Journal of the American Chemical Society
Article
competitive SN2 reaction of 1e with in situ generated α-
borylcarbanion.8 The absolute stereochemistry of the major
enantiomer of 3c was determined as S after oxidation of
B(nep) to a known compound.9
Scheme 1. Copper-Catalyzed Enantioselective Coupling of
gem-Diborylalkanes with Allylic Electrophiles
Substrate Scope. With the optimized conditions in hand,
the scope of allyl bromides was explored using 2c as a coupling
reagent (Table 2a). Allyl bromides bearing methyl-, hexyl-, and
cyclohexyl group as a substituent at C2-position furnished 3c−
3f in good yields and er. Substrates containing an ester, a
protected amine, a tert-butyldimethylsilyl (TBS)-protected
ether or a trimethylsilyl (TMS) group resulted in the formation
of 3g−3j with good to high selectivity. Furthermore, 4- bromo-
2-(bromomethyl)but-1-ene was successfully reacted with 2c,
leaving the bromo group intact (3k) for further elaborations.
Reactions of allyl bromides bearing alkenyl moiety smoothly
underwent the allylation, affording 3l and 3m in good
efficiency. Whereas CuH-catalyzed coupling of alkenes with
allylic electrophiles has been well-established,10 the reaction
scope is rather limited especially for substrates containing
additional alkene substituent because of competitive hydro-
cupration. Therefore, our developed protocol offers an
attractive alternative to CuH-catalyzed allylation of alkenes.
Various C2-aryl-substituted allyl bromides bearing electroni-
cally neutral, donating, or withdrawing substituents led to
products 3n−3r in good yields with high er. Naphthyl- and
heteroaryl-containing 3-bromoprop-1-ene readily engaged in
the reaction leading to 3s and 3t. 2-Bromo-substituted and
simple allyl bromides also underwent the allylation, delivering
3u and 3v. Next, we investigated the scope of gem-
diborylalkanes under the standard reaction conditions.
Reactions of gem-diborylpropane and gem-diboryl-3-phenyl-
propane with 1e yielded 4a and 4b in good to moderate yields
with good er. gem-Diborylalkanes bearing a TBS-protected
alcohol and alkenes led to 4c−4e in good efficiencies. 2-
Bromo-substituted and simple allyl bromides also proceeded to
complete the coupling with various gem-diborylalkanes to give
corresponding products 4f−4i. Interestingly, the reaction of 2-
phenylallyl bromide and complex gem-diborylalkanes contain-
ing pinacolato groups, derived from lithocholic acid and liquid
crystal,11 furnished 4j and 4k in good yields with high
stereoselectivity.
Mechanistic Studies. To understand how the enantio-
topic-group-selective transmetalation occurs between gem-
diborylalkanes and copper catalyst, we performed quantum
mechanical calculations based on density functional theory
(DFT). The theoretical investigation utilizes gem-diborylalkane
2c as a representative substrate, L5 as the ligand bound to
copper, and LiOtBu as the base. Figure 1 shows the calculated
free energy profile of the enantiotopic-group-selective trans-
metalation to furnish chiral copper species C. The reaction
model starts from CuOtBu, which is formed by ligand
exchange of CuBr. Once tBuO−Cu(L5) is formed to
accommodate the gem-diborylalkane substrate, two mecha-
nistic scenarios of transmetalation can be imagined. One
involves the assistance of LiOtBu as marked in blue and red
trajectories, while the other excludes participation of the base
as shown in the black dotted line. Our calculations indicate
that LiOtBu renders the transmetalation much more viable. As
illustrated in Figure 1, LiOtBu first binds to the tBuO−Cu(L5)
intermediate to form a cyclic Lewis acid−base pair, A, which
can act as a bridge during the C−B bond cleavage and Cu−C
bond formation between the copper complex and 2c when
traversing what could be best characterized as an open
diborylalkanes with chiral copper catalyst proceeds. Further
synthetic transformations of the obtained enantioenriched
homoallylic boronate esters are also demonstrated.
RESULTS AND DISCUSSIONS
■
Optimization Studies. We tested the reaction of 2-
benzylallylic electrophiles (1) and gem-diborylethane bearing a
pinacolato moiety (2a) in the presence of CuBr as a catalyst,
(R)-BINOL-derived phosphoramidite as a ligand (L1), and
LiOtBu as a base. Although no reaction took place when 2-
benzylallyl acetate (1a) was employed as an electrophile
(Table 1, entry 1), the reactions of 2a with tert-butyl-(2-
benzylallyl)carbonate (1b) or diethyl-(2-benzylallyl)-phos-
phate (1c) afforded desired allylation product 3a in low to
high yields (entries 2 and 3) with poor enantiomeric ratios
(er). The er of 3a was increased up to 70:30 when (2-
(chloromethyl)allyl)benzene (1d) was used as an electrophile
(entry 4), and (2-(bromomethyl)allyl)benzene (1e) gave 3a
with a slightly higher er (entry 5).7 Next, we surveyed the
effect of an amino group of (R)-BINOL-derived phosphor-
amidite ligands (L2−L4) and found them to have negligible or
negative effects on the er (entries 6−8). Pleasingly, subjecting
the (R)-H8−BINOL-based phosphoramidite (L5) as a ligand
afforded 3a in good yield with 95:5 er (entry 9).
Spirobiindanediol (L6) or TADDOL-derived phosphoramidte
(L7) ligand displayed lower efficiencies (entries 10 and 11).
To improve the enantioselectivity further, substituent effects of
boronate ester were subsequently examined. While gem-
diborylalkane containing 1,3-propanediolato group(2b)
showed low efficiency and selectivity (entry 12), the
employment of gem-diborylalkane having neopentylglycolato
group (2c) delivered 3c in good yields with the highest er
(entry 13) found in this series. Increasing the reaction
temperature to 50 °C shortened the reaction time (12 h),
even though 3c was obtained in a slightly decreased er (entry
14). Using NaOtBu (entry 15) or KOtBu (entry 16) instead of
LiOtBu gave 3c in low to poor er, probably because of the
1070
J. Am. Chem. Soc. 2021, 143, 1069−1077