COMMUNICATION
entry 1). In the reaction, the choice of the base appears to
be crucial for the reactivity and the selectivity, and sodium
N-aryltosylamide gave promising results (entries 2–4).
Among them, the amides bearing naphthalene ring systems
performed best, and sodium N-5,6,7,8-tetrahydro-1-naph-
thyltosylamide showed the high diastereomeric ratio of anti/
syn=90:10. We then investigated the effect of the metal
moiety in the amide catalyst and found that magnesium salts
showed the best stereoselectivity (entry 6). It was confirmed
that Et2O also worked as an acceptable solvent, but the
yield was a little lower (entry 7). Further, reduction of the
catalyst amount was conducted, and just 2 mol% Mg cata-
lyst gave the desired product in high yield without any loss
of diastereoselectivity (entry 9). The effect of the fluorenyli-
dene group is very important, and the typical glycine Schiff
base 1b derived from benzophenone did not work at all
under the reaction conditions (entry 10). It should be noted
that the use of fluorenylidene group enhanced the reactivity
of the glycine ester dramatically in the aldol reaction.
Scheme 2. Modification of the product.
dene group and hydrolysis of the ester (Scheme 2). Acid hy-
drolysis of the aldol adduct 3a gave the corresponding b-hy-
droxy-a-amino ester HCl salt 5 in 94% yield, and following
basic hydrolysis, the desired b-hydroxy-a-amino acid 6[17]
was obtained. This sequence could also be performed with-
out any purification of the intermediates.
With optimum reaction conditions in hand, we next inves-
tigated the scope of aldehydes (Table 2). Typical aliphatic al-
During the investigations of this aldol reaction, we could
easily observe the retro-aldol process. When we carried out
the reaction at room temperature, the desired aldol product
3a was obtained in moderate yield with very low diastereo-
selectivity (Scheme 3). When we put the product 3a with
Table 2. Aldehyde scope.[a]
Entry
R
anti/syn
Yield [%]
1
2
n-C5H11 (2b)
iPr (2c)
89/11
94/6
92/8
95/5
95/5
94/6
55/45
54/46
–
91
88
90
85
87
74
35
79
NR
53
3
ACHTUNGTRENNUNG(CH3)2CHCH2 (2d)
4
PhCH2 (2e)
5
6
PhCH2CH2 (2 f)
Ph2CH (2g)
7
Ph (2h)
8
9
10
o-O2NC6H4 (2i)
(E)-PhCH=CH (2j)
(E)-CH3CH=CH (2k)
52/48
[a] The reaction was performed using 1a and 2 (1.1 equiv) in THF at
À788C for 15 h in the presence of 4 f (2 mol%). [b] Determined by
1H NMR analysis of the crude mixture. [c] Yield of isolated product.
dehydes, n-hexanylaldehyde (2b), isobutylaldehyde (2c),
and isopentylaldehyde (2d), worked well, similarly to 2a,
and high yields and high diastereoselectivities were obtained
(entries 1–3). The reaction was not sensitive to substitutions
at the a-position of the aldehyde. Aliphatic aldehydes bear-
ing an aromatic ring were also found to be effective, and
good to high yields and high selectivities were observed (en-
tries 4–6). Interestingly, easily enolizable aldehydes, phenyl-
acetoaldehyde (2e) and diphenylacetoaldehyde (2g), gave
the desired products in good yields without significant side
reactions. However, use of aromatic and a,b-unsaturated al-
dehydes was not fruitful, and the desired aldol products
were obtained in low to good yields with quite disappointing
selectivities (entries 7–10). That is presumably because
a retro-aldol process from the aldol product in the reaction
system seemed to be significant when conjugated aldehydes
were employed.
Scheme 3. Consideration of the reaction process.
low diastereomeric ratio under the reaction conditions at
À788C, no change in the diastereomeric ratio was observed.
On the other hand, when we put 3a with high diastereo-
meric ratio under the reaction conditions at room tempera-
ture, the ratio dramatically decreased, with some loss of the
product. These results indicate that the aldol reaction pro-
ceeded at À788C under kinetic control to afford the desired
product in high diastereoselectivity, and the diastereomeric
ratio of the product decreased at high reaction temperature
to lead to a low diastereomeric ratio under thermodynamic
control via the retro-aldol process.
The products could be easily converted to the correspond-
ing b-hydroxy-a-amino acid by removal of the fluorenyli-
Finally, the structure of 4 f was investigated by NMR stud-
ies. In THF-d8, the tosyl amide and 4 f showed very different
&
&
&2
&
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 0000, 00, 0 – 0
ÝÝ These are not the final page numbers!