A chiral iron(II)-pybox catalyst stable in aqueous media. Asymmetric Mukaiyama-aldol reaction

Article abstract of DOI:10.1016/j.tetlet.2006.05.140

Among various transition metals, iron is nontoxic, cheap and hence of great potential synthetic use. We report herein a water stable chiral Lewis acid containing an iron(II) ion and a pybox-type ligand. The resulting cationic aqua complex of C₂

Full text of DOI:10.1016/j.tetlet.2006.05.140

Tetrahedron Letters 47 (2006) 5281–5284
A chiral iron(II)–pybox catalyst stable in aqueous media.
Asymmetric Mukaiyama–aldol reaction
Joanna Jankowska, Joanna Paradowska and Jacek Mlynarski*
Institute of Organic Chemistry, Polish Academy of Sciences, Kasprzaka 44/52, 01-224 Warsaw, Poland
Received 20 February 2006; revised 13 May 2006; accepted 24 May 2006
Abstract—Among various transition metals, iron is nontoxic, cheap and hence of great potential synthetic use. We report herein a
water stable chiral Lewis acid containing an iron(II) ion and a pybox-type ligand. The resulting cationic aqua complex of C -sym-
2
metry is an effective Lewis acid catalyst for asymmetric Mukaiyama–aldol reactions in aqueous media. The aldol products have been
obtained in good yields, syn-diastereoselectivities and ca. 70% levels of enantioselectivity.
Ó 2006 Elsevier Ltd. All rights reserved.
9
The aldol reaction is considered as one of the most
important carbon–carbon bond forming reactions in or-
ganic synthesis. Its usefulness for the construction of
ongoing development. In this respect enantioselective
transformations promoted by iron complexes constitute
1
10
precious and rare examples.
natural products has promoted rapid evolution of effi-
cient chiral catalysts. Among other methodologies,
the chiral Lewis-acid-catalyzed condensation of silyl
enol ethers with aldehydes (the Mukaiyama reaction)3
constitutes one of the most convenient and relevant vari-
ants of the classical aldol reaction.
2
11
Previously, dicarbonyl cyclopentadienyl iron halides
1
2
and cationic iron complexes were demonstrated as cat-
alysts in diastereoselective Mukaiyama-type additions of
ketene acetals to aldehydes in anhydrous solvents. More
recently, Kobayashi reported that Fe(II) and Fe(III)
salts showed considerable activity in stereoselective
Mukaiyama–aldol reactions in aqueous THF,1 espe-
cially when the reaction was carried out in the presence
of catalytic amounts of surfactants. Thus, we consid-
ered iron salts to be interesting candidates for designing
water compatible chiral Lewis acids. To the best of our
knowledge, chiral iron-complexes have never been dem-
onstrated as catalysts for asymmetric Mukaiyama–aldol
reactions, not to mention in aqueous media, where the
said catalytic system must meet much more stringent
3
In recent years, stereoselective organic reactions in aque-
4
,5
ous media have attracted a great deal of attention. In
this respect, the enantioselective Mukaiyama reaction in
aqueous solvents presents an important and challenging
1
4
6
issue. Since Kobayashi showed that a chiral copper
complex can act as a water tolerant catalyst for the
7
Mukaiyama–aldol reaction, several other Lewis acids
have been utilized in asymmetric aldol reactions in aque-
8
ous solvents.
5
a,6b
requirements.
However, most of those catalysts are composed of heavy
or rare earth metals which create drawbacks in their
applications because of the toxicity or high price. In
contrast, iron is the most abundant metal on earth,
and consequently one of the cheapest and environmen-
tally friendly. Interest in well-defined iron complexes
as catalysts for bond forming reactions is an area of
As a part of an ongoing program towards new asymmet-
ric aldol methodologies, we reported recently the appli-
cation of a zinc-based chiral Lewis acid in an asymmetric
Mukaiyama–aldol reaction in aqueous media. Here we
present the first example of water compatible iron(II)
complexes as non-toxic, environmentally benign cata-
lysts for asymmetric Mukaiyama carbon–carbon bond
forming processes.
1
5
Keywords: Asymmetric synthesis; Mukaiyama reaction; Water; Iron
complex.
*
Corresponding author. Tel.: +48 22 343 21 30; fax: +48 22 632 66
1; e-mail: mlynar@icho.edu.pl
We initially screened various chiral ligand-iron(II) chlo-
ride complexes for the best catalytic activity in the
8
0040-4039/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.05.140

5282
J. Jankowska et al. / Tetrahedron Letters 47 (2006) 5281–5284
aqueous asymmetric Mukaiyama reaction. Our preli-
was found to be an unpromising combination. Iron
fluoride did not form a catalyst soluble in the reaction
mixture.
1
7
minary studies revealed that pybox-type ligands 1
1
8
and 2 were the most promising candidates for further
studies.
The effect of the solvents was then examined. The aldol
reaction proceeded most efficiently in water–ethanol
solutions. The Fe(II)-based chiral Lewis acid was stable
in alcohol-type solvents but the catalytic system decom-
posed rapidly when an aqueous mixture with aprotic
solvents, such as THF and DME, was engaged as the
reaction medium.
The iron(II)–pybox complexes were prepared by stirring
a deoxygenated EtOH/H O mixture of ligand and
2
iron(II) chloride until all the solid was observed to
dissolve. To the resulting dark red solution, benzalde-
hyde and silyl enol ether 3 were added under an argon
atmosphere. The reaction in EtOH/H O (9/1) proceeded
2
smoothly at 0 °C affording excellent selectivity (Table 1,
entry 1).
We have found that a water content exceeding 20% in the
reaction medium resulted in a lower yield and selectivity.
On the other hand, the essential role of water was proved
by using dry ethanol. In this case, a lower yield and ste-
reoselectivity were observed (entry 12). In dry dichloro-
methane, only traces of aldol product was formed in
the reaction.
The reaction outcome depended, however, on the sol-
vent ratio giving the best enantioselectivity for EtOH/
H O (4/1) (entry 2). We were delighted to find that the
2
reaction proceeded in the same yield and good enantio-
selectivity using 10 mol % of the catalyst (entry 3).
Below this level a lower yield and selectivity were
observed (entry 4).
In all cases careful deoxygenation of the solvents played
an essential role in obtaining satisfactory levels of ee and
was crucial for reproducibility of the experiments.
Although iron(III) chloride with 1 was an active Lewis
acid, it delivered aldol 5a as a racemate (entry 14). For-
mation of an iron(III) catalyst was thus expected to be
The use of Fe(BF ) Æ6H O (entry 6) or Fe(ClO ) (entry
4
2
2
4 2
7
) with 1 resulted in a slightly better reaction yields
without significant change in the diastereoselectivity,
but a drop in enantioselectivity was observed. FeSO -1
4
Table 1. Solvent and catalyst studies. Reaction of benzaldehyde with Z-enol silyl ether 3
O
O
O
O
N
N
N
N
N
N
1
2
OH
HO
OH
OTMS
O
Fe-salt (n mol%)
or 2 (n mol%)
CHO
1
+
3
4
solvent
(
1.2 equiv.)
0 °C / 5 h
5a
a
b
Entry
Solvent
Fe-salt
n
Yield (%) (syn/anti)
ee (%) (syn)
Ligand 1
c
1
2
3
4
5
6
7
8
9
EtOH/H
EtOH/H
EtOH/H
EtOH/H
EtOH/H
EtOH/H
EtOH/H
MeOH/H
n-PrOH/H
i-PrOH/H
2
2
2
2
2
2
2
O 9/1
O 4/1
O 4/1
O 4/1
O 7/3
O 4/1
O 4/1
FeCl
FeCl
FeCl
FeCl
FeCl
2
2
2
2
2
Æ4H
Æ4H
Æ4H
Æ4H
Æ4H
2
O
O
O
O
O
20
20
10
5
66 (94/6)
62 (96/4)
60 (95/5)
50 (92/8)
42 (94/6)
65 (95/5)
69 (94/6)
55 (95/5)
50 (92/8)
45 (92/8)
70 (85/15)
30 (86/14)
30 (81/19)
39 (86/14)
40 (S,S)
2
2
2
2
62
62
24
30
45
38
33
28
28
rac.
47
40
rac.
10
10
10
10
10
10
10
10
10
10
Fe(BF
Fe(ClO
4
)
2
Æ6H
2
O
4
)
2
ÆnH
2
O
2
O 4/1
O 4/1
O 4/1
O 4/1
FeCl
FeCl
FeCl
FeCl
FeCl
FeCl
FeCl
2
2
2
2
2
2
3
Æ4H
Æ4H
Æ4H
Æ4H
Æ4H
2
O
O
O
O
O
2
2
2
2
2
10
11
12
13
14
2
DMF/H
EtOH
2
EtOH
EtOH/H
2
O 4/1
Ligand 2
c
15
6
EtOH/H
EtOH/H
2
O 9/1
O 4/1
FeCl
FeCl
2
Æ4H
Æ4H
2
O
O
10
10
68 (91/9)
66 (85/15)
70 (R,R)
58
1
2
2
2
a
b
c
Isolated yield after silica gel chromatography.
Determined by HPLC analysis using a chiralpak AD-H column.
The absolute configuration of the aldol product 5a was determined by comparing the optical rotation and HPLC analyses with the literature
16,8c
data.

J. Jankowska et al. / Tetrahedron Letters 47 (2006) 5281–5284
5283
Table 2. Asymmetric aldol reaction of silyl enol ether 3 with various
In summary, we have shown that the FeCl –pybox com-
2
19
aldehydes catalyzed by FeCl
2
/hm-pybox 2 (10 mol %)
plex is an efficient chiral catalyst for asymmetric aldol
reactions in aqueous media. To the best of our knowl-
edge, this is the first example of a chiral iron complex
being active in aqueous asymmetric Mukaiyama reac-
tions. Although the stereoselectivity remains to be im-
proved further, the present work provides a useful
concept for the design of chiral catalysts composed of
iron(II) salts which function effectively in aqueous
a
b
Entry Aldehyde
Product Yield (%) ee (%) (syn)
(
syn/anti)
CHO
20
1
5a
72 (91/9)
75 (92/8)
70 (R,R)
CHO
2
5b
70
2
1
Me
media.
CHO
3
5c
5d
65 (93/7)
87 (93/7)
75
70
References and notes
MeO
1
. Modern Aldol Reactions; Mahrwald, R., Ed.; Wiley-VCH:
Weinheim, Germany, 2004.
CHO
4
2. (a) Alcaide, B.; Almendros, P. Eur. J. Org. Chem. 2002,
1595; (b) Palomo, C.; Oiarbide, M.; Garcia, J. M. Chem.
Soc. Rev. 2004, 33, 65; (c) Machajewski, T. D.; Wong,
C.-H. Angew. Chem., Int. Ed. 2000, 39, 1352.
. (a) Mukaiyama, T.; Narasaka, K.; Banno, K. Chem. Lett.
973, 1011; (b) Mukaiyama, T. Org. React. 1982, 28, 203;
c) Mukaiyama, T.; Matsuo, J. In Modern Aldol Reactions;
Mahrwald, R., Ed.; Wiley-VCH: Weinheim, Germany,
004; p 127.
. Organic Synthesis in Water; Grieco, P. A., Ed.; Blackie
Academic & Professional: London, 1998.
. (a) Lindstr o¨ m, U. M. Chem. Rev. 2002, 102, 2751; (b)
Sinou, D. Adv. Synth. Catal. 2002, 344, 221; (c) Li, Ch.-J.
Chem. Rev. 2005, 105, 3095.
OMe
CHO
3
5
5e
5f
79 (9/1)
72 (7/3)
72
44
23
1
(
Cl
CHO
2
6
4
5
CHO
c
7
5g
25 (8/2)
6
. (a) Manabe, K.; Kobayashi, S. Chem. Eur. J. 2002, 8,
4095; (b) Kobayashi, S.; Manabe, K. Acc. Chem. Res.
2002, 35, 209.
a
Isolated yield after silica gel chromatography.
Determined by HPLC analysis using a chiralpak AD-H column.
Reaction time 72 h.
b
c
7. Kobayashi, S.; Nagayama, S.; Busujima, T. Tetrahedron
999, 55, 8739.
1
8
. For a lead-based catalyst see: (a) Nagayama, S.; Koba-
yashi, S. J. Am. Chem. Soc. 2000, 122, 11531; for a
gallium-based catalyst see: (b) Li, H.-J.; Tian, H.-Y.; Wu,
Y.-Ch.; Chen, Y.-J.; Liu, L.; Wang, D.; Li, Ch.-J. Adv.
Synth. Catal. 2005, 347, 1247; for rare earth metal
catalysts see: (c) Hamada, T.; Manabe, K.; Ishikawa, S.;
Nagayama, S.; Shiro, M.; Kobayashi, S. J. Am. Chem.
Soc. 2003, 125, 2989; (d) Ishikawa, S.; Hamada, T.;
Manabe, K.; Kobayashi, S. J. Am. Chem. Soc. 2004, 126,
12236; for a recent example of a bismuth-based catalyst
see: (e) Kobayashi, S.; Ogino, T.; Shimizu, H.; Ishikawa,
S.; Hamada, T.; Manabe, K. Org. Lett. 2005, 7, 4729.
. Bolm, C.; Legros, J.; Le Paih, J.; Zani, L. Chem. Rev.
responsible for the decrease in the reaction selectivity in
nondeoxygenated solvents.
Interestingly, other pybox-type ligands delivered essen-
tially different results. Application of (R,R)-diphenyl-py-
box resulted in the formation of a racemic aldol product.
On the other hand, the yield and selectivity were im-
proved by application of 10 mol % of (R,R)-hm-pybox
2
and Fe(II) chloride (entry 15). This catalytic system
was more stable and thus a far more attractive combina-
9
tion than iron(II)–pybox 1.
2
004, 104, 6217.
0. For examples of asymmetric Diels–Alder reactions, see:
a) K u¨ ndig, E. P.; Saudan, C. M.; Viton, F. Adv. Synth.
1
Finally, it was revealed that the aldol product 5a was
obtained in 72% yield with 70% ee using 10 mol % of
FeCl Æ4H O and 2 (Table 2, entry 1). It is noteworthy
(
Catal. 2001, 343, 51, and references cited therein; (b)
Kanemasa, S.; Oderaotoshi, Y.; Sakaguchi, S.-i.; Yama-
moto, H.; Tanaka, J.; Wada, E.; Curran, D. P. J. Am.
Chem. Soc. 1998, 120, 3074; (c) Corey, E. J.; Ishihara, K.
Tetrahedron Lett. 1992, 33, 6807; (d) Khiar, N.; Fernan-
dez, I.; Alcudia, F. Tetrahedron Lett. 1993, 34, 123; (e)
Matsukawa, S.; Sugama, H.; Imamoto, T. Tetrahedron
Lett. 2000, 41, 6461; for enantioselective sulfide oxidation
see: (f) Legros, J.; Bolm, C. Angew. Chem., Int. Ed. 2004,
2
2
that the same level of selectivity was attained using
FeCl Æ4H O and anhydrous FeCl . For practical rea-
2
2
2
sons, we chose for further experiments FeCl which is
easier to handle and is less prone to air oxygenation.
2
1
9
Several other substrates were subjected to this catalytic
system, and the results are summarized in Table 2. In
the case of aromatic aldehydes all the reactions provided
good yields (65–87%), diastereoselectivities (syn/anti =
4
3, 4225; for other examples see: (g) Nakamura, M.; Hirai,
A.; Nakamura, E. J. Am. Chem. Soc. 2000, 122, 978; (h)
Trost, B. M.; Lee, Ch. B. J. Am. Chem. Soc. 2001, 123,
9
1/9–93/7) and enantioselectivities for the syn isomer
3
2
671; (i) Christoffers, J.; Mann, A. Angew. Chem., Int. Ed.
000, 39, 2752; (j) Scheafer, C.; Fu, G. C. Angew. Chem.,
of 70–75%. a,b-Unsaturated cinnamaldehyde gave a rel-
atively lower ee of aldol product 5f (44%). An aliphatic
aldehyde (entry 7) was less compatible with the reaction
conditions (25% yield, 23% ee).
Int. Ed. 2005, 44, 4606.
11. Colombo, L.; Ulgheri, F.; Prati, L. Tetrahedron Lett. 1989,
30, 6435.

5284
J. Jankowska et al. / Tetrahedron Letters 47 (2006) 5281–5284
1
1
1
1
2. Bach, T.; Fox, D. N. A.; Retz, M. T. J. Chem. Soc. Chem.
2 equiv) and appropriate aldehyde (0.5 mmol) were added
and the resulting solution was stirred at 0 °C for 5 h under
an argon atmosphere. The reaction was diluted with
methyl tert-butyl ether (30 mL) and washed with water
and brine. The organic phase was dried, evaporated to
dryness and the residue was purified by silica gel chroma-
tography (typically, AcOEt/hexane, 1/4).
Commun. 1992, 1634.
3. Kobayashi, S.; Nagayama, S.; Busujima, T. J. Am. Chem.
Soc. 1998, 120, 8287.
4. Aoyama, N.; Manabe, K.; Kobayashi, S. Chem. Lett.
2004, 33, 312.
5. (a) Mlynarski, J.; Jankowska, J. Adv. Synth. Catal. 2005,
3
2
47, 521; (b) Jankowska, J.; Mlynarski, J. J. Org. Chem.
006, 71, 1317.
20. Data for (2R,3R)-3-hydroxy-2-methyl-1,3-diphenyl-pro-
1
pan-1-one (5a): H NMR (200 MHz) d (syn): 1.19 (d, 3H,
1
1
1
1
6. Denmark, S.; Wong, K.-T.; Stavenger, R. A. J. Am. Chem.
Soc. 1997, 119, 2333.
7. Desimoni, G.; Faita, G.; Quadrelli, P. Chem. Rev. 2003,
J 7.2 Hz), 3.67 (d, 1H, OH, J 1.7 Hz), 4.11 (dq, 1H, J 3.0,
7.3 Hz), 5.23 (br s, 1H), 7.20–7.65 (m, 8H), 7.85–8.20 (m,
1
3
2H); C NMR (50 MHz) d: 11.1, 47.1, 73.0, 125.9, 127.3,
128.2, 128.4, 128.7, 133.5, 135.6, 141.7, 205.8.; [a]
À3.5 (c
1.00 in CHCl , ee 70%), lit. for syn-(2S,3S): [a]D +11.7 (c
1
03, 3119.
8. Iwasa, S.; Nakamura, H.; Nishiyama, H. Heterocycles
000, 52, 939.
9. Representative procedure: a mixture of hm-pybox ligand 2
17 mg, 0.06 mmol) and iron(II) chloride (6.5 mg,
D
1
6
3
2
1.03 in CHCl
hexane/i-PrOH = 9/1, flow rate = 1 mL/min; syn isomer:
= 9.9 min (major), t = 12.2 min (minor), anti isomer:
t = 14.6 min, t = 16.0 min.
3
, 95% ee); HPLC Daicel Chiralpak AD-H,
(
t
1
2
0
.05 mmol, 10 mol %) in deoxygenated EtOH/H O (9/1,
2
1
2
1
.5 mL) was stirred at 0 °C under an Ar atmosphere until
21. During the evaluation of this manuscript we encountered a
more efficient catalyst composed of an iron salt and a
tuned pybox ligand. These results will be published soon.
all the solid had dissolved (15–20 min). To the resulting
deep-red solution, silyl enol ether 3 (230 lL, 1.0 mmol,