Diastereoselective IBX Oxidative Dearomatization of Phenols
COMMUNICATION
tries 1 and 11); 10b-TFA ester was also isolated as a single
diastereomer.
cause the greater the number of equivalents of TFA, the
faster the deprotection occurs (Figure 2). These kinetics are
correlated with the observed diastereoselectivity of 10a
(Table 1, compare entries 1, 2, and 7) and we propose that
the oxidative dearomatization of either monophenol 12a or
17a is diastereoselective, whereas the oxidation of 9a is not.
Furthermore, we have shown that adding water dramatically
accelerated the reaction. In this case, the kinetics of oxida-
tion (kox) increase but do not affect the kinetics of deprotec-
tion significantly (kdep2 and k’dep2). Therefore, in the pres-
ence of water, the oxidation of 12a and 17a takes place
before the second deprotection and results in high diastereo-
selectivity. The water–TFA combination presumably depoly-
merizes IBX and allows for a more favorable liquid–liquid
interaction instead of a solid–liquid interaction, with IBX
being poorly soluble in dichloromethane. In the case of 14a,
the observed diastereoselectivity was lower than with 8a
(Table 1, compare entries 6 and 9). This is due to the faster
deprotection of 17a, compared with 12a (k’dep2>kdep2),
which increases the amount of 9a, for which the oxidation is
not diastereoselective.
The oxidative dearomatization of in situ generated 9a or
9b was performed after TFA treatment of 8a or 8b and
overnight stirring (for complete deprotection of both MOM
groups), before addition of IBX. In these cases, 10a was ob-
tained with no diastereoselectivity (Table 1, entry 3) whereas
10b was obtained with a moderate 61/39 d.r. (Table 1,
entry 13). These two results show the lack of or low diaste-
reoselectivity of the oxidation of free diphenols 9a or 9b;
they also show that iPr substitution induces a better, al-
though to a moderate extent, diastereoselectivity. These re-
sults are also consistent with the d.r. observed in the pres-
ence of 20 equiv of TFA (Table 1, entries 1 and 11), in which
free diphenols are formed very quickly (vide infra,
Figure 2).
In the case of 16a, NMR studies in the presence of TFA
only showed that the benzyl protecting group tolerated
acidic conditions (since only the MOM at position 8 was
cleaved). Accordingly, a high diastereoselectivity was ex-
pected but, to our surprise, low diastereoselectivity was ob-
served. This can be explained by the IBX-mediated depro-
tection and oxidation of the benzyl group,[12] generating di-
phenol 9a, in which oxidative dearomatization is not diaste-
reoselective.
To summarize, the obtained diastereoselectivity of 10 is a
combination of the highly diastereoselective oxidation of
12a, 12b, or 17a, counterbalanced by the non- or low-diaste-
reoselective oxidation of 9a or 9b.
^
Figure 2. Kinetics of formation of 9a from 8a in various conditions.
=
~
&
TFA (7 equiv); =TFA (7 equiv)+H2O (2 equiv); =TFA (20 equiv).
Importantly, the formation of TFA esters do not arise
from esterification of the corresponding alcohols in the reac-
tion medium but certainly during dearomatization by TFA
transfer from TFA–IBX esters.[10] Moderate yields may be
explained by the high sensitivity of 10 on silica and aqueous
conditions, due to Michael addition of water on the enone
moiety. The use of SIBX instead of IBX did not allow us to
increase the yield.[11]
We can then postulate that diastereoselective oxidation
takes place when IBX coordinates to the phenol in position
8 or 6 (the other phenol still being protected) and delivers
oxygen on one side preferentially, anti to the methyl of the
dihydropyran ring (Scheme 4). Indeed, ligand exchange be-
The sequential, but different,
deprotection of protecting
groups on lactols 8a and 14a
has been observed (Scheme 3).
Treatment of either 8a or 14a
with TFA very quickly (mi-
nutes) yielded the oxonium ion
12a (MOM cleavage at posi-
tion 8), or 17a (TBS cleavage
at position 6), respectively.
Then, 12a slowly evolved to di-
phenol 9a (kdep2), whereas 17a
more rapidly evolved to diphe-
nol 9a (k’dep2). The kinetics of
these second deprotections
(kdep2 and k’dep2) were shown
to be [TFA] dependent be-
Scheme 4. Proposed mechanism for diastereoselectivity.
Chem. Eur. J. 2011, 17, 10241 – 10245
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10243