Novel effect of palladium catalysts on chemoselective oxidation of b-pinene by hydrogen peroxide
325
only a low conversion was obtained (ca. 8 %, Table 2).
Pd(acac)2/OAc-
Nevertheless, a greater conversion of b-pinene into oxi-
dation products was achieved when -OAc ions were added
to the solution (ca. 79 %, run 6, Table 3). The highest
selectivity for pinocarveol (1a) was achieved in the pres-
ence of the Pd(acac)2/-OAc catalyst. Consequently, it can
be concluded that in the reactions studied, acetate ions
efficiently replaced the acac ligands, thus resulting in a
more active catalyst.
Pd(acac)2/Cl-
Pd(TFAc)2/OAc-
Pd(TFAc)2/Cl-
80
60
40
20
0
An interesting aspect is the rate of anionic ligand
exchange reaction. Monitoring of substrate conversion may
be indicative of the ligand substitution rate throughout the
reaction. The kinetic curves shown in Fig. 6 reveal that
addition of chloride or acetate ions to Pd(TFAc)2 or Pd
(acac)2-catalyzed oxidation reactions drastically change the
b-pinene oxidation rates.
0
2
4
6
8
Time/h
A comparison between reaction rates measured in the
absence of chloride or acetate ions (Fig. 3) and those mea-
sured in their presence indicates that in all catalytic runs
these additional ions reduce the reaction rate (Fig. 6). In the
absence of these ions, Pd(TFAc)2 or Pd(acac)2-catalyzed
reactions reach maximum conversion after 2 h of reaction.
However, there is a positive and relevant aspect; although it
requires a longer reaction time, when compared to
Pd(OAc)2-catalyzed reactions, the Pd(acac)2/-OAc catalyst
was shown to be highly effective for promoting b-pinene
oxidation into pinocarveol. Thus, Pd(acac)2/-OAc can be
also used as a catalyst instead of Pd(OAc)2.
Fig. 6 Kinetic curves of Pd(TFAc)2 and Pd(acac)2-catalyzed oxida-
tion reactions of b-pinene by hydrogen peroxide in the presence of
–OAc and Cl- ions (reaction conditions: 0.15 mmol catalyst,
3.75 mmol b-pinene, 4.5 mmol H2O2, 15 cm3 CH3OH, 8 h, 55 °C)
complete conversion of b-pinene into allylic products has
not yet been reported.
Therefore, it was concluded that with appropriate
selection of the experimental conditions and the adequate
palladium catalyst, the chemoselectivity of allylic oxida-
tion reaction can be controlled, reaching high conversion
and selectivity for two attractive synthetic routes: ether and
alcohol synthesis from natural olefins.
Conclusions
Experimental
In this work, a significant advance was obtained in the
development of metal reoxidant-free protocols, which are
based on Pd(II)/H2O2 and applied for oxidation of mono-
terpenes. Use of CH3OH as a solvent provokes a drastic
reduction in reaction time (ca. 8–2 h) compared to the
other system described (i.e., PdCl2/H2O2/CH3CN system),
without compromising oxidation selectivity (as in the
PdCl2/H2O2/HOAc system). In general, b-pinene was
almost completely converted into allylic products by H2O2
in the PdCl2 or Pd(OAc)2-catalyzed reactions. Reaction
chemoselectivity was strongly dependent on the anionic
ligands on the palladium catalyst. When using the
Pd(TFAc)2 or PdCl2 catalyst, a rare allyl ether (namely
myrtenol methyl ether) was the major product; conversely,
when the reactions were catalyzed by Pd(acac)2 or
Pd(OAc)2, b-pinene was efficiently oxidized into pinocar-
veol. The effect of adding acetate and chloride ions to the
less effective catalysts was also assessed. In the presence of
acetate ions, the less efficient catalyst Pd(acac)2 becomes
notably more effective, selectively promoting b-pinene
oxidation into pinocarveol. To our knowledge, a virtually
All chemicals were purchased from commercial sources.
Palladium chloride (99 % w/w), palladium trifluoroacetate
(99 % w/w), palladium acetate (99 % w/w), and palladium
acetylacetonate (99 % w/w) were acquired from Sigma-
Aldrich. (-)-b-Pinene (99 % w/w) was also purchased
from Sigma-Aldrich. Methanol (Merck, HPLC grade) was
used as received without prior treatment. All others sol-
vents (dichloromethane, hexane) were acquired from a
commercial source (Nuclear, Brazil) and used as received.
The H2O2 aqueous solution (ca. 34 % w/w, Vetec, Brazil)
was the oxidant employed in all reactions, and its con-
centration was determined by titration against a KMnO4
solution.
Catalytic tests and reaction progress monitoring
The reactions were carried out in a glass reactor (50 cm3)
equipped with a reflux condenser under atmospheric pressure.
Typically, the appropriate palladium catalyst (0.15 mmol)
and aqueous hydrogen peroxide (3.75–11.3 mmol) were
123