Oxygen-Atom Transfer
COMMUNICATION
cation MnO+,[14] complex 1 does not react at an appreciable
rate with ethylene under ICR conditions (kexp
<
10ꢀ12 cm3 moleculeꢀ1 sꢀ1). Undoubtedly, the presence of the
porphyrin ligand in 1 moderates its reactivity and the onset
of the olefin epoxidation is only reached with propene. Also
the electron demand by the porphyrin macrocycle clearly
plays a role in the gas phase as well as in solution.[8,15,16] In
fact the OAT reactivity towards 2-methoxypropene is about
twenty times higher for 1 than with [(TPP)MnVO]+, devoid
of fluorine atoms on the phenyl rings (see footnote [f] in
Table 1).
The kinetic data summarised in Table 1 display a wide
range of efficiencies, which span from 0.02% for a terminal
olefin like propene, to 100% for electron-rich monoter-
penes. The OAT reactivity increases as the ionisation energy
(IE) of the olefin decreases. In the case of indene, the OAT
process is accompanied by an electron-transfer reaction,
which sets a lower limit of 8.1 eV for the IE value of gas-
eous [(TPFPP)MnIVO]. Another issue from the kinetic data
reported in Table 1 regards the reactivity of naked 1 towards
E/Z isomers. The performance of manganese complexes in
stereoselective epoxidations is of high synthetic value, al-
though the nature of the reactive intermediate is still under
debate. The gas-phase reaction of complex 1 with (Z)-2-
butene is ten times faster than the one with (E)-2-butene.
This inherent selectivity may provide a valuable reference
to account for the stereoelectronic factors affecting the in-
teraction between an olefin and a MnV–oxo complex.[17]
Facing the considerable OAT reactivity of 1 towards ole-
fins, the question arises whether the corresponding trans-di-
Figure 3. Time dependence of the relative ion abundances (A, %) follow-
ing the selection of ions at m/z 1043 in the presence of a gaseous mixture
of 2-F-pyridine (6.5ꢂ10ꢀ8 mbar) and b-pinene (5.0ꢂ10ꢀ8 mbar) in the FT-
ICR cell.
energy surfaces and of ensuing mixing has been proposed in
several oxidation processes, including Mn-catalysed epoxida-
tions.[3,13] In particular, recent detailed computational studies
show that three possible spin states (singlet, triplet and quin-
tet) are available to MnV–oxo–porphyrin complexes, sepa-
rated by relatively narrow energy gaps.[6,7]
The bimolecular rate constants (kexp) along with the rela-
tive efficiencies (F), both listed in Table 1, indicate that ion
1 bearing no axial ligand is indeed an active oxidant toward
olefins. Under the present experimental conditions, neither
hydrogen-atom-transfer processes nor addition of L to ion 1
are ever observed. In contrast to the bare metal oxide
oxomanganese(V) (MnVO2 ) species can be generated in
ꢀ
Table 1. Kinetic data for the reaction of [(TPFPP)MnVO]+ (1) with se-
the gas phase and tested for reaction with the same sub-
strates. Operating in negative ESI, complex 4 at m/z 1059,
formally corresponding to [(TPFPP)MnVO2]ꢀ, is observed.
When the ion is isolated in the FT-ICR cell in the presence
of selected neutrals, no reaction is found to occur [Eq. (3)].
Not only electron-rich olefins, but also compounds prone to
oxidation, such as sulfides, appear remarkably inert.
lected olefins in the gas-phase.[a]
[c,d]
Olefin (IE)[b]
kexp
F[e]
0.021
0.43
0.78
8.1
0.76
6.8
7.7
7.8
16
21
28
12
propene (9.73)
0.0020
0.041
0.072
0.75
0.070
0.63
0.75
0.69
1.7
3,3-dimethyl-1-butene (9.45)
(E)-2-butene (9.10)
(Z)-2-butene (9.11)
allylbenzene (7.8–8.7)
styrene (8.46)
cyclohexene (8.95)
½ðTPFPPÞMnVO2ꢁꢀ þ L ! unreactive
ð3Þ
1,4-cyclohexadiene (8.8)
1-propene, 2-methoxy (8.64)[f]
1,3,5-cycloheptatriene (8.30)
(+)-camphene (ꢃ8.86)
indene (8.14)[g]
1.9
2.6
1.2
In conclusion, the gas-phase ion chemistry of MnV–oxo–
porphyrin complexes has provided direct evidence that a
MnV–oxo–porphyrin cation such as [(TPFPP)MnVO]+ (1),
generated as a naked five-coordinate species, performs as an
efficient oxygen-atom donor to olefins. In contrast, the
dioxo species 4, [(TPFPP)MnVO2]ꢀ, a gas-phase counterpart
of the trans dioxo–MnV–porphyrins identified by Spiro,
Groves and their co-workers in the condensed phase,[5] has
proven unreactive.
(1S)-(ꢀ)-a-pinene (8.07)
9.0
9.3
9.2
100
100
100
b-pinene (n.a.)
(R)-(+)-limonene (8.3)
+
[a] In all cases, only the [(TPFPP)MnIII
]
product ion of the OAT chan-
nel is observed, unless stated otherwise. [b] Ionisation energies (IE [eV])
given in parentheses are from reference [19]; n.a. stands for not available.
[c] Second-order rate constants (kexp) in units of 10ꢀ10 cm3 moleculeꢀ1 sꢀ1
,
at the temperature of the FT-ICR cell of 300 K. The estimated error is
ꢂ30%; the internal consistency of the data is within ꢂ10%. [d] The fol-
lowing reactants failed to react with 1: CO; ethylene; cyclohexane; ben-
zene. [e] Reaction efficiency, F =kexp/kcoll ꢂ100. Collision rate constants
(kcoll) evaluated with the parameterised trajectory theory. [f] The reaction
of this olefin with [(TPP)MnVO]+ (TPP=5,10,15,20-tetraphenylporphina-
to dianion) is characterised by kexp =0.091ꢂ10ꢀ10 cm3 moleculeꢀ1 sꢀ1 and
F=0.85. [g] Electron transfer yielding L+· is also observed, accounting
for 40% of the product ions.
Experimental Section
The experiments were run on a Bruker BioApex 4.7 T Fourier transform
ion cyclotron resonance (FT-ICR) mass spectrometer equipped with an
Apollo I electrospray (ESI) ionisation source. Ions were driven into a
Chem. Eur. J. 2009, 15, 7863 – 7866
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7865