Toward the Total Synthesis of Haliclonin A
COMMUNICATION
bonylative cyclization side product 15 and a 70% yield of
the recovered starting material (Table 1, entry 5). Triethyl-
ACHTUNGTRENNUNGamine and ammonium formate were then tested as addi-
from which two pathways are possible. The first, path a, in-
volves an intramolecular insertion reaction, leading to Pd
complex M2, the syn-b-elimination[22g] of which (leading to
the bridgehead enone 3a)[24] is not possible. However, hy-
drolysis of Pd complex M2 via its tautomer M2’ gives the
major product 3.[22i,j] In the second, path b, Pd complex M1
undergoes a decarbonylation reaction to give Pd complex
M3,[22g,23] which goes through an intramolecular insertion re-
action to give M4. Protolysis of Pd complex M4 produces
the side product 15. Use of the bidentate dppp ligand and
MeCN as a chelating solvent inhibits the decarbonylation of
Pd complex M1 to give Pd complex M3, thus preventing the
formation of the side product 15.
tives; however, neither led to any improvement (Table 1,
entry 6). On the other hand, the yield of the desired product
could be improved by increasing the catalyst loading; how-
ever, an increased amount of the decarbonylative side prod-
uct 15 was obtained (Table 1, entries 5, 7, and 8).
These results indicated that a catalytic cycle was not
formed during the reaction.[22i] The decarbonylative side
product 15 might be produced as a result of the oxidative
addition of Pd0 to phenylthiocarbamate 14, followed by a
decarbonylation and intramolecular insertion reaction.[22g,23]
On the basis of this consideration, the addition of a chelat-
ing diphosphine ligand was envisioned to be able to inhibit
this side reaction. Indeed, in the presence of 1.0 equivalent
of dppp, the ratio of 3/15 was slightly improved (Table 1,
entry 9). Further optimization showed that when a combina-
For the aldol reaction of ketone 3 with aldehyde 4, pre-
pared in three steps from 6-bromohex-1-ene,[25] several
methods were attempted. It was found that treatment of 3
and 4 with TiCl4/iPr2NEt[26] gave the optimal result (58%
yield, Scheme 5), providing aldol 16 as the only observable
tion of 1.0 equivalent of PdACHTNUTRGNENG(U OAc)2 and 2.0 equivalents of
dppp was used as the source of Pd0, along with MeCN as a
chelating solvent, the reaction proceeded smoothly at 1008C
to yield the desired product 3 in 79% yield, with only a
trace amount of side product 15 observed (Table 1,
entry 12).
A plausible mechanism for this unprecedented Pd-mediat-
ed cyclization reaction is depicted in Scheme 4. Oxidative
addition of compound 14 with Pd0 gives Pd complex M1,
Scheme 5. Construction of the racemic tricyclic substructure 2.
1
isomer in the H NMR spectrum. The high regio- and dia-
stereomeric selectivity of this reaction was surprising be-
cause several regio- and diastereomers were possible. Since
no isomers were observed in either the 1H or 13C NMR spec-
tra, in order to check the purity of the product the crude
aldol product was subjected to LC-MS analysis. Indeed, only
a trace of an isomer was found, in a ratio of 16/isomer as
high as 56.5:1 (see the Supporting Information for the LC-
MS diagram). Thus, the high regio- and diastereomeric se-
lectivity of this reaction was confirmed.
As displayed in the retrosynthetic analysis (Scheme 1),
the formation of the macrocycle was planned to be accom-
plished by an RCM reaction. For this purpose, diene 16 was
subjected to the RCM reaction by using the Grubbs first-
Scheme 4. Proposed mechanism of the Pd-mediated cyclization.
Chem. Eur. J. 2013, 19, 87 – 91
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
89