A. Msutu, R. Hunter / Tetrahedron Letters 55 (2014) 2295–2298
2297
1) DBAD (2 equiv) / MCA (5.2 equiv) /
TFA (0.1 equiv) / tetrazole (20 mol%) /
CH2Cl2 (0.5 M) / H2O (5 equiv) / 4 °C
the superior H-bonding acceptor attributes of the aza-amidine
functionality.
OH
RO OR
Regarding scope (Fig. 3), changing the acetal from dimethoxy to
diethoxy (entries 1 and 2) increased the reaction rate, presumably
due to an increase in the stability of the intermediate oxocarbeni-
um ion as well as the greater basicity of the ethoxy group, making
it a better leaving group.23 Importantly, the yield rose to 93%
reflecting the instability of DBAD in an acid medium (shorter reac-
tion time). As with the model (entry 1), ee values were both very
good (>85%). Entry 4 reveals that quaternized stereocentres can
be accessed, albeit with a drop in ee to 56%, which nevertheless
compared reasonably well with other reports on the corresponding
aldehyde8a,e using proline-based catalysts. For ketals (entries 5–7),
as expected, rates depended on the OR groups in the order
cyclic > acyclic for the same OR group and dimethoxy > dioxolane,
which is in agreement with known hydrolysis rates23 and the sta-
bility of the intermediate oxocarbenium ion based on inductive
and hyperconjugative effects. Entry 5 gave a 96:4 ratio, by HPLC,
of more-substituted:less-substituted regioisomers in accordance
with the intermediacy of the more stable enamine. As indicated
in Figure 3, in all ketal cases, excess substrate was used (to 1 equiv
of DBAD) to drive the reaction to completion as well as to avoid
BnO
N
2) NaBH4 / EtOH
NH
OBn
O
(1 equiv)
O
OMe
(CH2)7CH3
OEt
OEt
entries 1-3
MeO
EtO
EtO
3
2
1
9 h; 73%; 94% ee
4 h; 93%; 86% ee
9 h; 79%; 90% ee
MeO OMe
OMe
MeO
MeO OMe
entries 4-6
4
5
6
Ph
20 equiv
5 equiv
72 h; 51%; 56% eea,b 48 h; 33%; 86% eea, c 9 h; 71%; 62% eea, c
O
O
a
b
c
Product isolated and evaluated without reduction
4 equiv of DBAD + 10.4 equiv of MCA used
DBAD (1 equiv)
entry 7
7
20 equiv
48 h; 66%; 64% eea, c
disubstitution. In these cases the a-aminated ketone was isolated
Figure 3. Scope of the acetal a-amination.
without carbonyl reduction to avoid the introduction of an extra
stereocentre into the product.
In view of an obvious application to lactol
decided to study the enantioselective -amination of d-lactol 8
containing a hemi-acetal masked aldehyde functionality, Table 2.
This is the first example of this type of amination, with obvious
application as a model for C-2-deoxy carbohydrate derivatives.
As expected, in view of its ability to open directly to an alde-
ee. The proposed rationale is that, in this case, the carboxyl group
(taken as representative) of the enamine intermediate is heavily H-
bonded to a hydronium ion/water conglomerate, whose hydro-
nium ion content is dependent on the strength and concentration
of the Brønsted acid. Such H-bonding is a sterically more severe
environment than that in List and Jorgensen’s case pertaining to
the reaction of aldehydes,2,3 which involves the production of only
one mol of water per mol of substrate from the condensation reac-
tion to generate the enamine, and in which there is no acid. Indeed,
in their model it is assumed that this water plays no important role
in the assisted transition state involving direct H-bonding between
the proline carboxyl group and the diazo nitrogen. In the acetal
reaction at high acid concentrations, assistance might be promoted
in two possible ways. Either protonation of the DBAD occurs, ren-
dering it an excellent H-bond donor to the carboxyl group carbonyl
oxygen, or a high hydronium ion concentration around the car-
boxyl group promotes general acid catalysis towards the electro-
phile. In either cases, direct H-bonding between the carboxyl OH
and the DBAD, as in the Houk–List model, is not considered to be
operating, (Fig. 2). AcOH, as a weaker acid, promotes a level of ste-
ric model due to the steric bulk of the water conglomerate around
the proline carboxyl group, which leads to an increase in the
percentage of (S)-enantiomer. Importantly, leaving the nine-hour
tetrazole-catalysed reaction mentioned above for 48 h only mar-
ginally reduced the ee (90–84%), indicating that racemization
was not a major event within this timeframe. Presumably, the
tetrazole catalyst promotes the fastest reaction because of
a-amination, it was
a
hyde form, lactol 8 was quite reactive, forming the
a-aminated
product as a 3:2 ratio of diastereomers in high yield with proline
as the organocatalyst within a few hours at room temperature,
and importantly without the need to include a Brønsted activator
(entry 1). However, interestingly, the ee was quite low (60%),
which was likely due to the free hydroxyl in the acyclic hydroxy-
enamine intermediate interfering with the assisted transition-
state. Introduction of acid at low equivalents compared to those
for the optimal acetal conditions, and using the tetrazole catalyst
returned a much higher and pleasing ee of 93% and a high yield
of product (entry 2), endorsing the idea of promoting the assis-
tance through a protonated environment. An extended study
involving pyran-based substrates is included in the Supporting
information.
A final demonstration of the potential scope of the reaction
came in the chemo-differentiation of bis-acetal 10. Dialdehydes
are notoriously prone to self-condensation reactions, whereas
bis-acetals are stable masked equivalents. Applying our conditions
using 10 in excess (5 equiv) followed by the normal borohydride
reduction returned a very good ee of 90% and a good yield (78%)
of the mono-aminated product 11, (Scheme 1), in which one of
Table 2
a
-Amination of a hemi-acetal
DBAD / organocatalyst (20 mol%)
/ acid / H2O (10 equiv)
/ CH3CN (0.5 M)
Cbz
N
H
Cbz
N
10 equiv
Lactol
O
OH
O
OH
8
9
Entry
Catalyst
Acid
No H2O or acid
MCA (0.1) + TFA (0.1)
Time (h)
Temp (°C)
Yield (%)
ee (%) (dr)
1
2
Proline
Tetrazole
4.5
9
20
0
90
80
60 (60:40)
93 (61:39)