RESEARCH ARTICLE
by the alkoxide base and inadvertent water, and thus were appropriate for the formation of corresponding
decided to include additional proton sources (entries 4 indanols (2d and 2e), but dihydronaphthalenol for-
and 5). Indeed, the addition of 1 equivalent of MeOH mation were less efficient, resulting in decreased
with LiOMe as the base offered the best results, yield or ee.
yielding 2a in 86% yield with 96:4 diastereomeric
ratio, which was chosen for the optimal conditions for aromatic ring of dienyl arenes were investigated. In
synthesis of 5-membered benzocycles 2 (entry 5). general, all substrates with either electron-donating
The effect of different substituents (R2) on the
Phosphine ligands such as PPh3 and Xantphos also (1f–1h) or electron-withdrawing substituents (1i–1l)
produced 2a, but in lower yields (entries 6 and 7). produced the corresponding indanols (2f–2l) in mod-
When the C2 symmetric chiral bisphosphine ligands, erate to good yield with high diastereoselectivity under
BINAP and Segphos were used with the NaOtBu base, condition A. However, in cases of 6-membered ring
1
both 2a and 3a-Bpin formed in a ~1:1 ratio via H formation, substrates with an electron-donating or
NMR analysis, and the corresponding hydroxy com- electron withdrawing group were slightly less efficient
pound 3a was isolated in moderate yield and ee after than neutral substrates, giving cyclized products (3) in
oxidation instead of unstable 3a-Bpin (entries 8 and moderately reduced yield or ee. Especially, in the case
9). The use of Josiphos ligand was not suitable for the of the strongly electron-withdrawing CF3-containing
formation of 3a, yielding a racemic product (0% ee) substrate 1i, the intramolecular copper-allyl cyclization
despite its improved yield. Finally, the optimal ligand was not very effective, forming the indanol byproduct
for the asymmetric formation of 3a was found to be 2i in a ~1:1 ratio.
(S,S)-Ph-BPE ligand with a high enantioselectivity
Next, we tried to explain the reaction pathways to
(92% ee) and with no detectable formation of 2a. The benzocycles 2 and 3 (Scheme 3). With IMesCu
relative configuration of 2a[10] was determined by catalyst, when the reaction was carried out without
NOE experiments (See SI, Fig S1) and the absolute additional MeOH in a glovebox, a low yield of 2a was
configuration of 3a was determined by X-ray crystal obtained. The yield increased to 67% through the
analysis of the corresponding 4-bromobenzoate deriva- addition of 0.5 equiv. of MeOH. Moreover, the
tive of 3a, which contained two molecules in a unit addition of 1 equiv. of MeOD resulted in 66%
cell (Figure 2).[11]
deuterium incorporation to the benzylic carbon of 2a,
Having established conditions for 2a and 3a, we indicating protonation of the benzylic CuÀ C bond
surveyed the scope of copper-catalyzed chemodiver- (Scheme 3a). Therefore, for IMesCu catalyst,[8a] it was
gent intramolecular cyclization of dienyl arenes (2) proposed that copper-catalyzed intramolecular addition
using condition A or B (Scheme 2). First, scaled-up of IMes-copper allyl intermediate III was not facile,[12]
reactions of 1a under both conditions were carried undergoing protonation of the CuÀ C bond by proton
out and each product was formed with no significant sources to yield (Z)-allylboronate IV.[13,14] Subsequent
change in yield or enantioselectivity. Dienyl arenes intramolecular cyclization of the allylboronate[15] to the
with an alkyl-substituted ketone moiety (R1 = alkyl; carbonyl group produced racemic indanols 2 with high
1b and 1c) were suitable for both reactions, yielding diastereoselectivity. Also, DFT calculations showed
each benzocycle in good yield and with high that IMes-copper allyl complex III did not allow
enantioselectivity for dihydronaphthalenols (3b and structural rotations necessary for the copper-allyl
3c). Substrates with an aromatic substituent (R1 =Ar) addition, requiring high activation energy of 20.2 kcal/
mol (See SI, Fig S4).
In the case of copper-BPE catalyst, copper-allyl
addition to the ketone moiety[17] efficiently occurs,
yielding dihydronaphthalene derivative (3a-Bpin) with
good enantioselectivity. Detailed density functional
theory (DFT) calculations were carried out (Sche-
me 3b). Copper-allyl intermediate I, a formal 1,4-
borocupration product is generated by 1,2-borocupra-
tion of the diene 1a and subsequent allylic
isomerization.[8,16] Then, intermediate I undergoes
intramolecular cyclization to produce the dihydronaph-
thalenol derivative (3a) through copper-catalyzed
allylation. Although attaining a favorable transition
state was expected to be difficult due to a low level of
flexibility, the copper-BPE catalyst was highly efficient
for preferred production of 3a via a boat-like transition
state with relatively low activation energy of 3.1 kcal/
mol (TS-I). Overall, the proposed catalytic cycles for
Figure 2. X-ray Structure of 4-bromobenzoate of 3a.
Adv. Synth. Catal. 2021, 363, 1–8
3
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!