Angewandte
Chemie
subtype represented by 5 and 6 as precursors to controllably
access structurally and biosynthetically diverse architectures
would be enhanced.
gram scale; 2) extensive efforts to form a variant of 14 by
Stobbe condensations (with a free carboxylic acid instead of
the tert-butyl ester) proved low yielding and capricious,
particularly on scale;[7] 3) attempted DIBAL-H reduction of
the ester within 15 into the aldehyde failed, with only PDBBA
(formed by admixing DIBAL-H with KOtBu)[8] giving the
desired chemoselectivity;[9] and 4) commercial sultone 17 had
to be recrystallized prior to use to achieve a high yielding
alkali fusion reaction en route to 18.[10]
With these fragments in hand, they were then united into
key intermediate 19 (a defined form of retron 5, see
Scheme 1) in 58% overall yield by an initial Horner–Wads-
worth–Emmons olefination between the anion derived from
11 and aldehyde 16, and subsequent halogen–lithium
exchange and nucleophilic attack onto the aldehyde function
of 18 (Scheme 3). As such, we could now test our ability to
convert this material into the entire dalesconol framework.
After extensive studies, this goal was indeed realized;
Scheme 3 presents the sequence in its current level of
optimization.
Our explorations to test this overall hypothesis began with
the preparation of three phenolic precursors, which were
anticipated to come together to form 5 through the retro-
synthetic disconnections indicated in Scheme 1; dalesconol B
(2) was specifically targeted. After several rounds of protect-
ing group selections to achieve proper reactivity in later steps
(see below), fragments 11, 16, and 18 were smoothly prepared
from commercial materials in four, five, and five linear steps,
respectively (Scheme 2). Given the conventional nature of
many of these operations, a detailed discussion of the entire
sequence is not warranted. However, we do wish to note the
following: 1) each fragment was readily synthesized on multi-
In the event, compound 19 was taken up in a mixture of
EtOAc and EtOH (2:3) and subjected to 1 atmosphere of H2
gas in the presence of a full equivalent of Pd/C (10%); under
these specific conditions, the benzyl protecting group was
excised and the double bond uniting the A and B rings was
reduced in quantitative yield. Use of any other solvent
combinations or ratios, as well as catalytic loadings of
palladium, led to significant amounts of material in which
the alcohol group on the carbon atom bridging the A and
E rings was reduced as well. After filtration and solvent
removal, the crude residue was resuspended in 2,2,2-trifluoro-
ethanol and treated with a full equivalent of TFA at ꢀ458C
for 15 minutes. During this time, the alcohol function was
ionized, thereby initiating a Friedel–Crafts reaction which
generated the seven-membered ring within 21.[11] Subsequent
addition of 1.1 equivalents of PhI(OAc)2 to the same pot at
ꢀ458C, and then 20 minutes of additional reaction time
converted the strategically deprotected phenol (B ring) into
an oxidized material with a para-disposed carbocation that
was engaged by the D ring to fashion the complete dalesconol
core as expressed in 22.[12] Globally, these operations provided
22 in 32% yield upon isolation, thereby accounting for an
overall efficiency level of 75% per step based on its four
distinct operations.
Having completed this critical operation, the completion
of dalesconol B (2) required several adjustments in oxidation
state prior to removal of the phenolic protecting groups. The
first of these events, hydrogenation of the double bond within
22, occurred chemoselectively when performed in a 3:1
mixture of EtOH and EtOAc at 258C. This step provided 23
as a single diastereomer of unknown configuration in 84%
yield;[13] other solvents or prolonged reaction times led to
unwanted conversion of the benzylic ketone into the corre-
sponding alcohol as well. After removal of the MOM-
protecting group (HCl, THF) and DDQ-mediated oxida-
tion[14] into the corresponding para-quinone methide, an X-
ray crystal structure of the resultant intermediate (not shown,
see the Supporting Information) confirmed the stereochem-
istry as that desired for the target structure and as drawn in
Scheme 2. Synthesis of key phenolic building blocks 11, 16, and 18:
a) POCl3 (3.0 equiv), DMF (6.0 equiv), 908C, 6 h; aq. KOH, 0!258C,
12 h, 99%; b) NaBH4 (2.0 equiv), MeOH, 08C, 30 min, 96%; c) PBr3
(1.0 equiv), pyridine (cat.), Et2O, 258C, 4 h, 96%; d) KHMDS (0.5m in
toluene, 1.8 equiv), HP(O)(OEt)2 (2.0 equiv), 08C, 15 min; then start-
ing material added, THF, 0!258C, 12 h, 94%; e) LiCl (1.3 equiv),
DBU (1.0 equiv), 13 (1 equiv), CH3CN, 258C, 12 h, 99%; f) TFA/H2O
(9:1), 258C, 90 min; NaOAc (1.4 equiv), Ac2O, 1408C, 1 h, 83%; g) H2
(1 atm), Pd/C (10%), MeOH/CH2Cl2 (2:1), 258C, 24 h; filter, NaOMe
(3.0 equiv), 0!258C, 2 h, 99%; h) NaH (2.0 equiv), BnBr (2.0 equiv),
DMF, 0!258C, 1 h, 77%; i) PDBBA (0.9 equiv), THF, ꢀ20!08C,
1.5 h, 67%; j) NaOH/KOH/17 (1:5:1 by weight), 2108C, 40 min, 53%;
k) NaH (1.0 equiv), THF, 08C, 10 min; Me2SO4 (1.0 equiv), 0!258C,
14 h, 99%; l) NBS (1.0 equiv), CH3CN, 258C, 1 h, 98%; m) NaH
(1.2 equiv), MOMCl (1.5 equiv), DMF, 08C, 1.5 h, 99%; n) nBuLi
(1.6m in hexanes, 1.2 equiv), THF, ꢀ788C, 20 min; DMF (4.0 equiv),
THF, ꢀ788C, 1.5 h, 96%. DBU=1,8-diazabicyclo[5.4.0]undec-7-ene,
DMF=N,N-dimethylformamide, NBS=N-bromosuccinamide,
PDBBA=potassium diisobutyl-tert-butoxyaluminum hydride,
KHMDS=potassium bis(trimethylsilyl)amide, MOM=methoxymethyl,
THF=tetrahydrofuran.
Angew. Chem. Int. Ed. 2010, 49, 5146 –5150
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5147