Journal of the American Chemical Society
Communication
Table 1. Optimal Conditions for Synthesis of α-CF3
a
Allylboronic Acids
a
Boroxine 2a (0.033 mmol, equivalent to 0.1 mmol of the boronic
acid), 3 (0.3 mmol), 4 (0.02 mmol, 20 mol %), and ethanol (0.2
mmol) were reacted in DCM (0.8 mL) for 48 h at 40 °C, and then
Figure 2. Concept of catalytic asymmetric homologation with 1,2-
borotropic migration.
b
DanH (0.15 mmol) was added. Yields of 5a determined by 19F NMR
c
d
spectroscopy. Isolated yields. A complex reaction mixture was
obtained.
omethyl diazomethane 3, catalytic amounts of BINOL
(4),30−36 and stoichiometric amounts of EtOH. Alkenylbor-
oxine 2 readily reacts with diazo compound 3.9−11 However,
this reaction results in racemic product, such as rac-1-OR. The
racemic background reaction can be avoided by addition of
EtOH to the reaction mixture, which forms unreactive
alkylboronic esters 2-OEt, which are weaker Lewis acids37
than the corresponding boroxines 2.30,31 Because of the
dynamic covalent bonding38 ability of boron, BINOL 4
undergoes transesterification with 2-OEt to form chiral alkenyl
boronate A. Exchange of the alkyl group to an aromatic moiety
leads to a substantial increase in the Lewis acidity of boron,37
and therefore, A and 3 form ate complex B in the
stereoinduction step of the process (see Figure S3). Then
the alkenyl group undergoes stereoselective 1,2-migration10,39
to afford C. Subsequently, ethanolysis of C gives product 1-
OEt.
can be rationalized by the racemic background reaction (2 + 3
→ rac-1-OR in Figure 2). The complex reaction mixture is a
consequence of the poor stability of 1 and its boroxine in the
absence of EtOH. Simple aliphatic alcohols esterify the boronic
acids/boroxines and thus protect them from decomposition
under the reaction conditions of the borylation (Figure
1a).2,32,41 When both EtOH and the BINOL catalyst were
omitted (entry 10), a complex reaction mixture was obtained
again. Without molecular sieves (entry 11), the yield was poor,
probably because the slow formation of chiral alkenyl-BINOL-
type intermediate A (Figure 2). At room temperature,
changing dichloromethane (DCM) to toluene leads to
lowering the yield and a slight decrease of the ee (entries
12−13).
The optimal conditions for the homologation involved using
2a with an excess of 3, 20 mol % 4 and 2 equiv of EtOH
(Table 1, entry 1). The oxygen-sensitive allylboronic ester 1a-
OEt was protected with diaminonaphthalene (DanH)40 to give
5a with 98% ee in 69% yield. When the reaction was repeated
with 10 mol % catalyst 4, the yield was substantially lowered
(12%), but the enantioselectivity was practically unchanged
(96% ee) (entry 2). Replacement of iodo-BINOL 4 with
bromo-BINOL (entry 3) led to decreases in the yield (9%)
and the enantioselectivity (88% ee). Interestingly, increasing
the loading of bromo-BINOL (entry 4) to 30 mol % led to a
high yield (73%) and selectivity (94% ee). When bulky γ-
substituents were employed in the BINOL catalyst (entry 5),
both the yield and selectivity strongly declined. Application of
the parent BINOL as the catalyst gave a low yield (4%) and
relatively low selectivity (72% ee). When a commercially
available alkenylboronic acid was used as the substrate (entry
7), the reaction proceeded in poor yield (18%) but with
excellent selectivity (97% ee). When EtOH was replaced by
iPrOH (entry 8), the yield dropped (44%) but the selectivity
was still high (96% ee). In the absence of EtOH (entry 9), a
complex reaction mixture was obtained, from which 5a was
isolated in 4% yield with 47% ee. The poor enantioselectivity
Under the optimal conditions, alkyl-substituted alkenylbor-
onic acids 2a−c reacted readily to give the corresponding α-
CF3 allylboronic acid esters 1(a-c)-OEt and Bdan derivatives
5a−c (Figure 3a). Aryl-substituted alkenylboronic acids (2d−
g) reacted somewhat slower than the aliphatic ones. Cinnamyl
derivative 5d was formed in 54% yield (93% ee) when 20 mol
% catalyst was used. However, with 20 mol % catalyst, 5e
formed only in 26% yield (89% ee). Therefore, the catalyst
loading was increased to 30 mol % to obtain acceptable yields
of 5e−g (50−70%). The absolute configuration of crystalline
5e was determined to be S by X-ray diffraction. On the basis of
the structural similarities of the substrates and the reaction
conditions, we assumed that the absolute configuration of the
other species (5a−d, 5f, and 5g) was the same. The reactions
can be easily scaled up. For example, the synthesis of 5a on 1
and 2 mmol scales occurred with 98 and 96% ee in 78 and 68%
yield, respectively.
The transient allylboron compounds 1-OEt reacted with
aldehyde 6a in situ (Figure 1b). The enantioselectivity for the
formation of 7a−d varied between 90 and 98% ee. In addition,
only one of the four possible diastereomers was formed in each
case. We did not detect any Z isomer of 7a−e in the crude
product of the reaction. Usually, α-substituted allylboron
B
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX