Organic Letters
Letter
allylic alcohols were screened (3r−3ac). As expected, a variety
of α-aromatic substituted allylic alcohols were found to be
suitable coupling partners regardless of the electron-with-
drawing or -donating group on the phenyl ring (3s−3w).
Remarkably, either acyclic (3x, 3y) or cyclic (3z) alkyl-
substituted allylic alcohols were well tolerated, affording the
corresponding 1,4-dienes in moderate to high yields. Even the
tert-allylic alcohol 2-phenylbut-3-en-2-ol, which favored the
competitive self-dehydration process with the generation of
conjugated 1,3-dienes, can be successfully incorporated into
the desired cross-coupling with butyl acrylate, efficiently
producing 3aa. With the attempt to access 3-substituted 1,4-
diene skeletons, (E)-4-phenylbut-3-en-2-ol was then selected
as the starting material. To our delight, the corresponding 3ab
was also successfully generated in moderate yield. Remarkably,
the β-blocked allylic alcohol 2-methyl-1-phenylprop-2-en-1-ol
was identified as a powerful precursor, affording 3ac in 67%
yield.
3) when reacted with butyl acrylate. These results reveal that
the same π-allylpalladium intermediates can be generated from
these two kinds of isomeric allylic alcohols (3r−3t). In
addition, sterically demanding cinnamyl alcohol derivatives can
be successfully incorporated into this transformation, albeit
affording 3ad in a slightly lower yield. (E)-Oct-2-en-1-ol was
also proven to be an amenable substrate for this protocol
(3ae). As presented in Scheme 3, a variety of alkenes readily
participate as coupling partners in this reaction with cinnamyl
alcohol. In addition to the standard butyl acrylate 2a, volatile
ethyl acrylate (3af) and multifluoro-substituted butyl acrylate
(3ag) also provided the respective corresponding products
effectively. Even the highly active ketone carbonyl groups
could be introduced into the corresponding 1,4-dienes when
oct-1-en-3-one was selected as the initial substance (3ah). To
our satisfaction, other electron-deficient alkenes, such as
phenyl ethenesulfonate (3ai) and diethyl vinylphosphonate
(3aj), were proven to be viable coupling partners for this
reaction. In addition to terminal olefins, diethyl maleate (3ak)
and (E)-3-(dimethylamino)acrylonitrile (3al) can also react
with cinnamyl alcohol well, affording the corresponding 1,4-
dienes in moderate to high yields. The L-(−)-methenyl group
can be incorporated into multifunctionalized 1,4-dienes (3am).
This newly developed catalytic protocol is not only suitable for
electron-deficient alkenes but is also applicable to styrene
derivatives (3an−3ap), regardless of the electron-withdrawing
or electron-donating groups on the phenyl ring. α-Alkyl- (3aq)
or aryl-substituted (3ar) styrenes smoothly completed the
desired cross-coupling process. Pleasingly, benzofuran (3as)
and 1H-indene (3at) were identified as practical substances
involved in this dehydrative cross-coupling. Unfortunately,
only a trace amount of the desired product was detected when
simple olefins such as dec-1-ene (3au) were used as the
coupling partner. Remarkably, the reaction also can perform
well on gram scale affording 3r in 1.43 g (73% yield), which
has been proven as a powerful synthetic intermediate for
multifunctionalized pyrrolidines.21
pri-Allylic alcohols were then explored under identical
conditions. As shown in Schemes 2 and 3, only the linear
selective products (3r−3t) were obtained with either 1-aryl-2-
en-1-ol (Scheme 2) or cinnamyl alcohol derivatives (Scheme
Scheme 3. Dehydrative Cross-Coupling between Primary
a
Allylic Alcohols and Various Alkenes
Next, we examined the allyl−alkene coupling concerning
several biologically active skeletons. The acrylates, bearing the
motifs of α-D-galactopyranose or cholesterol, can be consumed
smoothly, affording the corresponding 3av and 3aw in high
yields. Even an estraliol moiety with an active hydroxyl group
was also well tolerated (3ax). These results further highlight
the utility of this method in pharmaceutical-related studies.
Recently, 4,4′-((1E,4E)-penta-1,4-diene-1,5-diyl)bis(2-me-
thoxyphenol) (3ay) was recently found to be highly potent
against cancer and human promyelocytic leukemia (HL-60
cells with an IC50 of 65 μM).22 With our developed method,
the desired 3ay can be conveniently obtained in moderate
yield, which is much better than the reported method (less
than 10% yield22).
Having demonstrated the versatility of the resultant
products, we attempted to gain insight into the key step of
C−OH bond activation. We preliminarily studied the binding
of allylic alcohols and Ca(NTf2)PF6 by using the method of
continuous variation (Job’s method), which revealed that a 1/1
stoichiometry for the Ca(NTf2)PF6/C−OH complex was
formed, regardless of cinnamyl alcohol or MBH versions
(Scheme 4a). Moreover, diffusion-ordered spectroscopy
(DOSY) can also account for the formation of the
aforementioned complex (Scheme 4b). For cinnamyl alcohol
(Da = 11.6 * 10−10 m2 s−1), the translational diffusion
coefficient significantly decreased to Da = 8.9 * 10−10 m2 s−1
a
Reaction conditions: 1 (0.4 mmol), 2 (1.2 mmol), and Pd(PPh3)4 (5
mol %), Ca(NTf2)2 (10 mol %), KPF6 (10 mol %) in 4.0 mL of
solvent (DME) in argon atmosphere at 100 °C for 12 h. Isolated total
yield. Only one isomer was detected. Two isomers were detected
b
(5E/5Z = 10/1).
4370
Org. Lett. 2021, 23, 4368−4373