Angewandte
Communications
Chemie
allylic alcohols (Figure 2B). Specifically, we demonstrate
that, when employing the novel, easily prepared chiral
oxazoline ligand t-Bu-Pmrox (L1), highly efficient and
selective 1,2-reduction of a large collection of enones can be
achieved by NiH catalysis.[10]
We began our study by optimizing the conditions for the
reduction of (E)-4-phenylbut-3-en-2-one (S1). Investigation
of a range of parameters showed that the desired 1,2-adduct
could be obtained in excellent yield (99%) and ee (> 99% ee)
within 40 min (Table 1, entry 1). Ligand L1 (t-Bu-Pmrox) was
hydrosilanes) were comparatively ineffective (entry 9).
Finally, THF was also found to be a suitable solvent for this
reaction (entry 10).
With our optimized conditions, we next examined the
substrate scope. As illustrated in Table 2a, an array of b-aryl
substituted enones could be converted into secondary allylic
alcohol in a highly enantio- and 1,2-selective manner. A wide
variety of substituents on the b-aryl ring, including electron-
withdrawing (2–7) and electron-donating (8–11) substituents
were well-tolerated. It is noteworthy that, under these
exceptionally mild reaction conditions, potentially sensitive
functional groups, including a boronic acid pinacol ester (4),
an aryl chloride (5), a bromide (6), an iodide (7), and a triflate
(11), were left intact. An a,b-unsaturated amide (12) was also
compatible, thus highlighting the excellent chemoselectivity
of the reaction. Under these conditions, an unprotected
phenol could also be used as a substrate (10) if an additional
equivalent of pinacolborane was included. The phenol
hydroxy group is presumably transformed into the corre-
sponding boronate ester, which then undergoes the desired
1,2-reduction to give the desired a-chiral allylic alcohol
product upon chromatographic purification. In addition,
heteroaromatic substrates, such as those containing a furan
(13) or an indole (14), were also suitable for this reaction.
Moreover, chalcone, a substrate bearing an aryl substituent
on the carbonyl, also underwent reduction in high conversion
and enantiomeric excess, although the ambidoselectivity was
somewhat diminished compared to a’-alkyl substrates (15).
The reaction was also tolerant of a broad array of b-alkyl-
substituted enones. As shown in Table 2b, a variety of
functional groups were again readily accommodated, includ-
ing a secondary carbamate (18), ethers (19, 20), an alkene
(22), and a silane (23). Moreover, a’-substituents other than
a methyl group were handled without incident (17, 18, 21).
Moreover, the configuration of a conjugated double bond was
retained during the reaction (21), and the presence of
extended conjugation in b-ionone did not interfere with the
1,2-reduction (22). Finally, a b-silylated enone also underwent
1,2-reduction to afford highly enantioenriched allylic alcohol
containing a g-silicon substituent (23) suitable for subsequent
electrophilic functionalization.
Table 1: Variation of reaction parameters.
Entry
Variation from
1:1’[a]
Yield of 1[b]
ee of 1[c]
“standard” conditions
1
2
3
4
5
6
7
8
9
none
>25:1
>25:1
1:1
1:1
10:1
11:1
>25:1
ND
1:2
99
98
7
>99
99
ND
ND
97
84
0
ND
ND
99
L2 instead of L1
L3 instead of L1
L4 instead of L1
RT instead of À258C
w/o DABCO
w/o DABCO and Ni/L
w/o Ni/L
PMHS instead of HBpin
THF instead of toluene
4
91
92
34
NR
2
10
>25:1
98
1
[a] The ratios of 1/1’ were determined by H NMR spectroscopy of the
crude reaction mixtures. [b] Yields are yield of isolated product (average
of two runs). [c] Enantioselectivities were determined by chiral HPLC
analysis; the absolute configuration was assigned by chemical correla-
tion or analogy (See the Supporting Information for experimental
details).
found to be the most effective for this reaction. t-Bu-Pmrox is
easily accessible in one step from the corresponding chiral
amino alcohol and was readily prepared on a decagram scale
(Figure 2B, see the Supporting Information for synthetic
details). Although the use of the known ligand t-Bu-Pyrox
(L2) led to a comparable result (entry 2), it was subsequently
found that L1 provided high 1,2- and enantioselectivity for
a broader range of substrates. On the other hand, poor yields
were obtained when Box and phosphine ligands were used
(entries 3, 4). Conducting the reaction at RT instead of À258C
led to somewhat lower 1,2-selectivity (entry 5). Furthermore,
lower levels of 1,2-selectivity and enantioselectivity were
observed in the absence of DABCO as an additive (entry 6).
Control experiment showed that HBpin slowly reduced S1 in
the absence of Ni and ligand. The addition of DABCO
inhibited this background reaction, which accounts for the
role of this additive in improving overall levels of selectivity
(entries 7, 8). Additionally, other hydride sources (like
The broad scope of this 1,2-reduction method was further
demonstrated using a range of less reactive and more
challenging b,b-disubstituted enones[3f,11] (Table 2c). Indeed,
both b-aryl-b-alkyl-substituted and b,b-dialkyl-substituted
enones were found to work well. Moreover, substrates
containing electron-donating (25) or electron-withdrawing
(26) substituents on the b-aryl ring both proved compatible,
maintaining both high yield and excellent enantiomeric
excess. In addition, a heteroaromatic substrate based on
thiophene (27) could be successfully applied in this reaction,
again delivering the corresponding chiral alcohol in high yield
and ee. Notably, a’-substitution with an alkyl group other than
methyl was tolerated without issue (28). In the case of b,b-
dialkyl-substituted enone 29, high enantioselectivity was
obtained, although the reaction rate was somewhat attenu-
ated compared to b-aryl-b-alkyl-substituted enones.
Enantioenriched protected allylic alcohols are valuable
synthetic intermediates. To show the utility of this method in
2
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
These are not the final page numbers!