4
F. Bronston et al. / Polyhedron xxx (2016) xxx–xxx
the Ga(III) center in order to facilitate the oxygen atom transfer to
the alkene [20]. Anionic ligands are generally more effective at
competing for vacant coordination sites on cationic metal centers
[25], and the deprotonation of PA to peracetate should facilitate
the binding of the terminal oxidant to the Ga(III). The heightened
activity at acidic pH values, conversely, may be a consequence of
keeping the acetic acid (pKa = 4.8) administered with the PA in its
protonated form, which would have a lesser affinity for Ga(III).
Between pH 6 and pH 8, the acetic acid would be predominantly
acetate, which would out-compete PA for the coordination sites
on the Ga(III) center.
(Table 2). PhIO and oxone were completely ineffective in pure
MeCN due to their poor solubility. Switching to alternative solvent
systems did not substantially improve the activity, and poor yields
of epoxides are observed in both the presence and absence of 2.
The reaction between cyclooctene and MCPBA, conversely,
proceeds quantitatively without a catalyst. The reaction between
1-octene and MCPBA only proceeds to ꢁ15% without a catalyst,
but the presence of 2 did not substantially improve this number.
Since the addition of 2 failed to improve the reactivity in all cases,
most of these terminal oxidants were not tried in reactions in H2O.
H2O2 was tested as
a terminal oxidant for Ga(III)-catalyzed
cyclooctene epoxidation in pH 4.0 H2O; as with the MeCN system,
no epoxide was observed.
3.3. Functional group tolerance
Previous attempts to analyze the mechanism of Ga(III)-
catalyzed olefin epoxidation were largely limited to a computa-
tional study which predicted the agency of a Ga(III)-peracetate
complex [20]. Although the Ga(III)-peracetate oxidant was calcu-
lated to be relatively stable, we were unable to observe this species
experimentally. The large excesses of acetic acid present in all
grades of PA would preclude the isolation of clean samples of per-
acetate intermediates. MCPBA is an attractive terminal oxidant for
mechanistic studies since it is available in a form that is compara-
tively pure. The perbenzoic acid is the major component (ꢁ75%) in
commercially available supplies of MCPBA, whereas PA is a minor
component in both commercially available and custom-prepared
solutions of this oxidant.
Regrettably, we were unable to observe a meta-chloroperben-
zoate analog of the calculated Ga(III)-peracetate oxidant.
Substrate-free reactions between MCPBA and 2 were monitored
by 1H NMR. The observed resonances correspond to the starting
materials, and no evidence for a meta-chloroperbenzoate-Ga(III)
adduct was observed. This result suggests that the failure of
Ga(III) to improve the oxidation of 1-octene by MCPBA (Table 2)
may be a consequence of the inability of this peracid to coordinate
to the metal center. We speculate that MCPBA is too bulky to
coordinate to the Ga(III)-bispicen complex. The noted inability of
H2O2 to coordinate to Ga(III) may partly explain why this molecule
is also ineffective as a terminal oxidant [26].
Transition metal catalysts for olefin epoxidation often enable
reactions with other functional groups. The functional group
tolerance of reactions using the Ga(III)-containing catalysts had
not yet been fully explored. Prior work suggested that the mixtures
of Ga(III) complexes and PA are not capable of rapidly opening
epoxides since the epoxides are the only observed organic products
for the first 5 h [18].
Various derivatized cyclohexanes were tested as substrates in
MeCN using 2 as the catalyst. The initial concentrations of 2, PAR,
and substrate were 5.0 mM, 500 mM and 500 mM, respectively.
Cyclohexanol, cyclohexanone, and chlorocyclohexane each remain
unchanged. When the reactions are provided the standard 60 min,
no organic compounds aside from unreacted starting material are
observed. 1,2-Diaminocyclohexane was also explored as a sub-
strate, but we were unable to isolate it and its putative products
cleanly from the reaction mixtures. Cyclohexanecarboxyaldehyde
reacts with PA to yield a complex mixture of organic products;
the addition of 2 does not noticeably alter this reactivity. From
these studies, we can predict that alkene substrates with amine
and/or aldehyde groups will not react cleanly with mixtures of
PA and Ga(III) complexes.
3.4. Alternative terminal oxidants
Even if the [GaN4]3+ cores of the catalysts were to be further sta-
bilized, the instability of PA in H2O would still limit the reaction
times to 1 h [23]. Finding an alternative terminal oxidant to PA is
therefore essential for obtaining further advances in aqueous
gallium-catalyzed olefin epoxidation. Heretofore, only O2, H2O2,
and two grades of PA were tried as terminal oxidants for
Ga(III)-catalyzed olefin epoxidation [17,18]. Other common oxygen
transfer agents include iodosobenzene (PhIO), oxone, and meta-
chloroperbenzoic acid (MCPBA). These terminal oxidants were
tested in primarily organic solvents using complex 2 as the catalyst
4. Conclusions
Two previously discovered Ga(III) complexes with N-donor
ligands were found to be capable catalysts for olefin epoxidation
in water. The activity was better with the more highly chelating
but more electron-rich bispicen ligand, suggesting that the stabil-
ity of the catalyst is more important than its electronic character
in aqueous olefin epoxidation. The activity is poor around neutral
pH but improves substantially under both basic and acidic condi-
tions. Although a number of functional groups survive the reaction
conditions in MeCN, aldehydes react, and the stability of amines
could not be readily assessed. Lastly, we did not locate a common
terminal oxidant that could benefit from the presence of Ga(III) and
substitute for PA. We hypothesize that these other terminal
oxidants may be unable to coordinate to the Ga(III) center.
Table 2
Yields of cyclooctene and 1-octene epoxidation by other terminal oxidants with and
without 2.
Terminal oxidant
Cyclooctene (%)
1-Octene (%)
O2
0/0
0/0
0/0
0/0
0/0
0/0a
Acknowledgement
H2O2
PhIOb
Oxonec
MCPBA
0/0
11/8
100/100
The described work was financially supported by Auburn
University.
17/16
Reaction
conditions:
MeCN,
air,
298 K,
[Ga(III)]o = 1.0 mM/
0.0 mM,
[alkene]o = 100 mM, [terminal oxidant]o = 200 mM. The yields were measured at
30 min by GC. All measurements are the averages of the yields of at least three
separate experiments. The yields of the Ga(III)-catalyzed reactions are listed first.
References
a
Reaction also attempted in H2O, yielding identical results.
Reactions with both alkenes occurred in MeOH.
Reaction performed in 3:1 H2O/MeCN. [Ga(III)]o = 0.5 mM, [alkene]o = 50 mM,
b
c
[oxone]o = 100 mM.