326
Published on the web February 23, 2011
Development of a Practical Synthetic Method for N-tert-Butoxycarbonyl ¡-Ketimino Esters
Takuya Hashimoto, Kumiko Yamamoto, and Keiji Maruoka*
Department of Chemistry, Graduate School of Science, Kyoto University, Sakyo-ku, Kyoto 606-8502
(Received December 14, 2010; CL-101059; E-mail: maruoka@kuchem.kyoto-u.ac.jp)
TMS
O
Boc
Despite the potential synthetic utility of N-Boc ¡-ketimino
esters as prochiral ketimines to give chiral ¡-tertiary amines,
there has been no general method to access these molecules in a
practical fashion. We report herein a procedure for the one-step
synthesis of N-Boc ¡-ketimino esters starting from the
corresponding ¡-keto esters.
N
Li
TMS
N OLi
Li Li-1
Boc
N
OTMS
CO2R'
Boc
+
I
II
R
R
CO2R'
R
CO2R'
2
Boc
TMSCl
N
+
(TMS)2O + LiCl
R
CO2R'
4
3
In the realm of asymmetric catalysis aimed at the synthesis
of chiral amines, prochiral N-Boc aldimines have found
unlimited applications building on their good reactivity and
ease of deprotection after the planned transformation.1 However,
when it comes to their keto equivalent, N-Boc ketimines, there
has been essentially no report using these molecules in
asymmetric catalysis.2d We assumed that the reason for this
deficiency is partially due to the lack of supply of N-Boc
ketimines2 in addition to their elusive nature existing as a
tautomeric mixture in the case of ketimines having ¡-hydro-
gens,2c despite their high potency as valuable prochiral
substrates to produce N-Boc-protected chiral ¡-tertiary amines.3
We report herein the attempt to solve this issue by
establishment of a practical synthetic procedure for N-Boc ¡-
ketimino esters having no ¡-hydrogen which are particularly
attractive as a robust template for the asymmetric synthesis of
¡,¡-disubstituted ¡-amino acids.4 Our strategy to realize this
goal is the use of lithium N-Boc-N-TMS-amide Li-1 as a source
of the N-Boc imino group, designed according to early reports
on the synthesis of N-TMS and N-acyl imines (Figure 1).5
Nucleophilic addition of this lithium amide to ¡-keto ester 2
would give the intermediate I which might be in equilibrium
with II. We anticipated that addition of TMSCl to this
intermediate would deliver the corresponding N-Boc ¡-ketimino
ester concomitant with the extrusion of disiloxane 4.
Following this synthetic plan, we actually implemented the
synthesis of N-Boc ¡-ketimino esters. After some optimization
studies, we settled on the operationally simple one-pot sequen-
tial procedure shown in the scheme below (Table 1). Lithium
amide Li-1 could be generated by the reaction of carbamate 1
with butyllithium at ¹78 °C. Exposure of methyl benzoylfor-
mate (2a) to this solution and subsequent treatment with
chlorotrimethylsilane furnished N-Boc ¡-ketimino ester 3a in
67% yield as a single isomer (Entry 1). This reaction system
could also be applied to ¡-keto esters having bulkier esters, like
ethyl ester 2b and t-butyl ester 2c (Entries 2 and 3). A variety of
¡-ketimino esters bearing an aromatic substituent were obtained
in good yields (Entries 4-10), and even an ¡-ketimino ester
bearing an alkynyl moiety could be synthesized (Entry 11).
The limitation of this procedure is the difficulty to perform
the reaction with pyruvate (2, R = Me)2c and other alkyl-
substituted ¡-keto esters which might be due to the preferential
deprotonation of ¡-keto esters by Li-1. Probably for the same
reason, acetophenone could not be employed as well.
Figure 1. Synthetic scheme for N-Boc ¡-ketimino esters.
Table 1. Preparation of N-Boc ¡-ketimino estersa
O
R
CO2R'
TMSCl
(1.05 equiv)
Boc
BuLi (1 equiv) 2 (1 equiv)
Boc
N
TMS
N
H
THF
–78 °C, 1 h
R
–78 °C, 5 h –78 °C to rt, 2 h
CO2R'
1
3
Yield
/%b
Entry
R
R'
Product
1
2
3
Me
Et
2a
2b
2c
67
80
74
3a
3b
3c
t-Bu
Me
4
5
6
7
t-Bu
t-Bu
t-Bu
t-Bu
2d
2e
2f
71
73
78
57
3d
3e
3f
Me
MeO
2g
3g
MeO
Cl
8
9
t-Bu
t-Bu
2h
2i
54
69
3h
3i
O
10
11
t-Bu
t-Bu
2j
77
79
3j
TBS
2k
3k
b
aReactions performed at 0.50 mmol scale. Isolated yield.
After the establishment of a practical synthetic method for
N-Boc ¡-ketimino esters, we moved our attention to the
elucidation of the C=N double bond geometry. To our delight,
we could obtain a crystal of 3h suitable for X-ray crystallo-
Chem. Lett. 2011, 40, 326-327
© 2011 The Chemical Society of Japan