz, 3H, CH(CH3)2),
3
2
3
2JHH = 10.8 Hz, JHH = 7.1 Hz, 1H, SCHH), 3.33 (dd, JHH = 10.8 Hz,
1.42–1.66 (m, 2H, CH2CH3), 2.38 (sept, JHH = 6.9 Hz, 1H, CH(CH3)2),
3JHH = 8.6 Hz, 1H, SCHH), 3.91 (t, 1H, OCHH), 4.02–4.10 (m, 1H,
2.60 (br, 1H, OH), 3.58 (dd, 2JHH = 11.0 Hz, 3JHH = 5.8 Hz, 1H, CHHOH),
2
3
2
3
OCNCHEt), 4.27 (dd, JHH = 9.3 Hz, JHH = 8.1 Hz, 1H, OCHH), 4.39–
4.46 (m, 1H, SCNCHEt). 13C NMR (CDCl3, 101 MHz): d = 9.7 (CH2CH3),
10.7 (CH2CH3), 25.58 (C(CH3)2), 25.62 (C(CH3)2), 27.8 (CH2CH3), 28.2
(CH2CH3), 37.7 (SCH2), 43.1 (C(CH3)2), 67.2 (OCNCHEt), 72.4 (OCH2),
78.5 (SCNCHEt), 169.6 (OCN), 172.7 (SCN).
3.68 (dd, JHH = 11.0 Hz, JHH = 3.4 Hz, 1H, CHHOH), 3.80–3.88 (m,
1H, CHEt), 5.67 (br, 1H, NH). 13C NMR (CDCl3, 101 MHz): d = 10.7
(CH2CH3), 19.7 (CH(CH3)2), 19.9 (CH(CH3)2), 24.3 (CH2CH3), 35.9
(CH(CH3)2), 53.4 (CHEt), 65.7 (CH2OH), 178.3 (NCO). IR (cmꢀ1):
3278, 2967, 2932, 1641, 1542, 1236, 1101, 1046, 711. TOF MS ES+
[M+H]+ Calc. for C8H18NO2: 160.1338; Found: 160.1342. ES-MS m/z
(%): 160 (M+H, 60), 143 (21), 142 (5), 90 (100).
5.4.2. (R,S)-4-Ethyl-2-[1-methyl-1-(4-ethyl-4,5-dihydro)-
oxazolyl]ethyl-2-thiazoline (4b)
Was prepared according to the general procedure from oxazo-
line 10a using in the thioacylation step (S)-(+)-2-amino-1-butanol.
Purification by chromatography on silica gel (Et2O/pentane: 50/50,
5.5.2. (S)-N-(1-Hydroxymethyl-2,2-dimethyl)propyl-2-methyl-
propaneamide (8b)
Was prepared using (S)-(+)-tert-leucinol as aminoalcohol. Yield:
Rf = 0.22 in UV) afforded the product as yellow oil (546 mg,
1.29 g (6.9 mmol, 86%). Rf = 0.35 (EtOAc, blue in PMA). M.p. 124 °C.
20
ꢀ5.8 (c 1.05, CHCl3). 1H NMR
20
[a]
ꢀ18 (c 1.05, CHCl3). 1H NMR (CDCl3, 400 MHz): d = 0.95 (s, 9H,
2.15 mmol). Overall yield: 23%. [
a]
D
D
3
3
3
(CDCl3, 400 MHz): d = 0.92 (t, JHH = 7.4 Hz, 3H, OCNCHCH2CH3),
C(CH3)3), 1.17 (d, JHH = 6.9 Hz, 3H, CH(CH3)2), 1.18 (d, JHH = 6.9 Hz,
3H, CH(CH3)2), 2.43 (sept, 3JHH = 6.9 Hz, 1H, CH(CH3)2), 3.51–3.57 (m,
1H, CHHOH), 3.79–3.87 (m, 1 + 1H, CHHOH and CHt-Bu), 5.67 (br,
1H, NH). 13C NMR (CDCl3, 101 MHz): d = 19.7 (CH(CH3)2), 20.1
(CH(CH3)2), 27.1 (C(CH3)3), 33.6 (C(CH3)3), 36.1 (CH(CH3)2), 59.4
(CHt-Bu), 63.5 (CH2OH), 178.6 (NCO). IR (cmꢀ1): 3282, 2966, 1649,
1557, 1369, 1267, 1238, 1051, 910. TOF MS ES+ [M+H]+ calcd. for
3
0.98 (t, JHH = 7.4 Hz, 3H, SCNCHCH2CH3), 1.50–1.82 (m, 4H, 2x
CH2CH3), 1.52 (s, 6H, C(CH3)2), 2,96 (dd, 2JHH = 10.8 Hz, 3JHH = 7.1 Hz,
1H, SCHH), 3.34 (dd, 2JHH = 10.8 Hz, 3JHH = 8.6 Hz, 1H, SCHH), 3.92 (t,
2
1H, OCHH), 4.03–4.11 (m, 1H, OCNCHEt), 4.29 (dd, JHH = 9.3 Hz,
3JHH = 8.1 Hz, 1H, OCHH), 4.40–4.47 (m, 1H, SCNCHEt). 13C NMR
(CDCl3, 101 MHz): d = 9.7 (CH2CH3), 10.7 (CH2 CH3), 25.5 (C(CH3)2),
25.8 (C(CH3)2), 27.8 (CH2CH3), 28.2 (CH2CH3), 37.7 (SCH2), 43.1
(C(CH3)2), 67.3 (OCNCHEt), 72.4 (OCH2), 78.5 (SCNCHEt), 169.6
(OCN), 172.7 (SCN). IR (cmꢀ1): 2962, 2933, 2876, 1661, 1620, 1133,
1029, 981, 958, 922, 754. TOF MS ES+ [M+H]+ Calc. for C13H23N2OS:
255.1531; Found: 255.1538. ES-MS m/z (%): 255 (M+H, 29), 183 (28),
167 (100), 156 (2), 125 (5), 113 (2), 89 (13).
C10H22NO2: 188.1651; Found: 188.1655. ES-MS m/z (%): 188 (M+H,
70), 118 (100), 100 (21), 83 (18).
Cyclisation: To a solution of PPh3 (1.5 equiv.) in dry THF
(c = 0.5 M as to the amide concentration), DEAD (1.5 equiv.) was
added dropwise and the solution was stirred 40 min at 20 °C. The
corresponding amide (1 equiv.) was introduced in several portions
and the mixture was allowed to react for 24 h. After filtration of the
precipitate, solvents were evaporated to about a forth of the origi-
nal volume and the mixture was poured onto the same amount of
pentane. Filtration, evaporation of solvents and column chroma-
tography yielded the oxazolines as colourless oils.
5.4.3. (S,S)-4-tert-Butyl-2-[1-methyl-1-(4-tert-butyl-4,5-
dihydro)oxazolyl]ethyl-2-thiazoline (4c)
Complex 4a was prepared according to the general procedure
from oxazoline 10b using in the thioacylation step (S)-(+)-tert-leu-
cinol. Purification by chromatography on silica gel (Et2O/pentane:
20/80, Rf = 0.29 in UV) afforded the product as yellow oil (76 mg,
5.5.3. (R)-4-Ethyl-2-isopropyl-2-oxazoline (10a)
20
0.24 mmol). Overall yield: 20%. [
a
]
D
ꢀ89 (c 1.0, CHCl3). 1H NMR
Was prepared from amide 8a. Yield: 1.99 g (14.1 mmol, 59 %).
(CDCl3, 400 MHz): d = 0.89 (s, 9H, C(CH3)3), 0.95 (s, 9H, C(CH3)3),
20
Rf = 0.14 (Et2O/pentane: 10/90, green in PMA). [
a]
+79 (c 1.05,
1.516 (s, 3H, C(CH3)2), 1.520 (s, 3H, C(CH3)2), 3.08 (dd, 2JHH = 11.0 Hz,
D
CHCl3). 1H NMR (CDCl3, 400 MHz): d = 0.91 (t, JHH = 7.4 Hz, 3 H,
3
2
3
3JHH = 9.3 Hz, 1H, SCHH), 3.20 (dd, JHH = 11.0 Hz, JHH = 9.2 Hz, 1H,
3
CH2CH3), 1.18 (d, JHH = 7.0 Hz, 6 H, CH(CH3)2), 1.42–1.69 (m, 2H,
2
3
SCHH), 3.84 (dd, JHH = 10.1 Hz, JHH = 7.2 Hz, 1H, OCNCHt-Bu), 4.08
3
CH2CH3), 2.56 (septet, JHH = 7.0 Hz, 1H, CH(CH3)2), 3.82–3.86 (m,
2
3
2
(dd, JHH = 8.6 Hz, JHH = 7.2, 1H, OCHH), 4.15 (dd, JHH = 8.6 Hz,
2
1H, OCHH), 3.97–4.04 (m, 1H, CHEt), 4.25 (dd, 1H, JHH = 9.4 Hz,
3JHH = 10.1 Hz, 1H, OCHH), 4.20 (‘‘t”, JHH = 9.2 Hz, 1H, SCNCHt-Bu).
3
3JHH = 8.1 Hz, OCHH). 13C NMR (CDCl3, 101 MHz): d = 9.8 (CH2CH3),
19.9 (CH(CH3)2), 20.0 (CH(CH3)2), 28.3 (CH(CH3)2), 28.6 (CH2CH3),
67.2 (CHEt), 71.8 (OCH2), 171.7 (OCN). IR (cmꢀ1): 2969, 2935,
2878, 1665, 1463, 1199, 1147, 982, 920. TOF MS ES+[M+H]+ Calc.
for C8H16NO: 142.1232; Found: 142.1225. ES-MS m/z (%): 142
(M+H, 100), 126, 98, 88.
13C NMR (CDCl3, 101 MHz): d = 25.78 (C(CH3)2), 25.82 (C(CH3)2),
26.0 (OCNCHC(CH3)3), 26.7 (SCNCHC(CH3)3), 34.1 (C(CH3)3), 34.2
(SCH2), 35.5 (C(CH3)3), 43.3 (C(CH3)2), 69.1 (OCH2), 75.5 (OCNCHt-
Bu), 86.9 (SCNCHt-Bu), 169.4 (OCN), 171.8 (SCN). IR (cmꢀ1): 2954,
2869, 1666, 1627, 1466, 1364, 1130, 979, 925, 755. TOF MS ES+
[M+H]+ Calc. for C17H31N2OS: 311.2157; Found:311.2156. ES-MS
m/z (%): 311 (M+H, 67), 211 (100), 195 (68), 117 (3).
5.5.4. (S)-4-tert-Butyl-2-isopropyl-2-oxazoline (10b)
Was prepared from amide 8b. Yield: 438 mg (2.59 mmol, 55%).
5.5. General synthesis of oxazolines (10)
Rf = 0.20 (Et2O/pentane: 10/90, green in PMA). 1H NMR (CDCl3,
3
400 MHz): d = 0.87 (s, 9H, C(CH3)3), 1.18 (d, JHH = 7.0 Hz, 3H,
Were prepared in two steps (amide synthesis and intramolecu-
lar cyclisation) starting from 2-methyl-propanoylchloride and an
aminoalcohol.
Amide synthesis: The corresponding aminoalcohol (1 equiv.) and
triethylamine (1.1 equiv.) were dissolved in dry dichloromethane
(c = 0.2 M). After dropwise addition of 2-methyl-propanoylchloride
(1 equiv.), thesolutionwasstirred3 h,thenfilteredoversilica(eluent:
EtOAc). Evaporation of solvents afforded the products as white solids.
3
CH(CH3)2), 1.19 (d, JHH = 7.0 Hz, 3H, CH(CH3)2), 2.61 (septet,
3
3
3JHH = 7.0 Hz, 1H, CH(CH3)2), 3.80 (dd, JHH = 10.0 Hz, JHH = 7.2 Hz,
2
3
1H, CHt-Bu), 4.04 (dd, 1H, JHH = 8.6 Hz, JHH = 7.2 Hz, OCHH), 4.12
(dd, JHH = 8.7 Hz, JHH = 10.0 Hz, 1H, OCHH). 13C NMR (CDCl3, 101
MHz): d = 20.0 (CH(CH3)2), 20.1 (CH(CH3)2), 25.8 (C(CH3)3), 28.4
(CH(CH3)2), 33.8 (C(CH3)3), 68.5 (OCH2), 75.4 (CHt-Bu), 171.5
(OCN).
2
3

2508
A. Betz et al. / Journal of Organometallic Chemistry 693 (2008) 2499–2508
(b) A. Pfaltz, M. Lautens, Allylic substitution reactions, in: E.N. Jacobsen, A.
Pfaltz, H. Yamamoto (Eds.), Comprehensive Asymmetric Catalysis II, Springer,
New York, 1999, pp. 833–881.
5.5.5. Typical procedure for the Pd-catalyzed allylic substitution
(E)-1,3-Diphenyl-2-propenyl acetate (156 mg, 0.62 mmol), li-
gand (6 mol%, 0.04 mmol), allylpalladium chloride dimer (6 mg,
2.5 mol%, 0.02 mmol) and potassium acetate (3 mg, 5 mol%,
0.03 mmol) were introduced into a Schlenk flask and 2 mL of dic-
holormethane was added. The mixture was stirred for 20 min be-
fore the successive addition of dimethyl malonate (0.21 mL,
1.86 mmol) and N,O-bis(trimethylsilyl)acetamide (BSA) (0.49 mL,
1.86 mmol). The mixture was stirred at room temperature and
monitored by TLC (SiO2, Et2O/pentane: 20/80). After completion,
solvents were evaporated, the residue dissolved in ether and fil-
tered (syringe filter). After evaporation of ether, the residue was
analysed by HPLC using a Daicel Chiralpak AD analytical column
(n-heptane/2-propanol: 90/10, flow rate: 1 mL/min, 251.0 nm).
Enantiomeric excess of (E)-2-methoxycarbonyl-3,5-diphenylpent-
4-enoate was measured by HPLC; the separation of the two enan-
tiomers was calibrated with racemic product: (R)-enantiomer,
t1 = 12.4 min; (S)-enantiomer, t2 = 17.4 min.
[7] (a) P. von Matt, G.C. Lloyd-Jones, A.B.E. Minidis, A. Pfaltz, L. Macko, M.
Neuburger, M. Zehnder, H. Rüegger, P.S. Pregosin, Helv. Chim. Acta 78 (1995)
265–284;
(b) A. Pfaltz, Acc. Chem. Res. 26 (1993) 339–345;
(c) U. Leutenegger, G. Umbricht, C. Fahrni, P. von Matt, A. Pfaltz, Tetrahedron
48 (1992) 2143–2156;
(d) O. Reiser, Angew. Chem., Int. Ed. Engl. 32 (1993) 547–549;
(e) A. Pfaltz, Acta Chem. Scand. 50 (1996) 189–194.
[8] See for comparison (N,S) chelate ligands in Pd-catalyzed allylic substitution: H.
Adams, J.C. Anderson, R. Cubbon, D.S. James, J.P. Mathias, J. Org. Chem 64
(1999) 8256–8262.
[9] (a) See for comparison sulfanyl-oxazoline ligands: G.J. Dawson, C.G. Frost,
J.M.J. Williams, S.J. Coote, Tetrahedron Lett. 34 (1993) 3149–3150;
(b) J.V. Allen, S.J. Coote, G.J. Dawson, C.G. Frost, C.J. Martin, J.M.J. Williams, J.
Chem. Soc. Perkin Trans. I (1994) 2065–2072.
[10] (a) For some examples of non-symmetric bis(oxazolines) see: H. Nishiyama,
N. Soeda, T. Naito, Y. Motovama, Tetrahedron: Asymmetry 9 (1998) 2865–
2869;
(b) H.A. McManus, P.J. Guiry, J. Org. Chem. 67 (2002) 8566–8573;
(c) I. Atodiresei, I. Schiffers, C. Bolm, Tetrahedron: Asymmetry 17 (2006) 620–
633;
(d) J.I. García, J.A. Mayoral, E. Pires, I. Villalba, Tetrahedron: Asymmetry 17
(2006) 2270–2275;
(e) J.-M. Fraile, J.I. García, A. Gissibl, J.A. Mayoral, E. Pires, O. Reiser, M. Roldán,
I. Villalba, Chem. Eur. J. 13 (2007) 8830–8839.
Acknowledgements
We gratefully acknowledge the ‘‘Ministère de la Recherche et
des Nouvelles Technologies” for the grant to AB, and the CNRS,
the Région Basse-Normandie and the European Union (FEDER
funding) for financial support. Financial support from the Swiss
National Science Foundation (SNF) is gratefully acknowledged.
[11] (a) A. Bulpin, S.L. Roy-Gourvennec, S. Masson, Phosphorus, Sulfur, Silicon
Relat. Elem. 89 (1994) 119–132;
(b) E. Pfund, T. Lequeux, S. Masson, M. Vazeux, Org. Lett. 4 (2002) 843–846.
[12] S.E. Denmark, N. Nakajima, O.J.-C. Nicaise, A.-M. Faucher, J.P. Edwards, J. Org.
Chem. 60 (1995) 4884–4892.
[13] O. Mitsunobu, M. Yamada, Bull. Chem. Soc. Jpn. 40 (1967) 2380–2382.
[14] B.M. Trost, D.J. Murphy, Organometallics 4 (1985) 1143–1145.
[15] For comparison, conversions were: 90% after 30 h with bis(thiazoline) 1c, 20%
after 168 h with the corresponding bis(oxazoline) [3b], and 25% after 168 h
with oxazoline–thiazoline 4c (Table 2, entry 9).
References
[1] (a) For review see: K.A. Lutomski, A.I. Meyers, Asymmetric Synthesis via Chiral
Oxazolines, Asymmetric Synthesis, vol. 3, Academic Press, New York, 1984, pp.
213–274;
[16] However, 1c had better catalytic activity than 4c [3b].
[17] Electronegativities on the Pauling scale; see: L. Pauling, The Nature of the
Chemical Bond, 3rd ed., Cornell University Press, Ithaca, NY, 1960.
[18] (a) See for examples: S. Sakaki, M. Nishikawa, A. Ohyoshi, J. Am. Chem. Soc.
102 (1980) 4062–4069;
(b) A. Pfaltz, J. Heterocycl. Chem 36 (1999) 1437–1451;
(c) H.A. McManus, P.J. Guiry, Chem. Rev. 104 (2004) 4151–4202;
(d) A.K. Ghosh, P. Mathivanan, J. Cappiello, Tetrahedron: Asymmetry 9 (1998)
1–45;
(e) J.S. Johnson, D.A. Evans, Acc. Chem. Res 33 (2000) 325–335;
(f) P. Braunstein, F. Naud, Angew. Chem., Int. Ed. Engl. 40 (2001) 680–
699;
(b) B.M. Trost, M.H. Hung, J. Am. Chem. Soc. 106 (1984) 6837–6839;
(c) F. Delbecq, C. Lapouge, Organometallics 19 (2000) 2716–2723;
(d) K. Szabó, Organometallics 15 (1996) 1128–1133;
(e) S. Sakaki, H. Satoh, H. Shono, Y. Ujino, Organometallics 15 (1996) 1713–
1720;
(g) G. Desimoni, G. Faita, G.K.A. Jørgensen, Chem. Rev. 106 (2006) 3561–
3651.
[2] G. Helmchen, A. Krotz, K.-T. Ganz, D. Hansen, Synlett (1991) 257–259.
[3] (a) I. Abrunhosa, M. Gulea, J. Levillain, S. Masson, Tetrahedron: Asymmetry 12
(2001) 2851–2859;
(f) T.R. Ward, Organometallics 15 (1996) 2836–2838;
(g) J.D. Oslob, B. Åkermark, P. Helquist, P.-O. Norrby, Organometallics 16
(1997) 3015–3021;
(h) M. Svensson, U. Bremberg, K. Hallmann, I. Csöregh, C. Moberg,
Organometallics 18 (1999) 4900–4907;
(b) I. Abrunhosa, L. Delain-Bioton, A.-C. Gaumont, M. Gulea, S. Masson,
Tetrahedron 60 (2004) 9263–9272;
(c) P. Le Maux, I. Abrunhosa, M. Berchel, G. Simonneaux, M. Gulea, S. Masson,
Tetrahedron: Asymmetry 15 (2004) 2569–2573;
(d) B. Fu, D.-M. Du, Q. Xia, Synthesis (2004) 221–226;
(e) S.-F. Lu, D.-M. Du, S.-W. Zhang, J. Xu, Tetrahedron: Asymmetry 15 (2004)
3433–3441;
(i) O.A. Wallner, K.J. Szabó, Chem. Eur. J. 12 (2006) 6976–6983;
(j) D. Tang, X. Luo, W. Shen, M. Li, J. Molec. Struct. THEOCHEM 716 (2005) 79–
87.
[19] R. Ahlrichs, M. Bär, M. Häser, H. Horn, C. Kölmel, Chem. Phys. Lett. 162 (1989)
165–169.
[20] A.D. Becke, Phys. Rev. A 38 (1988) 3098–3100.
[21] J.P. Perdew, Phys. Rev. B 33 (1986) 8822–8824.
(f) D.-M. Du, S.-F. Lu, T. Fang, J. Xu, J. Org. Chem. 70 (2005) 3712–3715;
(g) S.-F. Lu, D.-M. Du, J. Xu, S.-W. Zhang, J. Am. Chem. Soc. 128 (2006) 7418–
7419;
(h) T. Nishio, Y. Kodama, Y. Tsurumi, Phosphorus, Sulfur, Silicon Relat. Elem.
180 (2005) 1449–1450;
[22] A.D. Becke, J. Chem. Phys. 98 (1993) 5648–5652.
[23] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785–789.
[24] D. Andrae, U. Häußermann, M. Dolg, H. Stoll, H. Preuß, Theor. Chim. Acta 77
(1990) 123–141.
[25] ftp://ftp.chemie.uni-karlsruhe.de/pub/basen.
(i) M. Yamakuchi, H. Matsunaga, R. Tokuda, T. Ishizuka, M. Nakajima, T.
Kunieda, Tetrahedron Lett. 46 (2005) 4019–4022;
(j) A. Tárraga, P. Molina, D. Curiel, D. Bautista, Tetrahedron: Asymmetry 13
(2002) 1621–1628;
(k) P. Molina, A. Tárraga, D. Curiel, Synlett (2002) 435–438;
(l) L. Bernardi, M.F. Bonini, M. Comes-Franchini, C. Femoni, M. Fochi, A. Ricci,
Tetrahedron: Asymmetry 15 (2004) 1133–1140;
[26] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
J.A. Montgomery, T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar, 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, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J.
Austin, R. Cammi, K. Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul,
S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, 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 03, revision b.03, Gaussian, Pittsburg, PA, 2002.
[27] A. Klamt, F. Eckert, Fluid Phase Equilibira 172 (2000) 43–72.
[28] A. Schäfter, A. Klamt, D. Sattel, J.C. Lohrenz, F. Eckert, Phys. Chem. Chem. Phys.
2 (2000) 2187–2193.
(m) M. Irmak, T. Lehnert, M.M.K. Boysen, Tetrahedron Lett. 48 (2007) 7890–
7893.
[4] (a) See thiazoline ligands in recent reviews: H. Pellissier, Tetrahedron 63
(2007) 1297–1330;
(b) M. Mellah, A. Voituriez, E. Schultz, Chem. Rev 107 (2007) 5133–5209.
[5] See also effects of the substitution of the oxygen by the sulfur for the couple
oxazole/thiazole: C. Hedberg, K. Källström, P. Brandt, L.K. Hansen, P.G.
Andersson, J. Am. Chem. Soc. 128 (2006) 2995–3001.
[6] (a) See for review: B.M. Trost, D.L. Van Vranken, Chem. Rev 96 (1996) 395–
422;
[29] G. Schaftenaar, J.H. Noordik, J. Comput.-Aided Mol. Des. 14 (2000) 123–134.
[30] A.F. Burchat, J.M. Chong, N. Nielsen, J. Organomet. Chem. 542 (1997) 281–
283.