Z. Fu et al. / Bioorganic Chemistry 59 (2015) 31–38
35
3.3. Effects of reaction conditions on Cu(II)(Im)4-catalyzed
autoxidation of 1 and 2
3. Results and discussion
3.1. Oxidation of alkylphenols by copper(II) bathocuproine disulfonate
complex. Are free phenoxy radicals susceptible to oxygenation?
As demonstrated in Table 3, a series of experiments on Cu(II)Im4
catalyzed autoxidation of 1 and 2 were conducted to examine the
effect of reaction conditions such as solvent, molar ratio of phenol/
Cu(II) and the concentration of O2 on the relative reaction rate
(conversion of starting phenols) and product distribution (oxygen-
ation vs. coupling). No quinone products (6 and 7) were detected
when Cu(II)Im4-catalyzed autoxidation of 1 and 2 was conducted
in pH 11 phosphate buffer, indicating that the active oxygenation
catalyst could not form in phosphate buffer. In the autoxidation
of 1 in alkaline aqueous MeOH, the use of stoichiometric amount
of Cu(II)(Im)4 increased the conversion of 1, but resulted in more
coupling than oxygenation, suggesting that the formation of active
oxygenation catalyst also requires excess phenol 1. The fact that
the conversion of phenol 1 in alkaline aqueous MeOH (9%) is much
higher than in phosphate buffer (ca. 1%) suggests that the active
oxygenation catalyst not only is able to promote the oxygenation
as opposed to coupling, but can also increase the overall conver-
sion. In the case of 2 which has an oxidation potential 0.1 V lower
than that of 1 [32], autoxidation proceeded smoothly in air in both
solvents, but only resulting in oxidative coupling. In fact, only
traces of 6 and 7 (corresponding to oxygenations on the tert-butyl
substituted and the unsubstituted o-positions of the phenol OH in
2) were detected in alkaline aqueous MeOH under 30 psi O2.
Theoretically, if a phenol has an oxidation potential lower
enough to be oxidized by Cu(II), one would expect trapping of
the resulting phenoxy radical by O2 and leading to oxygenation.
The fact that coupling reaction was observed as the major chemis-
try in most of the reported Cu(II)-mediated autoxidation of phenols
suggests that the phenoxy radical is more susceptible to coupling
than to oxygenation. To further diagnose the reactivity of free
phenoxy radicals toward O2, we initially examined the oxidation
of 1 and 2 by Cu(II)–bathocuproine disulfonate complex, a well-
known strong one electron oxidant (Table 1). Both reactions pro-
ceeded very quickly, with dimer 3 and trimer 4 being detected as
the major products in the case of 1 (entries 1–3), whereas dimer
5 was the only product in the case of 2 (entries 4–6). Similar results
were obtained under aerobic and anaerobic reactions, indicating
that oxidations of phenols 1 and 2 are completely O2 independent.
This is consistent with the fact that bathocuproine can lock the
copper species at Cu(I) state, thereby excluding its turnover to
Cu(II) by O2. Careful analysis of the concentrated organic extracts
did not show any oxygenation products. No oxygenation products
were also detected when the same reactions were conducted under
high dilution condition high pressure (30 psi) O2 (entries 3 and 6).
This confirms the low reactivity of phenoxy radicals toward oxy-
genation as opposed to coupling. These results clearly indicate that
free phenoxy radicals are relatively inert to O2 [30].
3.4. Effects of other reaction conditions on Cu(II)Im4 catalyzed phenol
autoxidation
As shown in Table 3, Cu(II)(Im)4 mediated autoxidation of 1 was
studied in details in 20% of pH 11 aqueous MeOH under 30 psi O2.
TLC analysis of the organic extracts indicated that, besides the
hydroxyquinone 6, the major side product was dimer 3, while only
traces of other oligomers were detected. As shown in Figs. 3 and 4,
profiles of time dependant consumption of 1 (left, Fig. 3), and for-
mation of 6 and 3 (right, Fig. 3; left, Fig. 4, respectively), and time
dependant deterioration of 6 (right, Fig. 4) under the same condi-
tions were obtained. The yield of 6 increased steadily and reached
maximum (4.2%) in 5 days and then decreased slowly. The conver-
sion of 1 and formation of 3 were kinetically parallel, and both slo-
wed down after 6 days. Control experiments did not show any
interchange between 6 and 3, suggesting that the slow decrease
of 6 after 5 days is due simply to its slow decomposition. Thus,
the theoretical yield of 6 should be slightly higher than observed.
The fact that conversion of 1 is less than 50% after 10 days suggests
that the cease of reaction is not due to exhaustion of starting 1, but
is due to the death of the solid active catalyst. Indeed, we found
that after 10 days the brown solid catalyst did not show any cata-
lytic activity on oxygenation of 1, but could still catalyze a rather
slow oxidative coupling of 1 to give dimer 3, suggesting that the
solid catalyst is more unique for oxygenation than for oxidative
coupling. In fact, slow oxidative coupling was observed in the pres-
ence of Cu(II) even without any ligand (not shown). Finally, an
examination of the reaction profile within 6 days reveals a total
material recovery of about 80%, indicating that other unknown
polar products were also formed.
3.2. Cu(II)-catalyzed autoxidation of 1 in the presence of different
ligands. What is the best ligand?
Rinaldi and coworkers chose imidazole as a ligand to promote
Cu(II)-catalyzed oxygenation of 1. The best yield of 6 was achieved
using four equivalents of imidazole with respect to copper(II) in 20%
aqueous MeOH at pH 11 [25]. Since redox property of Cu(II) can be
tuned using different ligands, which might offer better catalytic
performance for phenol oxygenation [8,17,18,25], we firstly ran
Rinaldi’s reaction [25] using a variety of ligands (Table 2). All efforts
failed to improve the yield of oxygenation product (quinone) 6.
Clearly, imidazole remains the best ligand, and significantly higher
yields of 6 can be achieved using this ligand under high pressure O2.
The imidazole containing chelating ligand cyclodihistidine (CDH)
showed comparable catalytic activity as imidazole, however, other
chelating ligands all failed to promote phenol oxygenation.
The nature of active catalyst using imidazole was then studied.
When a clear dark blue solution of Cu(II)Im4, and 1 in 20% of aque-
ous MeOH was titrated to pH 11, the mixture turned to a bluish-
green foaming suspension. After stirring in air for 1 day, the insol-
uble substance changed to a brown precipitate, and the yield of 6
was measured as 0.47%. At this point, the solid materials and the
solution were separated by centrifuge. The supernatant was fur-
ther stirred in air for 48 h. The brown solid was added to the same
volume and same concentration of a freshly prepared solution of 1
in 20% of pH 11 aqueous MeOH, and the resulting suspension was
also stirred in air for 48 h. The solid material afforded much more
product 6 (0.50%) than did the supernatant (only increased by
0.04%), suggesting that the brown solid is the active catalyst.
Although the structure of this catalyst is unclear, the ability of
imidazole to promote Cu(II)-catalyzed phenol oxygenation appears
to be associated with the fact that its two meta nitrogens both can
coordinate with Cu(II) [31], thereby permitting reorganization and
polymerization of Cu(II)Im4 into active catalyst in alkaline
medium.
3.5. Mechanistic discussion
Since the structure of Rinaldi’s active oxygenation catalyst
evolving from a complicated heterogeneous reaction system is
far reaching [25], the detailed mechanism for the copper version
of the reaction is unclear in our manuscript. However, put together
the results in our manuscript, we propose possible pathways for
the formation of oxygenation and coupling products (Fig. 5). The