Jiahua Chen, Zhen Yang et al.
Scheme 5. Synthesis of 12. a) PivCl (3.5 equiv), pyridine (solvent), room
temperature, 10 h, 98%; b) l-selectride (2.5 equiv), THF, À788C, 4 h,
93%; c) PDC (1.5 equiv), CH2Cl2, room temperature, 5 h, 87%;
d) KHMDS, 21 (1.0 equiv), THF, À788C, 2 h, 75 % (E/Z=3:1); e) HF/
H2O, CH3CN, 71%; f) AD-mix a (1.4 g/1 mmol), tBuOH in H2O (1:1),
CH3SOONH2, 08C, 7 days, 90%; g) diisopropyl azodicarboxylate (DIAD,
1.0 equiv), PPh3 (1.0 equiv), THF, room temperature, 6 h, 80%.
removal of the silyl group, allylic alcohol 23 underwent
a standard Sharpless asymmetric dihydroxylation[12] with
AD-mix-a as an oxidant to give triol 24 as a single diaste-
reoisomer in 90% yield. For the construction of epoxide 12,
triol 24 was then subjected to a Mitsunobu reaction[20] proto-
col to afford benzylic alcohol 12 in 80% yield.
For the preparation the key intermediate 25 for the AuIII-
catalyzed cycloalkylation, we attempted to couple benzylic
alcohol 12 with phenol 11[21] under a variety of conventional
Mitsunobu reaction conditions (Scheme 6). Unfortunately,
however, these efforts failed to provide the desired product
25, likely because of the steric hindrance of the substrates.
Following a period of systematic experimentation, we dis-
covered that product 25 could be generated as a single dia-
stereoisomer in 50% yield when the reaction was conducted
at 208C under sonication.[22]
Scheme 6. Syntheses of (+)-gallocatechin (3) and (À)-epigallocatechin
(4). a) DIAD (1.0 equiv), PPh3 (1.0 equiv), THF, 208C, 20 min, sonica-
tion, 50%; b) AuCl3 (0.1 equiv), AgOTf (0.3 equiv), PPh3 (0.1 equiv), 1,2-
dichloroethane, 508C, 7 h, 81%; c) Pd/C, H2 (balloon pressure), 94%;
d) LiAlH4 (10 equiv), THF, À788C to room temperature, 2 h, 42%;
e) Dess–Martin periodinane (DMP) (1.2 equiv), CH2Cl2, room tempera-
ture, 2 h, 65%; f) l-selectride (2.0 equiv), THF, À788C, 2 h, 87%; (g) Pd/
C, H2 (balloon pressure), 90%; h) LiAlH4 (10 equiv), THF, À788C to
room temperature, 2 h, 51%.
With the key intermediate 25 in hand, we then set out to
investigate our previously developed thiourea/AuIII-cata-
lyzed annulation. Unfortunately, under various reaction con-
ditions,[7b] product 9 was obtained in low yield, although dif-
ferent types of thiourea ligands, such as thioureas I, II, and
III (Scheme 6).
To improve the yield, we resort to electronic tuning of the
Au catalyst through its ligation with other types of ligands,
such as lutidine,[23] PPh3,[24] PCy3, BINAP (2,2’-bis(diphenyl-
phosphino)-1,1’-binaphthalene), and bisoxazoline,[25] and
found that the best result could be obtained when the reac-
tion was conducted in the presence of an AuCl3/AgOTf/
PPh3 mixture, to give the protected gallocatechin 9 endo-se-
lectively[7a] in 81% yield.
ty of different bases, including H2NNH2, NH3/MeOH,
K2CO3, and NaOMe, only gave low yield of desired product,
indicating that the resulting product was unstable under
basic conditions.[26] We therefore adopted a reverse depro-
tection sequence, with the debenzylation being affected first
followed by LiAlH4-mediated depivoylation.[27] Pleasingly,
this new protocol allowed for the successful synthesis of
(+)-gallocatechin (3) in 42% yield via a debenzylation–de-
pivoylation sequence (Scheme 6).
For the synthesis of (À)-epigallocatechin 4, substrate 9
was first subjected to a two-step C3-inversion protocol,[9] in-
volving a DMP oxidation and subsequent diastereoselective
l-selectride reduction to give product 26. Thus, following
the deprotection protocol developed above, (À)-epigalloca-
techin 4 was made in 51% yield. The physical properties of
the synthetic 3 and 4 were compared with those reported for
To complete the total synthesis of the natural product
(+)-gallocatechin 3, we originally planned to initially the
pivaloyl groups with subsequent hydrogenative debenzyla-
tion. Unfortunately, however, the treatment of 9 with a varie-
Chem. Asian J. 2013, 8, 700 – 704
702
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim