Synthesis of a fluorous ligand and its application for asymmetric addition of dimethylzinc to aldehydes

Article abstract of DOI:10.1021/ol070466n

A new fluorous ligand was synthesized from the acetonide of dimethyl tartarate, which showed excellent asymmetric induction on the addition of dimethylzinc to aldehydes. This ligand will be useful for synthesis of bioactive compounds with a methyl carbinol moiety. It could be recycled without using a fluorous solvent or a fluorous column.

Full text of DOI:10.1021/ol070466n

ORGANIC
LETTERS
2007
Vol. 9, No. 10
1927-1929
Synthesis of a Fluorous Ligand and Its
Application for Asymmetric Addition of
Dimethylzinc to Aldehydes
Yasser S. Sokeirik, Hiroyuki Mori, Masaaki Omote, Kazuyuki Sato, Atsushi Tarui,
Itsumaro Kumadaki, and Akira Ando*
Faculty of Pharmaceutical Sciences, Setsunan UniVersity, 45-1, Nagaotoge-cho,
Hirakata, 573-0101 Japan
aando@pharm.setsunan.ac.jp
Received February 22, 2007
ABSTRACT
A new fluorous ligand was synthesized from the acetonide of dimethyl tartarate, which showed excellent asymmetric induction on the addition
of dimethylzinc to aldehydes. This ligand will be useful for synthesis of bioactive compounds with a methyl carbinol moiety. It could be
recycled without using a fluorous solvent or a fluorous column.
Although there are a great number of biologically active
compounds containing a chiral methyl carbinol moiety,1 only
a small number of studies on the asymmetric addition of
dimethylzinc to aldehydes2 have been reported probably due
to the lower reactivity of dimethylzinc than those of its higher
homologues.3 Therefore, the development of a high enantio-
selective methodology for the addition of a methyl group to
a carbonyl group is strongly desired.
On the other hand, tartaric acid has attracted attention as
a chiral pool for synthesis of asymmetric ligands for a long
time due to its high availability of both enantiomers and its
easy derivatization to many ligands.
(1) Examples of the total synthesis of natural products having a chiral
methylcarbinol moiety: (a) Cohen, F.; Overman, L. E. J. Am. Chem. Soc.
2006, 128, 2604-2608. (b) Pattenden, G.; Critcher, D. J.; Remuinan, M.
Can. J. Chem. 2004, 82, 353-365. (c) Scott, M. S.; Luckhurst, C. A.; Dixon,
D. J. Org. Lett. 2005, 7, 5813-5816. (d) Hanessian, S. Total Synthesis of
Natural Products: The Chiron Approach; Pergamon: Oxford, 1983. (e)
Jones, G. B.; Guzel, M.; Chapman, J. Tetrahedron: Asymmetry 1998, 9,
901-905. (f) Sabitha, G.; Reddy, C.; Yadaf, J. S. Tetrahedron Lett. 2006,
47, 4513-4516.
(2) For enantioselective addition of Me2Zn, see: (a) Blay, G.; Ferna`ndez,
I.; Hernandez-Olmos, V.; Marco-Aleixandre, A.; Pedro, J. R. Tetrahe-
dron: Asymmetry 2005, 16, 1953-1958. (b) Cozzi, P. G.; Kotrusz, P. J.
Am. Chem. Soc. 2006, 128, 4940-4941. (c) Garcia-Delgado, N.; Fontes,
M.; Perica`s, M. A.; Riera, A.; Verdaguer, X. Tetrahedron: Asymmetry 2004,
15, 2085-2090. (d) Sprout, C. M.; Richmond, M. L.; Seto, C. T. J. Org.
Chem. 2004, 69, 6666-6673. (e) Cozzi, P. G.; Locatelli, M. Lett. Org.
Chem. 2004, 1, 208-211. (f) Kitamura, M.; Suga, S.; Kawai, K.; Noyori,
R. J. Am. Chem. Soc. 1986, 108, 6071-6072. (g) A unique example for
Me2Zn addition to ketones: Garc´ıa, C.; LaRochelle, L. K.; Walsh, P. J. J.
Am. Chem. Soc. 2002, 124, 10970-10971. (h) Addition of Me2Zn to
ketoesters: Wieland, L. C.; Deng, H.; Snapper, M. L.; Hoveyda, A. H. J.
Am. Chem. Soc. 2005, 127, 15453-15456.
TADDOL is one of the excellent chiral ligands derived
from tartaric acid.4
In this paper, we would like to describe the synthesis of
a new fluorous ligand 1a (see Figure 1) derived from tartaric
acid, which shows a high efficiency for asymmetric addition
of dimethylzinc to aldehydes.
On the other hand, we have reported the synthesis of
perfluoroalkylated chiral ligands.5 We had expected that large
(3) Kitamura, M.; Okada, S.; Suga, S.; Noyori, R. J. Am. Chem. Soc.
1989, 111, 4028-4036.
(4) For the use of TADDOL as a chiral ligand, please see: Seebach, D.;
Beck, A. K.; Heckel, A. Angew. Chem., Int. Ed. 2001, 40, 92-138.
10.1021/ol070466n CCC: $37.00
© 2007 American Chemical Society
Published on Web 04/11/2007

Scheme 1
Figure 1. TADDOL and our ligand 1a.
perfluoroalkyl groups would be sterically large enough to
induce high ee and that their electronic effect would increase
the Lewis acidity of the metal coordinated and hence increase
the activity of the complex. These expectations were
confirmed by the synthesis of a few perfluoroalkylated
ligands.5 As the extension of these works, we designed a
new chiral ligand (1a).
The ligand 1a has two different types of OH groups and
hence different coordination strengths that may help the
formation of a more effective metal complex with a fixed
electronic and steric environment.2a Perfluoroalkyl groups
would make the neighboring hydroxy group less sensitive
to elimination and/or oxidation than common alkyl groups
due to their high electronegativity5 and make 1a very stable.
Further, the high content of fluorine would make the ligand
recoverable by fluorous technology.6
The ligand 1a was synthesized according to Scheme 1.
Treatment of dimethyl (4R,5R)-2,2-dimethyl-1,3-diox-
olane-4,5-dicarboxylate (2) with a perfluorooctyl Grignard
reagent gave a keto alcohol (3). This reaction was highly
sensitive to temperature. At low temperature, the main
product was a keto alcohol (4), which must be formed by
reduction of a dicarbonyl intermediate with methoxymag-
nesium halide via the same mechanism as that on the
reduction of perfluoroalkyl ketone with lithium alkoxide.7
On the other hand, at a higher temperature, the Grignard
reagent was decomposed. A part of these difficulties was
due to the low solubility of highly fluorinated compounds
in common solvents. After several trials, we found that the
optimum condition is to use a high dilution condition and a
temperature range between -65 and -30 °C. The Grignard
reagent was formed at -65 °C. Then the mixture was
warmed rapidly to -40 °C, and the diester 2 was added.
The reaction mixture was stirred below -30 °C for 4 h.
The reduction of the keto alcohol 3 using LiAlH4 or other
common reducing agents afforded a mixture of two possible
isomers, 1a and 1b, with low selectivity. We had the best
results using a mixture of excess NaBH4/CeCl3. This reagent
gave a good selectivity (1a/1b ) 10:1) and high conversion.
The isomer 1a was obtained as crystals by crystallization of
the mixture from chloroform, whereas 1b was obtained as
an oil contaminated with 30% (estimated by GLC) of 1a
after concentration of the filtrate. We determined the structure
of 1a by single-crystal X-ray analysis. Thus, we could obtain
1a effectively.
We estimated the activity of asymmetric induction of the
isomers on the addition of diethylzinc to benzaldehyde.
Ligand 1a was found to be highly effective in the level of 3
mol % giving (R)-1-phenylpropanol in a quantitative yield
with 98% ee, whereas 1b gave the (S) isomer in only 40%
yield with 20% ee. These results showed that we obtained
the active isomer predominantly. This result encouraged us
to try the addition of dimethylzinc in the presence of 1a
(Scheme 2).
Scheme 2
(5) (a) Omote, M.; Kominato, A.; Sugawara, M.; Sato, K.; Ando, A.;
Kumadaki, I. Tetrahedron Lett. 1999, 40, 5583-5585. (b) Omote, M.;
Nishimura, Y.; Sato, K.; Ando, A.; Kumadaki, I. Tetrahedron Lett. 2005,
46 (2), 319-322. (c) Omote, M.; Nishimura, Y.; Sato, K.; Ando, A.;
Kumadaki, I. Tetrahedron 2006, 62 (8), 1886-1894. (d) Sokeirik, Y. S.;
Omote, M.; Sato, K.; Kumadaki, I.; Ando, A. Tetrahedron: Asymmetry
2006, 17, 2654-2658.
(6) Omote, M.; Nishimura, Y.; Sato, K.; Ando, A.; Kumadaki, I.
Tetrahedron 2006, 62, 1886-1894, and references therein.
(7) Sokeirik, Y. S.; Omote, M.; Sato, K.; Kumadaki, I.; Ando, A. J.
Fluorine Chem. 2006, 127, 150-152. Concerning the reduction of per-
fluoroalkyl ketones with alkoxymagnesium: Yamazaki, T.; Terajima, T.
unpublished data.
A portion of 3 mol % of 1a gave the product in a
quantitative yield with 80% ee. Increasing the amount of 1a
to 6 mol % afforded 96% ee. Solvents other than hexane
were shown to be noneffective. The reaction was completed
within only 3 h, which is unusually short for the addition of
dimethylzinc to aldehydes.
These results encouraged us to apply this reaction to other
aromatic aldehydes, including the ortho-substituted ones,
1928
Org. Lett., Vol. 9, No. 10, 2007

aliphatic aldehydes, and heterocyclic ones, most of which
were reported to give low ee’s even in the reaction of
diethylzinc.8 The results are shown in Table 1.
aldehydes. Cyclohexanecarbaldehyde (Table 1, entry 9) gave
lower yield and ee. R,â-Unsaturated aldehydes gave excellent
results especially in the case of 2-octynal (98% ee).
Next, we examined the recycling of the catalyst. We found
an economical way, taking advantage of the very low
solubility of the ligand in cold toluene. Our procedure was
as follows: the crude mixture obtained by concentrating the
extract from the reaction of dimethylzinc and benzaldehyde
was treated with cold toluene, in which the ligand is
practically insoluble. By simple filtration, we could separate
the ligand from the products. The recovered ligand was used
without purification. By this simple method, we could recycle
the ligand several times. The recycling data are shown in
Table 2.
Table 1. Methylation of Various Aldehydes in the Presence of
1a
entry
RCHO (R)
C6H5
p-CH3OC6H4-
p-CF3C6H4-
2-naphthyl
o-FC6H4-
time (h)
yield (%)a
ee (%)b
1
2
3
4
5
6
7
8
9
10
11
12
13
3
3
4
2
3
3
2
3
4
3
3
1.5
1
99
95
96
99
97
97
98
92
78
93
95
98
99
96
92
>99
95
89
o-ClC6H4-
91
n-C7H15
-
91c
88
PhCH2CH2-
cyclohexyl
PhCHdCH-
n-C5H11CtC-
2-furyl
84c
92
Table 2. Recycling of the Ligand
product
ligand
recovered(%)
98c
>99
84
cycle
yield (%)a
ee (%)b
benzofuran-2-yl
1
2
3
4
5
6
100
100
100
96
99
100
95
96
93
91
85
83
a Isolated yields. bDetermined by an HPLC OD-H column. Absolute
configurations are established by signs of the optical rotation reported. All
are R configuration. cDetermined by chiral HPLC analysis of the dinitroben-
zoate.
100
93
90
89
80
b
a Determined by GLC. Determined by HPLC.
The ligand gave excellent yields in every entry except
cyclohexanecarbaldehyde. The enantioselectivities were very
good to excellent. Aromatic aldehydes gave excellent yields
and ee’s, especially in the case of an electron-withdrawing
group which gave up to 99% ee (Table 1 entry 3), whereas
an aldehyde with an electron-donating group gave a little
lower ee of 92% (entry 2). Bulky aromatic aldehydes such
as 2-naphthaldehyde also gave excellent ee up to 95%. With
ortho-substituted aldehydes (entries 5 and 6), which were
reported to give low yields and ee’s even in the addition of
diethylzinc using TADDOL, we got an excellent chemical
yield up to 97% with very good to excellent ee. Furfural
gave an excellent result (entry 12), and the more bulky
derivative showed a lower ee of 84% (entry 13).
We observed that the yield was not affected through the
whole recycling process, whereas the ee decreased slightly.
The recycling of the ligand was not accompanied by
epimerization. The reason for the slight decrease of enan-
tiomeric excess is under investigation.
In conclusion, we have synthesized a new ligand based
on fluorous technology which showed an excellent asym-
metric induction on the addition of dimethylzinc to alde-
hydes. The high yield and ee and the short reaction time
will make it one of the best ligands ever made. This ligand
can be recycled without using expensive fluorous solvents
or fluorous columns.
Next, we tried aliphatic aldehydes including linear, cyclic,
and R,â-unsaturated aldehydes (entries 7-11). Saturated
linear aldehydes gave excellent results. Thus, the ee was very
good with hydrocinnamaldehyde (88%) and excellent with
octanal (91%) but a little lower than those of the aromatic
Supporting Information Available: General procedures
of experimental and spectral data including NMR, HPLC,
and X-ray analysis and NMR charts. This material is
available free of charge via the Internet at http://pubs.acs.org.
(8) Kitajima, H.; Ito, K.; Katsuki, T. Chem. Lett. 1996, 343-344.
OL070466N
Org. Lett., Vol. 9, No. 10, 2007
1929