C
A. Narayanappa et al.
Letter
Synlett
explore both the directing-group effect and also provide
useful methodology for isoprenoid homologation process-
es. The protected ketals 27–30 (Table 2) were readily ob-
tained following standard chemistry, however, compounds
27 and 28 proved hydrolytically unstable. In contrast, both
the neopentyl ketal 29 and acetone-pinacol derivative 30
proved to be chemically robust. The desired ketal-function-
alized phosphonium salt 1e was now readily accessed from
30 through quaternization with tripropylphosphine (Table
2).
O
Br
Br
23
O
O
CF3CO2H/CH2Cl2
r.t., 24 h
Br
O
9
24
OH
AD-mix-b
O
O
O
water/tBuOH
0 °C to r.t., 24 h
OH
O
11
25
Regioselective ylide formation using the salt 1e was
now investigated with 4-chlorobenzaldehyde (8b) under a
range of conditions (Table 3) in order to probe selectivity in
favor of the functionalized olefin 31 over the butene 32.
Considering that the ‘statistical’ distribution of olefins
should at most comprise 25% of 31 (lower than this figure
taking steric parameters into account) and >75% of the bu-
tene 32, the initial results proved disappointing (Table 3).
Olefin 31 was obtained in statistical mixtures (Table 3, en-
tries 1–10) with 32 or with slight selectivity (e.g., 40:60),
indicating only a 1.6 selectivity in favor of 31. This reaction
was investigated with various solvents and coordinating
cations with little improvement, leading us to consider that
salt 1e may now exceed the limit of the acetal/ketal-direct-
ed pathway. Finally, we investigated the reaction in the
noncoordinating solvent diethyl ether and in the presence
of lithium salts (Table 3, entry 11) resulting in a substantial
change in the olefin ratio. The desired functionalized olefin
31 was now obtained over 32 (57:43), demonstrating a syn-
thetically useful selectivity of 2.28 in the desired direction.
While solubility of the phosphonium salt 1e is an issue, this
positive result is a strong impetus to improve the regiose-
lective ylide formation further employing various coordi-
nating cations in diethyl ether and other noncoordinating
media.
Scheme 2 Deprotection/isomerization of the three-carbon homo-
logated intermediate 9 yielding the carbonyl-conjugated enal 24 (90%)
and dihydroxylation of 11 yielding the protected diol 25
The synthetic potential of the three-carbon homologat-
ed adducts summarized in Figure 1 has not been fully ex-
plored. Nonetheless, deprotection of the 4-bromophenyl
acetal derivative 9 was investigated under a range of condi-
tions, for example, use of PTSA in DCM gave a 40:60 mix-
ture of 23/24, whereas use of TFA (Scheme 2) gave the
deprotected/isomerized alkenal 24 cleanly in 90% yield.7 In-
termediates such as 24 appear ideal as substrates for di-
enamine-directed organocatalysis.9 Sharpless asymmetric
dihydroxylation10 on the protected olefin 11 give the pro-
tected diol 25,7 an intermediate that appears suited towards
the synthesis of naturally occurring arenediols.11
Table 2 Synthesis and Stability of 4-Bromo δ-Cyclic Acetals 27–30 and
Conversion of 30 into the Phosphonium Salt 1e
R1 R1
R2
33% HBr-HOAc (5 mol%)
CH(OCH3)3 (1.7 equiv)
O
R2
O
27, n = 0, R1 = R2 = H
28, n = 1, R1 = R2 = H
n
O
diol 4.0 equiv., r.t.
0.5 to 3 h
Br
Br
29, n = 1, R1 = H, R2 = CH3
30, n = 0, R1 = R2 = CH3
26
As in the case of salt 1d, the ketal present in 1e is posi-
tioned too far from the α-site of deprotonation and now
rules out inductive effects being involved in the selectivity
that favors functionalized olefin 31. The high degree of re-
gioselectivity can only now be explained in terms of a com-
plex-induced proximity effect8 of the acetal, directing ylide
formation. Such effect may be manifest thermodynamically,
though reversible deprotonation8b at the desired α-position.
Nonetheless, in consideration of the reaction conditions de-
scribed here (Table 1, entry 8; Table 3, entry 11) the results
are more consistent with kinetic deprotonation via pre-
complexation to the acetal/ketal and deprotonation at the
now closer α-methylene.8d No significant difference is ob-
served in regioselectivity as a function of time elapsed be-
fore addition of the aldehyde after ylide formation (Table 3,
entries 8 and 9), under otherwise identical conditions (THF,
lithium salts). Conversely, switching solvent to diethyl ether
and use of the strong base butyl lithium (Table 3, entry 11)
results in a significant shift in favor of the chelation-assist-
ed pathway.
Pr3P
1e
Entry
Ketal
Time (h)
Yield (%)a
26 after 24 h (%)b
1
2
3
4
27
28
29
30
0.5
3
74
33c
89
82
5
20
2
1.5
3
<1
a 1.0 mmol scale.
b Hydrolysis products determined after 24 h in CDCl3 by relative integration
of 1H NMR signals arising from methyl protons.
c Isolated as mixture of 26 and 28.
Given the success of the acetal-directed ylide formation
demonstrated using reagent 1d, we next desired to probe
the scope of regioselective ylide generation by pushing the
acetal protecting/directing group still one methylene fur-
ther. In terms of reagent design, we considered that ketals
derived from 5-bromo-2-pentanone (26) would be ideal to
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 28, A–E