Communication
In recent years, however, decarboxylative allylation (DcA) has
become a powerful tool for CÀC bond formation and has
drawn considerable attention due to its high efficiency, mild
reaction conditions, and green features with CO2 as the only
byproduct released.[9] Tunge,[10] Trost,[11] and Stoltz,[12] for in-
stance, have reported a series of catalytic decarboxylative
allylation or benzylation reactions.[13] Given the allylic feature
of enol ethers (3,4-dihydro-2H-pyran, 1), we envisaged that the
use of acetyl acetate as a leaving group could first activate the
pyranose system to promote the palladium-induced p-allyla-
tion and further decarboxylative coupling would facilitate the
CÀC bond formation (Scheme 1c).
Table 1. Ligand screening.[a,b]
We have achieved a stereoselective synthesis of cis-2,6-di-
substituted dihydropyrans by palladium-catalyzed decarboxyla-
tive allylation on readily available sugar scaffolds (glycals),
which allowed the achievement of challenging b-C-glycosyla-
tion.[14] The reaction is a formal intramolecular reaction on a gly-
cosyl acetylacetate scaffold in the presence of Pd(OAc)2 and
1,1’-bis(di-isopropylphosphino)ferrocene (DiPPF). Given the
high efficiency of this palladium-catalyzed decarboxylative cou-
pling reaction, we envisioned that this methodology could be
further applied to the common pyran system to provide a prac-
tical access to naturally occurring cis-2,6-tetrahydropyrans.
Herein, we further report the results of our efforts in extending
this reaction to the synthesis of cis-2,6-disubstituded-3,6-dihy-
dro-2H-dihydropyrans with high regio- and stereoselectivity.
Our investigation commenced with the decarboxylative cou-
pling reaction of (Æ)-3a[15] under the optimized reaction condi-
tions established previously (Pd(OAc)2: 5 mol%, DiPPF:
10 mol%, toluene, 608C; Table 1). After 3 h, the starting materi-
al was consumed and the desired product (Æ)-4a was ob-
tained in 92% yield. However, only a ratio of 7:1 cis to trans se-
lectivity was observed. Further screening of ligands revealed
that bulky ligands could increase the diastereoselectivity.
Among them, ligand B significantly enhanced the selectivity to
20:1, whereas ligand I afforded the cis product as a single
isomer in 91% yield. Other ligands, such as 1,2-bis(diphenyl-
phosphino)ethane (dppe) D and 1,5-bis(diphenylphosphino)-
pentane (dpppen), E gave poor selectivities, whereas xanphos
G and xphos H were observed to provide the desired product
with moderate selectivities, albeit in good yields. Interestingly,
ligand C could not promote this decarboxylative coupling
reaction.
[a] Reactions were carried out on a 0.2 mmol scale in the presence of
0.01 mmol of Pd(OAc)2 and 0.02 mmol of ligand in toluene (2 mL) at
608C. [b] Isolated yields, selectivity was determined by 1H NMR spec-
troscopy.
tion could be tolerated in this decarboxylative coupling reac-
tion and the reactions proceeded smoothly to provide the de-
sired products in good to excellent yields (6h–m).
The synthetic utility of this transformation was further dem-
onstrated by the total syntheses of selected natural products.
We have previously reported the formal total synthesis of as-
pergillide A from a sugar substrate by palladium-catalyzed de-
carboxylative C-glycosylation.[14] In this work, we focused on
the total syntheses of centrolobine 10 and decytospolides 19
and 20.
Centrolobine is a naturally occurring antibiotic, isolated from
the heartwood of Centrolobium robustum, tomentosum, and
the stem of Brosinium potabile.[16] Centrolobines have been
found to exhibit anti-inflammatory, antibacterial, and antileish-
manial properties.[17] The excellent bioactivities and relatively
simple chemical structure have drawn considerable interest to-
wards the total synthesis of centrolobine, both in the racemic
and optically pure forms.[18] Our synthesis started from the ra-
cemic cis-alcohol (Æ)-5. Esterification of 5 with b-keto acid 6
furnished b-keto ester 7, which was submitted to the decar-
boxylative coupling reaction conditions, and, as expected, cis-
2,6-disubstituted-3,6-dihydro-2H-dihydropyran 8 was obtained
in good yield. Further reduction of the carbonyl group of 8
with LiBH4 and triethylsilane produced compound 9 in 84%
yield. However, attempts to reduce the double bond by rou-
tine hydrogenation led to a low yield of the desired product
due to the unexpected ring opening of the tetrahydropyran.[19]
Fortunately, hydrogenation under acidic conditions both re-
The substrate scope was further examined as depicted in
Table 2. Gratifyingly, in all cases, good to excellent yields and
absolute cis selectivity were observed. A variety of functional
groups at the a-position of the ketone could also be tolerated
(4b–g). Interestingly, sterically hindered substituents, such as
isopropyl and tertiary butyl, gave slightly higher yields (6d,
6e) in comparison to the less hindered substituents (6b, 6c).
The reaction of substrates possessing aromatic substituents at
the a-position of the ketone proceeded very well and afforded
the corresponding products in higher yields than that of ali-
phatic substituents (6 f, 6g). The R1 substituent on C6 of the
dihydropyran ring was also modified to further investigate the
functional-group tolerance of the reaction conditions. Both
electron-rich and deficient aromatic substituents at the R1 posi-
Chem. Eur. J. 2014, 20, 405 – 409
406
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim