T.-H. Chen et al. / Applied Catalysis A: General 497 (2015) 121–126
123
1,2,4-trichlorobezene (0.1 mmol) was prepared (final vol-
ume = 0.5 mL). The internal standard was shown to be stable
to the oxidation conditions in control reactions. PhI(OAc)2
(0.2 mmol) as limiting reagent was added, and the mixture was
stirred at ambient temperature (23 2 ◦C) for 10 min. Relative
rate ratios for catalytic oxidations were determined based on the
amounts of substrates by GC (FID) as measured against an internal
with good epoxide yields (> 80%) and mass balance (> 95%), and in
all cases no traces of polymers or oligomers were detected. Thus,
the rate of alkene disappearance should reasonably reflect the
alkene reactivity toward the corrole-iron-catalyzed epoxidation.
The values reported in Table 3 are the averages of 2–3 runs with
standard deviations (1).
primarily epoxide with negligible amounts of allylic oxidation
products (entry 2). Remarkably, FeIII(TPFC) performs as well
the epoxidation of cyclohexene under similar reaction conditions
[27]. This catalytic activity (TOF) and product selectivity is a
major improvement over previously reported metallocorrole
catalysts [13,16] including perhalogenated MnIII(F8TPFC) [20] and
MnIII(Br8TPFC) [19]. Epoxidation of cis- and trans-stilbenes afforded
corresponding expoxides exclusively with complete stereoreten-
tion (entries 3 and 4). In the epoxidation of styrene and substituted
styrene, moderate conversions were observed albeit with small
amounts of aldehyde products (entries 6 and 7). Similarly, the
oxidation of secondary benzylic alcohols gave the corresponding
ketones with moderate catalytic activities (entries 8 and 9).
Activated alkanes including ethylbenzene and diphenylmethane
were oxidized to the corresponding alcohols and/or ketones from
over-oxidation with lowest activity (entries 10 and 11). It is note-
worthy that monitoring catalytic reactions by UV–vis spectroscopy
indicated no significant catalyst bleaching in the end of reactions
(see Fig. S1 in Supporting Information). Thus, in comparison to
other oxidants such as PhIO, the corrole catalyst stability against
degradation was much enhanced owing to the mild oxidizing
ability of PhI(OAc)2.
3. Results and discussion
3.1. Screening studies
The potential of PhI(OAc)2 as an oxygen source was first
evaluated in the catalytic epoxidation of cis-cyclooctene by the
electron-deficient iron(III) 5,10,15- tris(pentafluorophenyl) cor-
role, i.e. FeIII(TPFC).[32] Under mild homogeneous conditions, the
epoxidations were carried out with a catalyst: substrate: PhI(OAc)2
ratio of 1:200:220 (Table 1). After 1 h of reaction in CH3CN, cis-
cyclooctene oxide was obtained as the only identifiable oxidation
product (> 99% by GC) with ca. 33% conversion (Table 1, entry
1). Gratifyingly, the same reaction proceeded much more rapidly
with a small amount of H2O (5 l), and thus, 100% conversions
(entry 3). Fig. 1 depicts the time courses for the epoxidation in the
presence and absence of water. Similar water accelerating effect
observed in the reported iron(III) porphyrin-catalyzed oxidations
was rationalized in terms of the formation of more oxidizing PhIO
[27]. Remarkably, the catalyst loading can be as low as of 0.05 mol%
ocally the high efficiency (1400 TON). Although PhIO is a common
oxygen source generally used in metal-catalyzed oxidations, it was
found that the use of PhIO under the same conditions led to a lower
catalytic activity (entry 5 and Fig. 1), presumably due to its poor
solubility in CH3CN and/or causing more catalyst bleaching. More-
over, no accelerating effect of water was observed in the catalytic
oxidation by PhIO in the presence of acetic acid (data not shown).
The use of CH3OH or CH2Cl2 as solvent instead of CH3CN resulted
in reduced activity (entries 6 and 7). Quite interestingly, the oxi-
dation state on the metal gave a minor effect, and the FeIV(TPFC)Cl
(entry 8) and [FeIV(TPFC)]2O (entry 9) gave a slightly reduced cat-
alytic activities compared to FeIII(TPFC). Catalyst degradation was
a problem when the non-halogenated FeIII(TPC) (TPC = 5, 10,15-
triphenylcorrole) was used as catalyst, showing a sluggish catalytic
activity (entry 10). Control experiments showed that no epoxide
was formed in the absence of either the catalyst or the PhI(OAc)2
even at elevated temperature (50 ◦C) or in the presence of acetic
acid.
To show the synthetic utility of the method, the epoxidation of
cis-stilbene was scaled up to 2.0 mmol and to our delight, a simi-
lar result was obtained in 100% conversion and 95% isolated yield
exclusively for cis-stilbene oxide.
3.3. Catalytic competition studies
Prior to the present study, the use of metallcorroles for catalytic
oxidations has met with limited success in view of the poor selectiv-
ity, low efficiency and, in most cases, inherent catalyst degradation.
The synthetic value of the FeIII(TPFC)/PhI(OAc)2 system presented
above are indisputable, and the observed high catalytic activity
strongly implicates a high-valent corrole-iron-oxo species as the
for a high-valent corralizine-iron-oxo species as Compound I heme
analogues [24,25]. In fact, we reported LFP generation and kinetic
best described as iron(V)-oxo species [29,30]. To evaluate the
identity of the active oxidant during the catalytic conditions, the
competition studies with FeIII(TPFC) and PhI(OAc)2 were conducted
as described in Table 3. Evidently, the results of the competition
reactions between cyclohexene and cis-cyclooctene and between
ethylbenzene and ethlybenzene-d10 are in good agreement with
the ratios of the absolute rate constants found in direct kinetic
d10 revealed a kinetic isotope effect (KIE) of kH/kD = 4.40 0.21
at 298 K, similar to the KIE reported for the same reaction with
(model Compound I).[33] The observed KIE is larger than those
observed in autooxidation processes (typical KIE = 1–2),[34] sup-
porting a non-radical mechanism.
3.2. Substrate scope and scale
Consequently, the catalytic oxidations of a variety of organic
substrates were investigated under optimized conditions. Table 2
lists the oxidized products and corresponding substrate con-
versions and product yields including isolated yields using the
FeIII(TPFC) as catalyst. As evident in Table 2, in many cases, quan-
titative conversions, excellent selectivity and rapid turnovers
3.4. Hammett correlation studies
A further reflection of the high reactivity of active intermediate
involved is seen in the linear Hammett plot for competitive oxida-
tions of the series of substituted styrenes (Y-styrene, Y = 4-MeO,
4-Me, 4-F, 4-Cl, and 3-NO2). Fig. 2 depicts a linear correla-
tion (R = 0.996) of log krel [krel = k(Y-styrene)/k(styrene)] versus
(up to 20 TOF min−1
dation of cyclohexene was completed within 10 min, giving
) were observed. For example, epoxi-