Angewandte
Communications
Chemie
Table 1: Optimization of reaction conditions.[a]
treated as a potential strategy for asymmetric synthesis.
However, owing to multiple formidable challenges,[16] it has
rarely been realized so far.[17] Recently, we developed an
efficient method to access axially chiral biaryls via Pd0/chiral
NBE* cooperative catalysis.[17c] Inspired by this chemistry, we
envisioned a strategy for KR of tertiary benzyl alcohols. As
shown in Figure 1C, aryl iodide 2 undergoes sequential
Entry
Ligand
[NBE]
ee [%][b]
C [%][c]
S [%][d]
À
oxidative addition, chiral NBE* insertion and ortho-C H
1a
3a
activation to generate the chiral aryl-NBE palladacycle
(ANP) species I, which is then selectively oxidized by one
enantiomer of the racemic tertiary benzyl alcohol bearing an
aryl bromide motif (eg. (S)-1) to produce the PdIV intermedi-
ate II, leaving the other enantiomer (eg. (R)-1) unreacted.
Intermediate II subsequently undergoes reductive elimina-
tion and b-carbon elimination to form axially chiral complex
III and replenish the chiral NBE*. Finally, an intramolecular
etherification occurs via complex III to give the central chiral
benzo[c]chromene product (S)-3,[18] alongside disappearance
of the transient axial chirality. Potential features of this KR
method included the unprecedented resolution mode, the use
of simple aryl iodides as the resolution reagent and providing
two valuable chiral molecules in one operation. Nevertheless,
multiple challanges are foreseeable. For example, the iden-
tification of a suitable chiral NBE* co-catalyst to ensure both
good reactivity and enantiodiscrimination,[16] the innate low
reactivity of 2,6-disubstituted arylbromide motif of the
tertiary alcohols,[17c] and the dichotomy between required
high reaction temperature and expected good KR effciency.
To probe this challenging process, we initiated our studies
with a model reaction between racemic tertiary benzyl
alcohol 1a (1.0 equiv) and 1-iodonaphthalene 2a (0.5 equiv)
in the presence of Pd(OAc)2 (2.5 mol%), tri(2-furyl)phos-
phine (TFP) (7.5 mol%), and (1S,4R)-2-methyl ester-substi-
tuted NBE (N1*, 99% ee, 50 mol%)[17a] (Table 1). Gratify-
ingly, KR of racemic tertiary benzyl alcohol (Æ)-1a was
indeed achieved with good efficiency (the selectivity factor
(S) is 104), providing the desired benzo[c]chromene product
3a with 96% ee and recovered 1a with 71% ee (entry 1).
Subsequently, the KR effect of other chiral NBEs* was
investigated. The C2-ethyl ester substituted chiral NBE*
(N2*)[17b–c] behaved similarly to N1* (entry 2), and the C2-
amide substituted chiral NBE* (N3*)[16] led to erosion of the S
factor (entry 3). In sharp contrast, the C5-methyl ester
substituted chiral NBE* (N4*) resulted in almost a complete
loss of KR selectivity (entry 4). Then we investigated the
effect of phosphine ligand on KR efficiency (entries 5–8). The
use of PPh3 as the ligand led to a lower conversion (entry 5),
while other ligands such as JohnPhos, XPhos or PCy3, resulted
in low efficiency or even no reaction (entries 6–8). Notably,
the loading of N1* could be reduced to 30 mol% with a slightly
lower conversion (entry 9). Lastly, increasing the amount of
2a (to 0.8 equiv) and TFP (to 10 mol%) gave the best results
(51% conversion, S = 127), which was determined as the
optimal reaction conditions (entry 10).
1
2
3
TFP
TFP
TFP
N1*
N2*
N3*
N4*
N1*
N1*
N1*
N1*
N1*
N1*
71
68
89
21
44
10
0
0
53
96
96
96
92
18
95
94
–
–
96
94
43
41
49
54
31
10
0
0
36
51
104
100
72
2
60
36
–
–
83
127
4
TFP
5[e]
6[e]
7[e]
8[e]
9[f]
10[f,g]
PPh3
JohnPhos
XPhos
PCy3
TFP
TFP
[a] All reactions were performed on a 0.2 mmol scale. [b] Determined by
chiral HPLC analysis. [c] Conversion (C)=ees/(ees +eep). [d] S=ln-
[(1ÀC) (1Àeep)]/ln[(1ÀC) (1+eep)]. [e] N1* (97% ee) was applied.
[f] 30 mol% of N1* was applied. [g] 10 mol% of TFP and 0.8 equiv of 2a
were applied.
providing the desired chiral benzo[c]chromene products as
well as the recovered 1a with good to excellent enantiopurity
(78–99% ees). In general, electron-withdrawing aryl iodides
were inferior to the reaction (2e and 2m) and provided
smaller S factors. Besides the alkyl substituents (2b–d),
various functional groups of aryl iodides, including fluoro (2 f,
2h and 2k), chloro (2g), bromo (2l and 2n), trifluoromethyl
(2e) and nitro (2m) groups were tolerated. In addition,
bicyclic aryl iodides (2n and 2o) and heteroaryl iodide (2p)
were competent to afford excellent KR results.
Subsequently, the reaction scope of racemic tertiary
benzyl alcohols (1) were evaluated, with 1-iodonaphthalene
2a as the resolution reagent (Table 3). Firstly, we investigated
the aryl/Me-type tertiary benzyl alcohols, and found that good
KR results could be obtained with various groups (eg. methyl,
fluoro and methoxy) at different positions of the phenyl motif
(1b–g). The S factor reached 544 for the alcohol with an
ortho-methyl phenyl group (1d), although longer reaction
time was required (probably due to the increased steric
hindrance). Notably, tertiary alcohols with a heteroaryl motif
(1h and 1i) were also smoothly resolved with good efficiency.
Next, the more challenging alkyl/Me-type tertiary alcohols
were investigated. After screening, we found only the ones
with a proper steric differentiation of the two alkyl groups (1j
and 1k) gave good S factors. We also explored the reaction
scope of Ph/alkyl-type tertiary alcohols. Both the Ph/Et- and
Ph/CF3-substituted ones (1l and 1m) afforded good S factors.
Finally, we probed the scope of ortho substituent of the aryl
bromide moiety of tertiary alcohols, and found ethyl (1n),
cyanomethyl (1o), TBS-protected hydroxymethyl (1p), N,N-
dimethylaminomethyl (1q), pyrrolidin-1-ylmethyl (1r), mor-
With the optimal reaction conditions established, we first
set out to explore the reaction scope of aryl iodide 2, with (Æ)-
1a as the reaction partner (Table 2). To our delight, a wide
range of aryl iodides containing electron-donating or elec-
tron-withdrawing groups all proved to be suitable substrates,
2
ꢀ 2021 Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2021, 60, 1 – 6
These are not the final page numbers!