Chiral ferrocenyl ligands with bidentate pyridine donors. Synthesis and application in Pd-catalyzed asymmetric allylic alkylation of 1,3-diphenylpropenyl-1-esters

Article abstract of DOI:10.1016/j.tetasy.2010.06.015

A synthetic procedure relying on the Friedlaender condensation of enantiopure α-amino ferrocenecarboxaldeyde has been devised for the regio-designed elaboration of a pyridine nucleus fused onto the ferrocene scaffold. Three novel bidentate ligands with different pyridine nitrogen donors featuring the [3,2-b]ferrocenopyridine fragment a as the sole chirogenic element have been prepared in enantiopure form through a multi step route involving the diastereoselective deprotonation of a chiral acetal of ferrocenecarboxaldehyde in the stereodetermining step. The ligands were assessed in the Pd-catalyzed allylic alkylation of 1,3-diphenyl-2-propenyl esters with good stereoselectivity.

Full text of DOI:10.1016/j.tetasy.2010.06.015

Tetrahedron: Asymmetry 21 (2010) 1921–1927
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Chiral ferrocenyl ligands with bidentate pyridine donors. Synthesis
and application in Pd-catalyzed asymmetric allylic alkylation of
1,3-diphenylpropenyl-1-esters
Agnieszka Mroczek, Giulia Erre, Rossana Taras, Serafino Gladiali *
Dipartimento di Chimica, Università di Sassari, via Vienna 2, 07100 Sassari, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
A synthetic procedure relying on the Friedländer condensation of enantiopure a-amino ferro-
Received 29 April 2010
Accepted 10 June 2010
Available online 6 July 2010
cenecarboxaldeyde has been devised for the regio-designed elaboration of a pyridine nucleus fused onto
the ferrocene scaffold. Three novel bidentate ligands with different pyridine nitrogen donors featuring
the [3,2-b]ferrocenopyridine fragment a as the sole chirogenic element have been prepared in enantio-
pure form through a multi step route involving the diastereoselective deprotonation of a chiral acetal
of ferrocenecarboxaldehyde in the stereodetermining step. The ligands were assessed in the Pd-catalyzed
allylic alkylation of 1,3-diphenyl-2-propenyl esters with good stereoselectivity.
Ó 2010 Elsevier Ltd. All rights reserved.
1
. Introduction
Planar chirality in chiral ligands with pyridine nitrogen donors
is quite rare and there are a few examples of such a kind of deriv-
atives in the current literature. The cyclophane bipyridine 1 and
the related pyridyl quinolinophane 2 (Fig. 1) were isolated in low
yield at the end of an elaborate synthetic procedure involving
enantiomeric resolution.1a Both ligands were tested in Cu-cata-
lyzed asymmetric cyclopropanation of styrene and in Ir-catalyzed
asymmetric H-transfer reduction of acetophenone with modest re-
N
N
N
N
R
N
X
1
2
3
3
a:
b:
X=OH
1
b
sults (25–30% ee). A similar quinolinophane template was used
X=PPh2
2
as the stereogenic scaffold of N,O- and P,N-bidentate ligands 3. Li-
gand 3a gave excellent ee’s (99%) in the diethylzinc addition to
Figure 1.
2
a
aldehydes while the corresponding P,N-derivatives were much
less efficient in the Pd-catalyzed allylic alkylation.2b
and the relevant N-oxide 4b (Fig. 2) are effective chiral inducers
The large majority of ligands featuring a stereogenic plane as
the chiral element contain in their structure a ferrocene unit which
has been properly elaborated as to become the source of chirality
of the molecule. The combinations of donors appended onto the
enabling ee’s higher than 90% in a range of asymmetric reactions.4
2
A few years later, the same group reported the synthesis of the C -
symmetrical chiral ferrocenyl bipyridine 4c which was assessed in
the Cu-catalyzed cyclopropanation of styrene with very good results
To the best of our knowledge, this is the sole bidentate
pyridine nitrogen ligand obtained in enantiopure form whose planar
chirality comes from a stereogenic ferrocenyl framework. The rela-
tively similar 2,2 -biquinoline 4d has been obtained but only in
racemic form along with the meso diastereoisomer.
ferrocene template for meeting this purpose are the most varied,
5
a
phosphorus derivatives being the most popular choice.3 Less fre-
(94% ee).
quent is the case where these ligands feature only nitrogen atoms
as donor centers and even rarer are the chiral ferrocenylpyridines
with the nitrogen atom directly linked onto the cyclopentadienyl
ring. The efficiency of ferroceno-condensed planar-chiral pyridine
ligands in asymmetric catalysis has been first pointed out by Fu
several years ago when he showed that the DMAP-surrogate 4a
0
5b
Pursuing our long standing interest in chiral pyridine ligands,6
some time ago we thought it worthwhile to expand the structural
diversity of ligands of this type with planar chirality. To this purpose,
we set out to design a synthetic strategywhich could pave the way to
a modular procedure useful for the stereoselective preparation of a
library of pyridine ligands featuring the [3,2-b]ferrocenopyridine
*
Corresponding author. Tel.: +39 079 229546; fax: +39 079 229559.
E-mail address: gladiali@uniss.it (S. Gladiali).
0
957-4166/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2010.06.015

1
922
A. Mroczek et al. / Tetrahedron: Asymmetry 21 (2010) 1921–1927
Me
Fe
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Fe Me
Fe
NMe2
N
Fe
Me
Me
N
Me
N
N
Me
O
Me
S)
-4a
(S)-4b
(+)-(S,S)-4c
(
N
Fe
N
N
a
4
d
CpFe
FeCp
Figure 2.
fragment a as the common chirogenic element. Herein we report on
the synthesis of the first three ligands of this category and on their
application in the Pd-catalyzed asymmetric allylic alkylation of
N
Fe
N
1
,3-diphenyl-2-propenyl esters.
N
O
9
a
(S)-5
51 %
2
2
. Results and discussion
a
.1. Preparation of ligands
The main goal of this investigation resided in establishing a
straightforward procedure for the preparation in enantiopure form
of the planar-chiral tag a distinctive of this category of ligands. At
the same time the synthetic procedure had to meet the request of
making readily accessible a library of bidentate ligands of modular
diversity featuring a second pyridine donor in a predetermined po-
sition suitable for supporting the chelate binding of the ligand to a
metal center. Among the synthetic strategies amenable for elabo-
rating a fused pyridine ring onto one cyclopentadienyl fragment
N
CHO
NH2
O
Fe
N
9
b
N
Fe
a
(S)-6
52 %
(
S)-8
O
7
of the ferrocene template, the Friedländer condensation was
deemed the best suited tool for the synthesis of the targeted li-
N
gands 5, 6, and 7 with the desired directional control. From our
8
previous experience in the field, this condensation provides good
9
c
yields of the desired pyridine derivatives under reasonably mild
conditions and allows to introduce a wide structural diversity.
We were reasonably confident that such operative conditions
might be compatible with the utilization of an enantiopure 1,2-
disubstituted ferrocene with a stereogenic plane as the chiral
synthon while tolerating the presence of different structural deco-
rations in the condensation partner. Retrosynthetic analysis of the
Friedländer condensation led us to single out the enantiopure
N
Fe
a
N
(
S)-7
54 %
Scheme 1. Reagents and conditions: (a) KOH, EtOH, 25 °C, 4–12 h.
a-amino ferrocenecarboxaldeyde 8 as a common chiral intermedi-
approach is best suited for ferrocenes substituted with an achiral
ate providing the stereogenic plane and the pyridylketones 9a–c as
the suitable partners for the preparation of our targets (Scheme 1).
The asymmetric synthesis of enantiopure 1,2-disubstituted fer-
rocenes is most frequently accomplished according to two basic
strategies both relying on heteroatom-assisted diastereoselective
ortho-directing group and relies on an external source of chirality
1
0
such as (ꢀ)-sparteine that can promote the formation of diaste-
reomeric complexes at the level of the lithiated ferrocene.
Notably, the preparation of our key intermediate 8 was reported
for the first time several years ago by Kagan et al. according to the
9
11
deprotonation. In the first one, diastereoselectivity is achieved
reaction sequence in Scheme 2.
due to the presence on the ferrocenyl template of a chiral substitu-
ent aiding the recognition of the diastereotopic vicinal proton to be
selectively extracted by the base. Extensive use of temporary chiral
auxiliaries such as acetals, oxazolines, amines, and sulfoxides has
been exploited more or less successfully in this task. The second
In that seminal paper, the installation of the amino group on the
vicinal carbon of ferrocenecarboxaldeyde was achieved with com-
plete diastereoselectivity through a six-step synthetic route. Key
steps of that synthesis were the heteroatom-assisted deprotona-
tion of a temporary chiral acetal and the quenching of the relevant

A. Mroczek et al. / Tetrahedron: Asymmetry 21 (2010) 1921–1927
1923
O
O
OR
CHO
CHO
NH2
O
OR
O
a, b
c
d
e
f
X
Fe
Fe
Fe
Fe
1
0
11a: R=H
12a: X=N3
8
1
1b: R=CH3
12b: X=NH2
+
+
Scheme 2. Reagents and conditions: (a) trimethylorthoformate, H , MeOH, 80 °C, 3 h; 96%; (b) (S)-(ꢀ)-1,2,4-butanetriol, H , CHCl
c) NaH, CH I, THF, 25 °C, 12 h; 99% 11b; (d) t-BuLi, Et O, ꢀ78 °C, TosN , THF, 25 °C; 75–85% 12a; (e) NaBH , Bu NHSO , CH Cl , H
5 °C, 30 min; 97%.
3
, 4 Å molecular sieves, 25 °C, 12 h; 97% 11a;
2
O, 25 °C, 12 h; 74% 12b; (f) HCl 10%, CH Cl ,
(
2
3
2
3
4
4
4
2
2
2
2
mixed cyanocuprate with N,O-bis(trimethylsilyl)hydroxylamine,
an electrophilic source of nitrogen. The -aminoaldehyde 8 was
eventually obtained via the intermediate amino acetal 12b by
one. The condensation of these two reagents was best accom-
plished under acid rather than under basic catalysis.14 Thus,
refluxing overnight a THF solution containing a 5:1 mixture of
diketone and aminoaldehyde in the presence of 5% sulfamic acid
gave in one step the desired ketone in 36% isolated yield after chro-
matographic purification. This preparation is straightforward and
compares favorably with the multistep procedures reported in
the literature.
a
hydrolytic work up.
The advantage of using a chiral acetal as the ortho-directing
group is immediately apparent in our case since the required formyl
group can be regenerated under very mild conditions that are com-
patible with the presence of the delicate combination of functional-
ities displayed by 8. Thus, transacetalization of the dimethyl acetal
of ferrocenecarboxaldehyde 10 with (S)-(ꢀ)-1,2,4-butanetriol,
2.2. Palladium-catalyzed allylic alkylation
1
2
either commercially available or prepared from (S)-malic acid,
gave the chiral 1,3-dioxane 11a (90% yield) in perfect agreement
with the literature. Methylation of the free hydroxyl group was
The utility of chiral bidentate pyridine donors in metal-cata-
lyzed enantioselective catalysis is well established and their effi-
then performed using NaH and CH
3
I in THF to give the desired ace-
ciency has been expressed in a wide number of reactions
1
5
tal 11b (99% yield). The following step was the most critical of the
entire sequence. In our hands the methodology devised by Kagan
gave low and erratic yields never higher than 30%. In spite of our
numerous attempts the yields of this reaction could not be stabi-
lized at a satisfactory level and this forced us to search for a more
performant and reliable electrophilic source of nitrogen. Many ef-
forts have been spent to solve this problem and a wide range of
nitrogen reagents (diazonium salts, diphenyl phosphoryl azide,
diethyl azodicarboxylate, etc.) have been screened for this purpose.
At the end of this study tosylazide resulted as the best reagent for
the introduction of an amino group synthon onto the ortho-posi-
tion. Thus, treating the lithiated acetal with tosylazide at ꢀ78 °C
and further stirring at room temperature overnight lead to the for-
mation of the azidoacetal 12a. From 1H NMR the azido derivative
accounts for 75–85% of the crude product and is obtained as a single
diastereomer. The introduction of the azido group was confirmed
centered on a variety of different metals. For screening the po-
tential of ligands 5, 6, and 7 as chiral inducers we selected two
benchmark asymmetric reactions: the Cu-catalyzed cyclopropana-
tion of styrene by diazoacetate and the Pd-catalyzed allylic
alkylation of racemic 1,3-diphenyl-2-propenyl acetate 13 with di-
methyl malonate anion. In the first process C
2
-symmetrical bipyri-
consistently outperform the
-symmetric counterparts pro-
1
6
dines such as 4c, 15, and 16
analogue phenanthrolines and the C
1
viding ee’s in the range of 90% ee. In the Pd-catalyzed allylic alkyl-
ation the opposite trend is usually observed. In this reaction
1
7
phenanthrolines such as 17 and 18 are comparably better chiral
inducers than bipyridines and C -symmetry is apparently better
suited for high ee’s to be attained.
1
When ligands 5 and 6 were tested in the first reaction very low
catalytic activities and negligible ee’s were observed. The results
obtained in the second asymmetric process were more gratifying
ꢀ1
by the presence of a strong band at 2127 cm in the IR spectrum
of a sample of 12a purified by flash chromatography.
and are detailed below. The active catalyst was generated in situ
3
from [PdCl(
g
-C
H
3 5
)]
2
and the appropriate chiral ligand (Scheme 3)
Due to its poor stability the crude azido acetal 12a was directly
subjected to reduction with NaBH in biphasic conditions to pro-
4
while the nucleophile was generated by the combined action of
BSA with a catalytic amount of an inorganic base.
vide the corresponding amino derivative 12b in 70–75% isolated
yield after flash chromatography. Removal of the acetal protection
was quantitatively obtained by stirring a dichloromethane solution
of 12b with dilute hydrochloric acid for 30 min at room tempera-
ture. Neutralization of the aqueous solution gave the desired ami-
no aldehyde 8 pure enough as to be used straight in the Friedländer
condensation. This last step was performed by stirring overnight at
room temperature an ethanolic solution of the aminoaldehyde
with the suitable enolizable pyridyl ketone 9 in the presence of
potassium hydroxide as a condensing agent. Under these condi-
tions the aminoaldehyde undergoes some decomposition and a
The first set of results shown with ligands 5 and 6 are reported
in Tables 1 and 2.
For gaining an insight in the coordination mode of our ligands
during the catalytic reaction, the effect of the variation of the Pd/
ligand (S)-5 ratio was checked first. In dichloromethane reducing
the ratio of the Pd/ligand from 1:2 to 1:1 had no effect on rate
and stereoselectivity of the reaction. A further decrease of the rel-
ative amount of ligand to 0.25 brought about only a modest drop
of rate and selectivity (Table 1, entries 1–3). This was taken as an
indication that in this range of concentrations a chelate Pd-com-
plex is the operating catalyst. Addition of AgPF
ing the potential interference of the chloride anion
6
aiming at remov-
18
5
0% molar excess was found appropriate in order to counterbal-
ance its poor stability. With this trick the targeted ligands 5 and
could be obtained after chromatographic purification in isolated
had a
negative effect on the catalytic activity which was almost
suppressed.
6
yield as high as 50–55%. For the Friedländer synthesis of ligand 7,
the required partner was 2-acetylquinoline 9c.13 Keeping strictly to
the Friedländer chemistry we elaborated a straightforward route
for its preparation from 2-aminobenzaldehyde and 2,3-butanedi-
Conversion and stereoselectivity of the reaction were strongly
dependent on the solvent and on the operative conditions and
the reaction times required for
a complete conversion to
be achieved were quite different. With both ligands 5 and 6

1
924
A. Mroczek et al. / Tetrahedron: Asymmetry 21 (2010) 1921–1927
OAc
Ph
CH(COOCH3)2
Ph
3
[
Pd(η -C H )Cl] / (S)-L*
3 5 2
Ph
Ph
CH2(COOMe)2 / BSA, salt
1
3
14
N
N
N
N
OCH3
H3CO
R
R
1
6
1
5
N
H
N
N
N
17
18
Scheme 3.
Table 1
Allylic alkylation of 1,3-diphenylprop-2-enyl acetate with dimethyl malonate in the presence of Pd/(S)-5
3
[
Pd(η -C3H5)Cl]2 / (S)-5
OAc
Ph
CH(COOCH3)2
Ph
CH2(COOMe)2 / BSA
A=KOAc
B=LiOAc
Ph
Ph
o
A or B, solvent, 25 C
13
(S)-14
Entry
Solvent
Time
Conv. (%)
ee (%)
AcOK
1
CH
CH
CH
CHCl
THF
THF
2
2
2
Cl
Cl
Cl
2
2
2
5 min
5 min
20 min
48 h
78 h
4 h
100
100
100
50
51 (S)
50 (S)
45 (S)
49 (S)
37 (S)
47 (S)
50 (S)
47 (S)
a
2
3
b
4
5
3
a
40
6
7
100
100
100
a
Toluene
Toluene
15 min
10 min
8
AcOLi
a
9
CH
2
Cl
2
45 h
45 h
48 h
45 h
32 h
30 min
90
25
50
22
98
58 (S)
64 (S)
66 (S)
75 (S)
52 (S)
52 (S)
1
1
1
1
1
0a
1
2
3
4
CHCl
CHCl
CHCl
THF
3
3
3
c
a
Toluene
100
3 5 2
Pd(g H )Cl] 0.005 mmol, (S)-5 0.02 mmol, substrate 0.2 mmol, dimethyl malonate 0.6 mmol, BSA 0.6 mmol, acetate salt
3-C
[
0
.006 mmol, room temperature.
a
Ratio Pd/L = 1:1.
Ratio Pd/L = 1:0.25.
b
c
Temperature = 0 °C.
chlorinated solvents performed slightly better than oxygenated
ones or hydrocarbons (Tables 1 and 2), the difference being more
pronounced with lithium acetate as a base. With this base, a dra-
matic decrease of the reaction rate accompanied by a moderate in-
crease of the stereoselectivity was observed in dichloromethane
and THF (Table 1, compare entries 2 and 9 and entries 6 and 13;
Table 2, compare entries 1 and 5 and entries 3 and 8).
The reaction was even slower in chloroform but this solvent
turned out to be consistently more effective than dichloromethane
in terms of selectivity (Table 1, compare entries 9 and 10; Table 2,
compare entries 5 and 6). A definite improvement of the enantiose-
lectivity of the reaction was obtained by lowering the temperature
to 0 °C (Table 1, compare entries 11 and 12; Table 2, compare entries
6 and 7). When the reaction was performed at this temperature in
chloroform, 75% ee was scored with the bipyridine derivative 5
(Table 1, entry 12) and 68% ee with the dihydrophenanthroline
ligand 6 (Table 2, entry 7). This result follows in keeping with the
general trend observed in this study where the dihydrophenanthro-
line derivative was consistently outperformed by the bipyridine
ligand.

A. Mroczek et al. / Tetrahedron: Asymmetry 21 (2010) 1921–1927
1925
Table 2
Notably, complete conversions were observed only with
Allylic alkylation of 1,3-diphenylprop-2-enyl acetate with dimethyl malonate in the
presence of in situ-generated Pd/(S)-6
CH
CH
3
COONa and Na
2 3
CO for (S)-5 (Table 3, entries 5 and 6) and with
3
COOK for (S)-6 (Table 3, entry 8).
Entry
Solvent
Time
Conv. (%)
ee (%)
On the contrary, ligand (S)-7 gave complete conversion with
AcOK
1
2
3
4
both CH
excess.
3 3
COOLi and CH COOK, but with very low enantiomeric
CH
CH
THF
Toluene
2
Cl
Cl
2
2
45 min
45 min
45 min
1.2 h
100
100
100
100
40 (S)
46 (S)
42 (S)
52 (S)
a
2
The influence of the leaving group of the substrate has been in-
spected by exchanging the acetate ester of 13 with the methoxy-
carbonate of 19. This variation did not substantially alter the
stereoselectivity, as almost identical ee’s were obtained with both
the esters (Table 4).
AcOLi
5
CH
2
Cl
2
45 h
45 h
48 h
32 h
2 h
54
25
50
44
100
53 (S)
59 (S)
68 (S)
49 (S)
40 (S)
a
6
CHCl
CHCl
THF
3
b
7
3
8
9
3
. Conclusions
Toluene
[
Pd(
g
3-C
3
H
5
)Cl]
2
0.005 mmol, (S)-6 0.02 mmol, substrate 0.2 mmol, dimethyl
In the course of this investigation we have elaborated and vali-
malonate 0.6 mmol, BSA 0.6 mmol, acetate salt 0.006 mmol, room temperature.
dated a practical procedure useful for the modular synthesis of fer-
roceno-condensed planar-chiral pyridine ligands featuring as the
unique chiral element a stereogenic plane arising from the ferro-
cene scaffold. This methodology has been successfully exploited
in the preparation of one pyridine, one quinoline, and one phenan-
throline ligand, 5, 6, and 7, respectively, using enantiopure 2-amin-
oferrocenecarboxaldeyde (S)-8 as the common chiral synthon for
a
Ratio Pd/L = 1:1.
Temperature = 0 °C.
b
In the presence of both potassium and lithium acetate the com-
plex obtained from the quinolyl derivative (S)-7 displayed a good
catalytic activity. Quantitative conversions were obtained within
1
all these derivatives. These C -symmetry bidentate ligands have
2
–3 h in dichloromethane with KOAc and, more surprisingly, even
been assessed in the palladium-catalyzed alkylation of rac-1,3-di-
phenyl-2-propenyl esters 13 and 19 with dimethyl malonate. The
outcome of the reaction is strongly dependent on the solvent, the
temperature, and the nature of the cation of the salt used as a basic
promoter. Under optimized conditions both ligands 5 and 6 gave
similar stereoselectivities of (S)-alkylated product 14, the top value
being 75% ee. In the same reaction ligand 7 produced an active cat-
alyst that led to 100% conversions, but with very low enantiomeric
excess (22%).
with LiOAc in chloroform but the stereoselectivities for (S)-14
were quite low, 12% and 22%, respectively. When this ligand
was used in THF the catalytic activity collapsed (9% conversion
after 48 h).
The remarkable differences observed with potassium and lith-
ium acetate prompted us to explore in more detail the influence
of the salt on the enantioselectivity and the conversion when the
reaction is run in chloroform (Table 3). As it is apparent from
Table 3, the reaction is pretty sensitive both to the cationic and
the anionic components of the base. The effect of the anion is just
moderate and acetate is generally a better suited base than carbon-
ate, enabling higher ee’s to be obtained. On the contrary, the nature
of the alkaline cation has a profound influence on the reaction rate
which spans over two order magnitude when moving from the
most to the least efficient salt sodium (entry 5) and cesium acetate
4. Experimental section
4.1. General
Nuclear magnetic resonance spectra were obtained on a Varian
1
31
(
entry 7), respectively. Even if of some significance, the fluctuations
VXR-5000 spectrometer at 300 MHz for H, 121.42 MHz for P and
1
3
of stereoselectivity induced by the different cations are in a more
restricted range. The best ee’s of this set were obtained with
CH COOLi (Table 3, entries 3 and 9), however, with incomplete
3
75 MHz for C. Chemical shifts were reported in ppm downfield
for H and C and from H PO for
3 3 4
P NMR. Optical rotations were measured on a Jasco P-1010 polar-
1
13
from internal Me
4
Si in CDCl
3
1
conversion even after a prolonged reaction time. In the presence
of other alkaline cations the selectivity was slightly lower.
imeter. Melting points were obtained with a Buchi 530 melting
point apparatus. For column chromatography silica gel
Table 3
Allylic alkylation of 1,3-diphenylprop-2-enyl acetate with dimethyl malonate in the presence of in situ-generated Pd/(S)-5 or Pd/(S)-6:
effect of salt
3
[
Pd(η -C H )Cl] / (S)-5/6
3
5
2
OAc
Ph
CH(COOCH3)2
Ph
CH2(COOMe)2 / BSA
salt, CHCl3, 25 o
Ph
Ph
C
1
3
(S)-14
Entry
Ligand
Salt
KOAc
CO
LiOAc
Li CO
NaOAc
Na CO
Cs CO
Time (h)
Conv. (%)
ee (%)
1
2
3
4
5
6
7
8
9
(S)-5
(S)-5
(S)-5
(S)-5
(S)-5
(S)-5
(S)-5
(S)-6
(S)-6
(S)-6
48
48
48
48
1.5
1.5
48
8
50
84
50
49 (S)
54 (S)
66 (S)
59 (S)
60 (S)
56 (S)
52 (S)
43 (S)
59 (S)
51 (S)
K
2
3
2
3
68
100
100
32
100
50
2
3
2
3
KOAc
LiOAc
36
48
10
Li
2
CO
3
45
General experimental conditions [Pd(
BSA 0.6 mmol, salt 0.006 mmol, room temperature, CHCl
g
3-C
3
H
5
)Cl]
2
0.005 mmol, (S)-5 /6 0.02 mmol, substrate 0.2 mmol, dimethyl malonate 0.6 mmol,
2 ml.
3

1
926
A. Mroczek et al. / Tetrahedron: Asymmetry 21 (2010) 1921–1927
Table 4
Allylic alkylation of 1,3-diphenylprop-2-enyl carbonate in the presence of in situ-generated Pd/(S)-5 or Pd/(S)-6: effect of the salt
O
3
[
Pd(η -C3H5)Cl]2 / (S)-5/6
CH(COOCH3)2
Ph
O
OCH3
CH2(COOMe)2 / BSA
Ph
Ph
Ph
o
salt, CHCl3, 25 C
1
9
(
S)-14
Entry
Ligand
Salt
Conv. (%)
ee (%)
1
2
3
(S)-5
(S)-5
(S)-5
(S)-6
KOAc
K CO
2 3
LiOAc
LiOAc
64
64
100
71
57 (S)
54 (S)
68 (S)
56 (S)
a
4
General experimental conditions [Pd(
g
3-C
3
2 ml, 48 h.
3 5 2
H )Cl] 0.005 mmol, (S)-5/6 0.02 mmol, substrate 0.2 mmol, dimethyl malonate 0.6 mmol,
BSA 0.6 mmol, salt 0.006 mmol, CHCl
a
Time: 7 h.
(
230–400 mesh, Merk 60) and neutral Al
2
O
3
(70–230 mesh, Merck)
then diluted with equal volume of brine (36 ml), extracted three
times withdichloromethane, dried (Na SO ), filtered and the solvent
was evaporated. The crude was analyzed by H NMR. The analysis
showed the presence of the desired product 12b, some unreacted
starting material 12a, and a trace of the parent acetal 11b. The pure
were used, according to the literature. All commercial reagents
were used without further purification and solvents were distilled
after refluxing under nitrogen or argon and stored under an inert
atmosphere until used. (2S,4S)-4-(Methoxymethyl)-2-ferrocenyl-
2
4
1
1
1
1
,3-dioxane 11b was prepared as reported in the literature.
product 12b (3.8 g, 74%) was isolated after flash chromatography on
1
silica gel (AcOEt:Et
2
O:Et
3
N = 1:4:0.1). H NMR (CDCl
3
) d 5.40 (s, br,
4
.2. Synthesis of (2S,4S,SFc)-4-(methoxymethyl)-2-( -azidoferro-
a
1H), 4.22 (m, 2H), 4.13 (s, 5H), 4.04 (m, 1H), 3.91 (m, 2H), 3.81 (s,
br, 1H), 3.52 (m, 2H), 3.45 (s, 3H), 3.03 (m, br, 2H), 1.83 (m, 1H),
cenyl)-1,3-dioxane 12a
1
3
1
.51 (m, 1H). C NMR (CDCl
3
) d 101.1, 76.2, 75.8, 73.2, 69.8, 66.7,
ꢀ
1
(
2S,4S)-4-(Methoxymethyl)-2-ferrocenyl-1,3-dioxane 11b
63.2, 62.1, 59.6, 58.5, 28.1. IR (KBr) cm 3422, 3343.
(
4.39 g, 13.88 mmol) was dissolved in ether (72 ml, 5–10 ml of
ether per 2 mmol of acetal) and stirred at ꢀ78 °C for 30 min under
argon. t-BuLi (1.1 equiv as a 1.5 M solution, 13.9 ml, 20.82 mmol)
was added dropwise, yielding after a few minutes a bright yellow
precipitate. After 30 min of stirring at ꢀ78 °C, the mixture was al-
lowed to warm up to room temperature and stirring was continued
until a bright orange precipitate of lithiated acetal gradually
formed. The mixture was cooled again to ꢀ78 °C and at this tem-
perature a solution of anhydrous tosylazide (3.55 g, 18.01 mmol)
in THF (36 ml) was added through cannula. The mixture was
warmed up to room temperature and left stirring overnight result-
ing in a bright red solution. The reaction was quenched with water
and the solvent was removed under reduced pressure. The result-
ing slurry was diluted with water and extracted three times with
dichloromethane. The combined organic phases were dried
4.4. Synthesis of (S)-a-aminoferrocenecarboxaldehyde 8
At first, 10% HCl (17 ml) was added to a solution of (2S,4S,SFc)-4-
(methoxymethyl)-2-( -aminoferrocenyl)-1,3-dioxane 12b (0.3 g,
0.9 mmol) in CH Cl (17 ml) in a separatory funnel and shaken vig-
a
2
2
orously. The organic layer, initially light yellow, becomes immedi-
ately colorless while the aqueous phase becomes dark red,
indicating the almost instantaneous deprotection. The aqueous
phase was separated, made alkaline with NaOH 10%, and extracted
with CH Cl . The organic phase was dried (Na SO ), filtered, and
2 2 2 4
evaporated under reduced pressure. The resulting red-colored
product 8 (0.2 g, 97%) was used directly in the next step without
1
further purification. H NMR (CDCl
3
) d 9.99 (s, 1H), 4.41 (m, 1H),
4.33 (m, 1H), 4.20 (m, 1H), 4.16 (s, 5H), 3.83 (br s, 2H), identical
1
1
(
Na
SO
2 4
) and the solvent was evaporated under reduced pressure
with the data reported by Kagan.
to give crude 12a as a dark oil (5.3 g). NMR analysis of this product
showed the presence of the desired product 12a (75–85%) impure
of starting material 11b and excess of tosylazide. The crude com-
pound can be used as such in the next step. A sample purified by
flash chromatography showed in the IR spectrum (KBr) an intense
0
4.5. Synthesis of (S)-6-(2 -pyridyl)[3,2-b]ferrocenopyridine 5
Saturated ethanolic KOH (1.78 ml) was added to a suspension of
ferrocenyl amino aldehyde 8 (0.5 g, 2.18 mmol) and 2-acetylpyri-
dine 9a (0.15 g, 1.75 mmol) in absolute ethanol (40 ml) under ar-
gon. The mixture was stirred for 12 h at room temperature. Then
the reaction was quenched with water and the solvent was evapo-
rated. The resulting slurry was diluted with water and extracted
ꢀ1
stretching band at 2127 cm diagnostic for the presence of the
1
azido group. H NMR (CDCl
(
(
3
) d 5.47 (s, br, 1H), 4.30 (m, 2H), 4.28
s, 5H), 4.20–4.17 (m, 2H), 4.10–3.96 (m, 2H), 3.54 (m, 2H), 3.82
s, 3H), 1.82 (m, 1H), 1.51 (m, 1H).
2 2
three times with CH Cl . The combined organic phases were dried
4
.3. Synthesis of (2S,4S,SFc)-4-(methoxymethyl)-2-(a-aminoferro-
(Na SO ) and the solvent was evaporated under reduced pressure.
2
4
cenyl)-1,3-dioxane 12b
The residue was purified by flash chromatography on neutral alu-
mina with EtOAc yielding 5 (0.35 g, 51%) as a green solid. H NMR
1
At first, NaBH
added to the mixture of (2S,4S,SFc)-4-(methoxymethyl)-2-(
errocenyl)-1,3-dioxane 12a (5.5 g,14.14 mmol) in CH Cl
and Bu NHSO
argon. The resulting two-phase solution was vigorously stirred at
room temperature overnight. The reaction was monitored by TLC
4
(2.14 g, 56.56 mmol) in distilled H
2
O (14 ml) was
3
(CDCl ) d 8.69 (d, 1H, J = 4.8 Hz), 8.47 (d, 1H, J = 8.1 Hz), 8.28 (d, 1H,
a-azidof-
J = 9.3 Hz), 7.96 (d, 1H, J = 9.0 Hz), 7.84 (t, 1H, J = 6.9 Hz), 7.35 (t,
1H, J = 6.0 Hz), 5.46 (s, 1H), 4.95 (s, 1H), 4.26 (s, 1H), 3.90 (s, 5H).
2
2
(23 ml)
1
3
4
4
(1.92 g, 5.65 mmol) at room temperature under
C NMR (CDCl
121.9, 116.8, 108.9, 80.8, 71.7, 68.9, 63.8, 60.1. Anal. Calcd for
14FeN : C, 68.82; H, 4.49; N, 8.92. Found: C, 68.70; H, 4.65;
N, 9.14.
3
): d 159.5, 157.2, 149.3, 141.8, 137.1, 124.1,
C
18
H
2
2 3
on silica gel (Et O/AcOEt/Et N = 4:1:0.1). The reaction mixture was

A. Mroczek et al. / Tetrahedron: Asymmetry 21 (2010) 1921–1927
1927
4
.6. Synthesis of (S)-[2,3-b]-ferrocenyl-5,6-dihydrophenantro-
BSA (0.6 mmol, 122 mg) and a small amount of potassium acetate
(or other salt). The mixture was stirred at the given temperature,
line 6
2
while observing the conversion by tlc (silica, Et O/petroleum
Saturated ethanolic KOH (2.86 ml) was added to the solution of
ferrocenyl amino aldehyde 8 (0.8 g, 3.49 mmol) and 6,7-dihydro-
H-quinolin-8-one 9b (0.32 g, 2.8 mmol) in absolute ethanol
55 ml) under argon. The mixture was stirred for 12 h at room tem-
ether = 1:3). After the reaction was complete, the mixture was di-
luted with dichloromethane and washed with saturated aqueous
ammonium chloride solution. The organic phase was dried
5
(
2 4
(Na SO ) and filtered and the solvent was evaporated under re-
perature. Then the reaction was quenched with water and the sol-
vent was evaporated. The resulting slurry was diluted with water
duced pressure. Conversion was determined by NMR on this resi-
due. The crude product was purified by column chromatography
and extracted three times with CH
2
Cl
2
. The combined organic
2
on silica gel (Et O/petroleum ether = 1:3) to give 14. Ee’s of the
phases were dried over (Na SO ) and the solvent was evaporated
under reduced pressure. The residue was purified by flash chroma-
2
4
product were determined from the integrals of the methoxy groups
of (1,3-diphenylprop-2-enyl)malonate, as split by the chiral shift
reagent europium(III) tris[3-(heptafluoropropylhydroxymethyl-
ene)-d-camphorate]. The absolute configuration was assigned by
comparison of the sign of the measured value of the specific rota-
tion with the published data.
tography on neutral alumina with EtOAc yielding 0.62 g of the ex-
1
pected product 6 (52%). H NMR (CDCl
3
) d 8.78 (d, 1H, J = 4.2 Hz),
7
.94 (s, 1H), 7.57 (d, 1H, J = 7.5 Hz), 7.28 (s, 1H), 5.58 (s, 1H), 4.90
13
(
s, 1H), 4.22 (s, 1H), 3.89 (s, 5H), 2.96 (m, 4H). C NMR (CDCl
d 155.6, 152.4, 149.5, 137.3, 136.3, 134.9, 128.4, 123.8, 108.9,
1.2, 71.5, 68.7, 64.4, 59.3, 29.9, 28.6. Anal. Calcd for C20 16FeN
C, 70.61; H, 4.74; N, 8.23. Found: C, 70.42; H, 4.65; N, 8.14.
3
):
8
H
2
:
Acknowledgment
Financial support from Marie Curie INDAC-CHEM project and
MUR through the contract PRIN 2007HMTJWP ‘Product oriented
chemo- and stereo-selective syntheses by innovative transition
metal catalysts’ is gratefully acknowledged by S.G. and A.M.
4
.7. Synthesis of 2-acetylquinoline 9c
At first, 5% sulfamic acid (0.02 g, 0.21 mmol) was added to a
solution of 2-aminobenzaldehyde (0.5 g, 4.13 mmol) and 2,3-
butanedione (1.78 g, 20.65 mmol) in THF (15 ml) under argon.
The mixture was refluxed overnight, then the reaction was
quenched with aqueous sodium bicarbonate and the solvent was
evaporated. The resulting slurry was taken up with water and ex-
References
1. (a) Wörsdörfer, U.; Vögtle, F.; Glorius, F.; Pfaltz, A. J. Prakt. Chem. 1999, 341,
445–448; (b) Wörsdörfer, U.; Vögtle, F.; Niegler, M.; Waletzke, M.; Grimme, S.;
Glorius, F.; Pfaltz, A. Synthesis 1999, 597–602.
tracted three times with CH
dried (Na SO ) and the solvent was evaporated under reduced pres-
2 2
Cl . The combined organic phases were
2.
(a) Ricci, G.; Ruzziconi, R. Tetrahedron: Asymmetry 2005, 16, 1817–1827; (b)
Ruzziconi, R.; Santi, C.; Spizzichino, S. Tetrahedron: Asymmetry 2007, 18, 1742–
1749.
2
4
sure. The residue was purified by column chromatography on neu-
tral alumina (petroleum ether/EtOAc = 8:2) to give the expected
3.
(a) Hayashi, T.; Togni, A. Ferrocenes; VCH: Weinheim, 1995; (b) Blaser, H.-U.;
Brieden, W.; Pugin, B.; Spindler, F.; Studer, M.; Togni, A. Top. Catal. 2002, 19, 3;
product 9c (0.25 g, 36%). 1H NMR (CDCl
): d 8.25–8.18 (m, 2H),
.12 (d, 1H, J = 6 Hz), 7.85 (d, 1H, J = 6 Hz), 7.77 (t, 1H, J = 6 Hz),
.63 (t, 1H, J = 6 Hz), 2.87 (s, 3H). Anal. Calcd for C20 16FeN : C,
3
(
c) Chen, W.; Blaser, H.-U.. In Phosphorus Ligands in Asymmetric Catalysis;
8
7
7
Börner, A., Ed.; Wiley-VCH: Weinheim, 2008; Vol. 1, pp 359–390.
4.
5.
6.
(a) Ruble, J. C.; Latham, H. A.; Fu, G. C. J. Am. Chem. Soc. 1997, 119, 1492–1493;
H
2
(
b) Tao, B.; Lo, M. C.; Fu, G. C. J. Am. Chem. Soc. 2001, 123, 353–354.
0.61; H, 4.74; N, 8.23. Found: C, 70.70; H, 4.65; N, 8.04.
(a) Rios, R.; Liang, J.; Lo, M. C.; Fu, G. C. Chem. Commun. 2000, 377–378; (b)
Mamane, W.; Fort, Y. J. Org. Chem. 2005, 70, 8220–8223.
(a) Botteghi, C.; Chelucci, G.; Chessa, G.; Delogu, G.; Gladiali, S.; Soccolini, F. J.
Organomet. Chem. 1986, 304, 217–225; (b) Gladiali, S.; Chelucci, G.; Chessa, G.;
Delogu, G.; Soccolini, F. J. Organomet. Chem. 1987, 327, C15–C17; (c) Gladiali, S.;
Chelucci, G.; Chessa, G.; Delogu, G.; Soccolini, F. J. Organomet. Chem. 1989, 370,
0
4
.8. Synthesis of (S)-6-(2 -quinolinyl)[3,2-b]ferrocenopyridine 7
Saturated ethanolic KOH (0.8 ml) was added to a solution of
285–294.
ferrocenyl amino aldehyde 8 (0.21 g, 0.9 mmol) and 2-acetylquin-
oline 9c (0.15 g, 0.88 mmol) in absolute ethanol (17 ml) under ar-
gon. The mixture was stirred for 3 h at room temperature. Then
the reaction was quenched with water and the solvent was evapo-
rated. The resulting slurry was taken up with water and extracted
7. (a) Riesgo, E. C.; Jin, X.; Thummel, R. P. J. Org. Chem. 1996, 61, 3017–3022; (b)
Hung, C.-Y.; Wang, T.-L.; Shi, Z.; Thummel, R. P. Tetrahedron 1994, 50, 10685–
10692.
8.
Gladiali, S.; Chelucci, G.; Mudadu, M. S.; Gastaut, M.-A.; Thummel, R. P. J. Org.
Chem. 2001, 66, 400–405.
9. Reviews: (a) Arrayás, R. G.; Adrio, J.; Carretero, J. C. Angew. Chem., Int. Ed. 2006,
5, 7674–7715; (b) Atkinson, R. C. J.; Gibson, V. C.; Long, N. J. Chem. Soc. Rev.
004, 33, 313–328.
4
2
three times with CH
Na SO ) and the solvent was evaporated under reduced pressure.
The residue was purified by flash chromatography on neutral alu-
2 2
Cl . The combined organic phases were dried
(
2
4
10. Metallinos, C.; Szillat, H.; Taylor, N. J.; Snieckus, V. Adv. Synth. Catal. 2003, 345,
370–382.
1
11. Riant, O.; Samuel, O.; Flessner, T.; Taudien, S.; Kagan, H. B. J. Org. Chem. 1997,
2, 6733–6745.
mina (Et
2
O) yielding 0.18 g of the expected product 7 (54%).
H
6
NMR (CDCl
3
): d 8.69 (d, 1H, J = 9 Hz), 8.31–8.27 (m, 3H), 8.19 (d,
1
2. (a) Hanessian, S.; Ugolini, A.; Dube, D.; Glamyan, A. Can. J. Chem. 1984, 62,
2146–2147; (b) Herradon, B.; Cueto, S.; Morcuende, A.; Valverde, S.
Tetrahedron: Asymmetry 1993, 4, 845–864; (c) Boerjesson, L.; Welch, C. J.
Tetrahedron 1992, 48, 6325–6334.
1
H, J = 6 Hz), 7.87 (d, 1H, J = 6 Hz), 7.75 (t, 1H, J = 7.5 Hz), 7.58 (d,
H, J = 15 Hz), 5.49 (s, 1H), 4.98 (s, 1H), 4.30 (s, 1H), 3.92 (s, 5H).
1
1
3
C NMR (CDCl
29.7, 128.5, 127.9, 127.1, 119.6, 117.0, 108.7, 80.9, 71.8, 68.8,
3.8, 60.2. Anal. Calcd for C22 16FeN : C, 72.55; H, 4.43; N, 7.69.
3
): d 159.4, 156.7, 148.0, 141.5, 136.7, 130.1,
13. Legros, J.-Y.; Primault, G.; Fiaud, J.-C. Tetrahedron 2001, 57, 2507–2514.
1
6
14. Yadav, J. S.; Purushothama Rao, P.; Sreenu, D.; Srinivasa Rao, R.; Naveen Kumar,
V.; Nagaiah, K.; Prasad, A. R. Tetrahedron Lett. 2005, 46, 7249–7253.
H
2
1
1
5. Review: Chelucci, G.; Thummel, R. P. Chem. Rev. 2002, 102, 3129–3170.
6. (a) Lötscher, D.; Rupprecht, S.; Stoeckli-Evans, E.; von Zelewsky, A. Tetrahedron:
Asymmetry 2000, 11, 4341–4357; (b) Bolm, C.; Zehnder, M.; Bur, D. Angew.
Chem., Int. Ed. Engl. 1990, 29, 205–207; (c) Bolm, C.; Ewald, M. Tetrahedron Lett.
1990, 5011–5012; (d) Bolm, C.; Ewald, M.; Felder, M.; Schlingloff, G. Chem. Ber.
Found: C, 72.77; H, 4.31; N, 7.95.
4
.9. General procedure for allylic alkylation
1992, 125, 1169–1190.
3-C
17.
(a) Peña-Cabrera, E.; Norrby, P.-O.; Sjögren, M.; Vitagliano, A.; De Felice, V.;
Oslob, J.; Ishii, S.; O’Neill, D.; Åkemark, B.; Helquist, P. J. Am. Chem. Soc. 1996,
118, 4299–4313; (b) Chelucci, G.; Pinna, G. A.; Saba, A.; Sanna, G. J. Mol. Catal. A
2000, 159, 423–427.
A mixture of ligand (0.01–0.02 mmol) and [PdCl(
g
3 5 2
H )]
(
1.83 mg, 0.005 mmol) in THF (1 ml) was stirred at room tempera-
ture for 30 min. The resulting solution was added with a solution of
rac-1,3-diphenyl-2-propenyl acetate 13 (0.2 mmol, 50.5 mg) in
THF (1 ml), followed by dimethyl malonate (0.6 mmol, 79 mg),
18. Evans, L. A.; Fey, N.; Harvey, J. N.; Hose, D.; Lloyd-Jones, G. C.; Murray, P.;
Orpen, A. G.; Osborne, R.; Owen-Smith, G. J. J.; Purdie, M. J. Am. Chem. Soc. 2008,
1
30, 14471–14473.