A diiron(III,IV) imido species very active in nitrene-transfer reactions

Article abstract of DOI:10.1002/anie.201307429

Metal-catalyzed nitrene transfer reactions arouse intense interest as clean and efficient procedures for amine synthesis. Efficient Rh- and Ru-based catalysts exist but Fe alternatives are actively pursued. However, reactive iron imido species can be very short-lived and getting evidence of their occurrence in efficient nitrene-transfer reactions is an important challenge. We recently reported that a diiron(III,II) complex is a very efficient nitrene-transfer catalyst to various substrates. We describe herein how, by combining desorption electrospray ionization mass spectrometry, quantitative chemical quench experiments, and DFT calculations, we obtained conclusive evidence for the occurrence of an {Fe^IIIFe^IV=NTosyl} intermediate that is very active in H-abstraction and nitrene-transfer reactions. DFT calculations revealed a strong radical character of the tosyl nitrogen atom in very low-lying electronic configurations of the Fe^IV ion which are likely to confer its high reactivity. Nitrene transfer: An Fe^IIIFe^IV imido intermediate is identified in nitrene-transfer reactions by desorption electrospray ionization mass spectrometry (DESI-MS). DFT calculations show that low-lying Fe^IIIFe^III-^.N-tosyl configurations play a major role in the high reactivity of the intermediate. Copyright

Full text of DOI:10.1002/anie.201307429

.
Angewandte
Communications
DOI: 10.1002/anie.201307429
Nitrene Transfer Very Important Paper
A Diiron(III,IV) Imido Species Very Active in Nitrene-Transfer
Reactions**
Eric Gourꢀ, Frꢀdꢀric Avenier, Patrick Dubourdeaux, Olivier Sꢀnꢁque, Florian Albrieux,
Colette Lebrun, Martin Clꢀmancey, Pascale Maldivi,* and Jean-Marc Latour*
Abstract: Metal-catalyzed nitrene transfer reactions arouse
intense interest as clean and efficient procedures for amine
synthesis. Efficient Rh- and Ru-based catalysts exist but Fe
alternatives are actively pursued. However, reactive iron imido
species can be very short-lived and getting evidence of their
occurrence in efficient nitrene-transfer reactions is an impor-
tant challenge. We recently reported that a diiron(III,II)
complex is a very efficient nitrene-transfer catalyst to various
substrates. We describe herein how, by combining desorption
electrospray ionization mass spectrometry, quantitative chem-
ical quench experiments, and DFT calculations, we obtained
proven so far too reactive to be isolated except in a few
particular instances.[7,9,10] Metal imido species can indeed be
very reactive and extremely short-lived (sub-microsecond)
making their probing by usual techniques difficult and their
isolation impossible. Thus getting evidence of their occur-
rence in efficient nitrene-transfer reactions is currently an
important challenge.
We have reported that the diiron(III,II) complex
1 (Scheme 1) was very efficient in catalytic nitrene transfers
to alkenes and thioanisole.[12] In addition, its unprotected
analogue (with hydrogen atoms in place of chlorine on the
conclusive evidence for the occurrence of an {FeIIIFeIV
=
NTosyl} intermediate that is very active in H-abstraction and
nitrene-transfer reactions. DFT calculations revealed a strong
radical character of the tosyl nitrogen atom in very low-lying
electronic configurations of the FeIV ion which are likely to
confer its high reactivity.
M
etal-catalyzed nitrene-transfer reactions arouse intense
interest as clean and efficient procedures for the synthesis of
biologically or pharmacologically active amines.[1] Efficient
rhodium-based catalysts have been described[2] but intensive
efforts are currently devoted to replacing these heavy metals
by iron systems.[3] Owing to their close analogy with the now
well-established FeIV oxo complexes, obtaining reactive FeIV
imido species has become a major endeavor. Such FeIV imido
species could be formed by using highly encumbered ligands
of low denticity to isolate three- and four-coordinate imido Fe
complexes, but these systems until recently did not exhibit
a strong reactivity.[4–6] In contrast, five- and six-coordinate Fe
imido species mediate efficient nitrene transfers[7–11] and have
=
Scheme 1. Reaction of ArI NTs with the mixed-valent complex 1 with
NCCH3 as a labile ligand [FeIIIFeII(NCCH3)(L-Cl)(mpdp)]2+, mpdp=m-
phenylene dipropionate, Ar=o-tert-butylsulfonylphenyl, Ts=tosyl, HL-
Cl=[2-((bis(pyridyl-2-methyl)amino)methyl][6-((pyridyl-2-methyl)(2,6-
dichlorobenzyl))amino)methyl]-4-methylphenol. For simplicity the sol-
vent is not included in the formula.
ortho benzylic positions) was shown to mediate fast and
quantitative ortho-amination of the dangling benzyl group of
the ligand.[13] The remarkable reactivity of these systems
prompted us to investigate the nature of the species active in
the nitrene transfer to gain information useful to design
efficient Fe catalysts for these reactions. Herein we describe
how, by using a combination of desorption electrospray
ionization mass spectrometry (DESI-MS), quantitative chem-
ical quench experiments, and DFT calculations, we have
obtained conclusive evidence for the occurrence of an
[*] Dr. E. Gourꢀ, Dr. F. Avenier, P. Dubourdeaux, Dr. O. Sꢀnꢁque,
Dr. M. Clꢀmancey, Dr. J.-M. Latour
Universitꢀ Grenoble Alpes, LCBM, and
CEA, DSV, iRTSV, LCBM, PMB, and
CNRS UMR 5249, LCBM
38054 Grenoble (France)
E-mail: jean-marc.latour@cea.fr
{FeIIIFeIV NTs} intermediate that is very active in HC abstrac-
C. Lebrun, Dr. P. Maldivi
=
Universitꢀ Grenoble Alpes, UMR E3, and
CEA, DSM, INAC, SCIB, RICC
38054 Grenoble (France)
tion and nitrene-transfer reactions.
=
When 1 was treated with a tosyl aryliodinane ArI NTs,
the initially blue solution immediately turned red owing to the
formation of a new chromophore absorbing at 482 nm.[12] This
chromophore was identified as the diferric tosylamido
derivative [FeIIIFeIII(NHTs)(L-Cl)(mpdp)]2+ 4 (Scheme 1).
As observed in several instances, 4 could originate from an
intra- or intermolecular HC abstraction by a higher valent
E-mail: pascale.maldivi@cea.fr
Dr. F. Albrieux
CCSM UMR 5246 CNRS-Universitꢀ Claude Bernard Lyon 1 (France)
[**] J.-M.L. and P.M. acknowledge the partial support of Labex ARCANE
(ANR-11-LABX-0003-01).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201307429.
III
IV
2+
imido intermediate [Fe Fe ( NTs)(L-Cl)(mpdp)] (3).[4,14]
=
1580
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1580 –1584

Angewandte
Chemie
In the hope to detect it, we monitored the formation of 4 by
optical stopped-flow spectroscopy in acetonitrile at room
temperature. Its formation was evidenced by an absorption
growing at 482 nm with a simultaneous decrease of the band
at 611 nm and an isosbestic point at 595 nm (Supporting
Information, Figure S1). The same spectroscopic features
were observed in low-temperature experiments (down to
À808C) in propionitrile (Figure S2). The 611 nm absorption
could be associated with an FeIV NTs species since they have
=
been reported to have a low-intensity absorption near
650 nm.[8,9] We thus monitored the reaction by Mçssbauer
spectroscopy at À808C in the hope of trapping 3. Figure S3
shows the evolution of spectra recorded at times spanning the
range 0 to 21 min. These spectra can be explained by the
transformation of 1 into another FeIIIFeII species 2 and into 4
but do not uncover an FeIV derivative. Interestingly, 4
appeared as a mixture of two FeIIIFeIII components which
differ in the quadrupole splitting (QS) of a single site
(Table S2 and see below). These experiments therefore
demonstrate that 3 does not accumulate in the millisecond
time scale because its rate of formation is smaller than its rate
of reaction through HC abstraction. A faster technique was
thus needed. Such a situation was successfully addressed very
recently by DuBois and co-workers[15] who could detect
intermediates in the nitrene-transfer reactions of a dirhodium
catalyst using the DESI-MS technique, suited to identifying
intermediates with lifetimes shorter than milliseconds.[16,17]
We used this technique to study the reaction of 1 with
=
Figure 1. DESI-MS monitoring of the reaction of 1 with ArI NTs. Inset:
superimposition of the experimental (line) and calculated (bars) iso-
topic patterns of the nitrenyl derivative 3-15N obtained by reaction with
ArI 15NTs.
=
Scheme 2.[18] Reaction of 1 with ArI NTs rapidly forms the
iodinane adduct 2 detected in the optical (at 611 nm) and
Mçssbauer low-temperature experiments. Compound
decomposes into the imido intermediate 3 that accumulates
only to a small extent (detectable solely in DESI-MS experi-
ments) owing to its very high reactivity and either transfers
the NTs group to thioanisole to give 5 or abstracts a HC to give
the amide 4.
=
2
ArI NTs or its 15N analogue at room temperature. Com-
=
pound 1 deposited on a polymer surface, was treated with
=
a solution of ArI NTs in anhydrous acetonitrile. Apart from
Chemical quench experiments were then undertaken to
=
the signals of 1 and ArI NTs at m/z 464.0727 and 493.9967,
further probe and quantitate the formation of 3. In a typical
=
respectively, the experimental mass spectrum (Figure 1)
showed patterns of signals for dications at m/z values of
approximately 549.1 and 710.5655. The signal at 710.5655 is
experiment 1 was treated with ArI NTs in the presence of
a 1-electron or 2-electron donor (2,4-ditertbutylphenol,
DTBPH, or thioanisole, Ph-S-Me, respectively). The reaction
was monitored by UV/Vis spectroscopy to determine the
III
II
=
assigned to 2, the iodinane adduct of 1 [Fe Fe (ArI NTs)(L-
Cl)(mpdp)]2+. This assignment was based on the perfect
match with the expected mass, the close similarity of the
experimental and theoretical isotopic distribution, and
the shift of half a mass unit of the whole pattern when
using ArI 15NTs (Figure S6). The isotopic pattern in the
=
range m/z 547.5–552.0 is more complicated and does not
correspond to a single species. Indeed the observed
masses could be associated to a mixture of 4 (m/
z 549.0850) and the sought after imido intermediate 3
(m/z 548.5804). This conclusion is supported by the fact
that 1) the isotopic patterns recorded at various times
could be reproduced very satisfyingly by mixtures of 3
and 4 with various ratios ranging from 70/30 to 10/90
(Figure S7), 2) when thioanisole, a substrate of the
nitrene-transfer reaction, was added to the iodinane
reactant solution the isotopic pattern matched that of 4
alone (Figure S8), which is consistent with a fast reac-
tion of 3 with thioanisole competing with HC abstraction,
and 3) most importantly, when ArI 15NTs was used the
=
pattern expected for 3-15N at m/z 549.0825 was indeed
observed (Figure 1 inset). These DESI-MS experiments
thus allows the spectroscopic and reactivity observa-
Scheme 2. Cycle accounting for the species identified in the DESI-MS experi-
tions to be rationalized in the catalytic cycle shown in ments.
Angew. Chem. Int. Ed. 2014, 53, 1580 –1584
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1581

.
Angewandte
Communications
4 but the source of the HC atom in absence of DTBPH is
unknown. In our conditions, simple consideration of the
respective bond dissociation energy (BDE) values suggest
that the eight benzylic positions of the ligand HL-Cl and
dicarboxylate mpdp2À are the most likely HC donors. This was
=
confirmed by titrating 1 by ArI NTs in absence and presence
of a very good HC donor, dihydroanthracene (DHA, BDE =
76 kcalmolÀ1) and monitoring the absorbance at 482 nm
(Figure S15). In absence of DHA, a plateau was obtained
=
after addition of approximately 0.9 equivalents of ArI NTs at
an absorption of 1.6. In presence of DHA, the plateau was
Scheme 3. Experiments used to determine the formation of 3, see text
for details.
=
obtained at approximately 1.0 equivalents of ArI NTs and an
absorption of 1.75. The higher value of the plateau in presence
of DHA indicates that in its absence part of the chromophore
is consumed. The quantitative difference of about 12% is
consistent with the presence of eight reactive benzylic
positions in the complex that would account for most of the
HC for the formation of 4. In the course of our low-temper-
ature experiments intended to trap 3 we noticed that at
À608C, after an initial time lag (Figure S2), formation of 4
occurred monoexponentially with t1/2 ꢀ 57 s. Since 3 does not
accumulate in the reaction, its decay by HC abstraction is faster
than its formation. Owing to the presence of 16 benzylic
hydrogen atoms it can be roughly estimated that at À608C the
rate of HC abstraction is kH @ 1/(t1/2)[complex] ꢀ 6mÀ1 sÀ1. This
value is in the range observed for the most reactive
compounds described by Sorokin[20] pointing to a very
strong HC abstraction capability of 3.
oxidation state (FeIIIFeII vs FeIIIFeIII) of the resulting complex.
The electron count of the reaction was determined by NMR
=
spectroscopy and HPLC quantification of the product/ArI
NTs ratio. These experiments are summarized in Scheme 3. In
a first experiment (A in Scheme 3), 1 was treated with
=
1 equivalent of ArI NTs in the presence of 1 equivalent of
DTBPH. UV/Vis spectroscopy showed the disappearance of
1 and formation of 4 while NMR analysis indicated the
formation of 1 equivalent of the dimer (DTBP)2 and there-
fore the abstraction of a single HC (Figure S9, S10). This result
showed that the oxidation state of 3 is one unit higher than
that of 4 and that 3 contains an NTs group. In a second
experiment (B in Scheme 3), 1 was again treated with
=
1 equivalent ArI NTs but this time in the presence of
10 equivalents of Ph-S-Me. The UV/Vis spectrum was
unchanged after the reaction and 1 equivalent of sulfilimine
5 Ph-S(NTs)Me was formed (Figure S11, S12). These obser-
vations revealed that the oxidation state of 3 is two units
higher than that of 1 and that it can transfer an NTs group. All
The experiments have provided evidence for the forma-
tion of 3 and shown that it cannot be intercepted and
spectroscopically characterized. To investigate its molecular
and electronic structures, we thus relied on DFT calculations.
Firstly, thanks to the extensive computational literature on Fe
complexes, we successfully calibrated our calculations of
structural and Mçssbauer parameters on related known
dinuclear Fe complexes as well as a mononuclear imido FeIV
species (Tables S5–7). Then both bridging and terminal
binding modes of the N(H)Ts group were considered. More-
over, because in 1 the solvent binds in the position trans to the
bridging phenoxide[21,22] whereas in the insertion product of
the non-chloroprotected ligand, the N-benzyltosylate is
bound in the cis position,[13] we considered both cis and
trans isomers for the calculations of 3 and 4. First, we explored
the tosylamido complex {FeIIIFeIII-NHTs} 4 and observed that
the terminal mode is clearly preferred, the bridging structure
being about 23 kcalmolÀ1 higher. The terminal cis and trans
isomers were very close in energy (cis 4.2 kcalmolÀ1 lower
than trans at the B3LYP level). An illustration of the
molecular structure and the bond lengths to each iron atom
are given in Figure S18 and Table S8. The relevant optimized
III
IV
=
these results suggest that 3 can be formulated as a {Fe Fe (
NTs)} species.
We then intended to characterize the high reactivity of 3
both in nitrene-transfer and HC-abstraction reactions in
a more quantitative manner. Owing to the very fast rate of
the sulfimidation reaction, we resorted to competition experi-
ments between substituted thioanisoles. In a typical experi-
=
ment, 1 was treated with 1 equivalent ArI NTs in the
presence of a mixture of 10 equivalents of thioanisole and
10 equivalents of a para-substituted thioanisole (p-X-Ph-S-
Me, X = NO2, COMe, Cl, Me, OMe). The reaction mixture
was analyzed by HPLC to quantify the two sulfilimines: Ph-
S(NTs)-Me and p-X-Ph-S(NTs)-Me. The ratio of the quanti-
ties of sulfilimines was taken as the ratio of the kinetic
constants of the two reactions. The results are summarized in
Table S4. Figure S13 shows that logkX/kH is linearly corre-
lated (r2 = 0.97) to the redox potentials of the thioethers with
a slope of À1.9(2). This result indicates that the sulfimidation
occurs by a direct nitrene transfer (instead of a one-electron
oxidation and recombination).[19] In addition, a linear corre-
lation (r2 = 0.98) of logkX/kH versus the Hammett constant
sp+(X) is observed with a slope À0.74(6) revealing the strong
electrophilicity of 3 (Figure S14).
parameters featuring the FeIII NHTs bond are in remarkably
À
good agreement with the X-ray structure of the six-coordinate
tolylamido FeIII compound described by Borovik et al.[14] with
an FeIII N bond of 1.968 ꢀ and a Fe-N-S angle of 1278,
À
whereas our optimized FeIII NTs bond is 1.94 ꢀ, with a Fe-N-S
À
angle of 1288. Compound 4 was isolated in the low-temper-
ature experiments as a mixture of two species differing mostly
by their QS values (Figure S3). The values of the Mçssbauer
QS calculated for the cis and trans isomers of 4 are in very
The strong HC-abstraction ability of 3 is the cause of its
transiency and we searched to assess this ability further. As
shown in Scheme 3A, DTBPH acts as an HC donor to produce
1582
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1580 –1584

Angewandte
Chemie
good agreement with those observed experimentally for the
two species thus allowing to assign their structures (Table S2).
After the validation of our computational procedure for
polarization from 5 to about 4 for a high-spin FeIII ion owing
to donation from ligands. For FeIV, spin polarized donation
from ligands also occurs, and in particular from the NTs and Ob
(phenoxide) atoms. In the Ob case, the absence of significant
radical character on the Ob atom (Table 1) and the presence
complex 4 we performed the same analysis on the {FeIIIFeIV
=
NTs} intermediate 3 taking into account the two possible spin
states for FeIV, S = 1 and S = 2. The bridging mode is found
again at higher energies than the terminal one by at least
16 kcalmolÀ1, while the cis mode proved to be more stable
than the trans mode by at least 7 kcalmolÀ1.
À
of six equivalent C C bonds in the phenoxide ring regardless
of the electronic configuration exclude any possibility of
a phenoxyl radical (see Supporting Information).
Two configurations arising from the S = 2 state (FeIV-aNTs
and FeIV-bNTs) exhibit partial radical character on CNTs. In
One prominent result of the geometry optimizations is the
close energy of both S = 1 and 2 spin states (difference less
than 4 kcalmolÀ1 at the B3LYP level). A second major result
is the existence of four low-lying configurations arising from
NTs-to-FeIV charge-transfer states for S = 1 and S = 2 which
are easily identified by characteristic spin densities (see below
and Supporting Information for details). Careful optimiza-
tions of the geometries of these different configurations by
control of the spin polarizations of Fe and NTs atoms yielded
structural parameters gathered in Tables S10 and S11 while
Figure 2 illustrates the structure of 3 in the most stable cis
mode. The Fe–ligand bond lengths in Tables S10 and S11 are
consistent with experimental FeIII or FeIV bond lengths. Of
particular interest is the short Fe–NTs distance (ca. 1.71 ꢀ) of
the iron-imido group in species 3 S = 1, very close to the
corresponding distance found in the complex reported by Que
et al. (1.73 ꢀ).[9]
contrast, the third configuration (FeIII CNTs) involves a com-
À
5/2
plete charge transfer resulting in a high-spin FeIII atom
bonded to a nitrenyl radical. A complete charge transfer also
gives rise to the FeIII CNTs configuration in the S = 1 state.
Both radical states are consistent with other Fe N(imido)
À
1/2
À
systems.[4,6,9,23] Note that the optimized geometries corre-
sponding to these two FeIII CNTs configurations exhibit Fe NTs
bonds consistent with FeIII spin state (S = 5/2 or 1/2) (see
Supporting Information). Such radical character on the
nitrogen atom gives rise to a low-energy vacant orbital on
this atom. They are represented on Figure 3 for both S = 1
and S = 2 states, with their energy level above the SOMO,
suggesting in both cases a very high electrophilicity of the NTs
atom.
À
À
Figure 3. Lowest unoccupied b MO on complex 3 in the a) S=1 state;
b) S=2 state corresponding to the FeIII CNTs species (see text),
together with its gap above the SOMO at the B3LYP level. Top view
from phenoxide axis.
À
Figure 2. DFT optimized structure of 3. a) front view, b) top view along
the phenoxide main axis magenta FeIII, green FeIV. Bond lengths and
angles are collected in Table S10.
Another remarkable feature emerges from our detailed
investigation of the electronic structure of 3 based on spin
densities. Indeed, all low-lying configurations have an
enhanced radical character of the NTs donor atom (Table 1
and Table S12). As is usual, we observed a reduced spin
These results are reminiscent of the widely studied FeIV
oxo species derived from heme or non-heme models, known
to activate the oxygen-transfer reaction. Their activation
mechanism through multiple spin-state pathways has been
successfully deciphered by several theoretical and spectro-
scopic approaches.[24–27] It was also developed for the nitrene-
transfer reaction for a heme-model system.[28] In particular,
these studies strongly suggest for the S = 2 pathway, an
Table 1: Spin densities (1s) and total bonding energies (eV) computed at
the B3LYP level, for complex 3 in cis coordination mode.[a]
activated species described as containing a high-spin FeIII CO
À
FeIV S=2
FeIV-bNTs
FeIV S=1
FeIII1/2-CNTs
moiety.[24,26] In the light of these seminal studies, our finding of
FeIII5/2-CNTs
FeIV-aNTs
FeIII CNTs configurations in both S = 1 and S = 2 almost
À
E B3LYP
1s FeIV
1s N
À774.44
4.03
1.1
À0.05
À774.48
3.36
0.33
À0.13
À774.63
4.07
À774.40
0.94
0.93
À0.16
isoenergetic states may explain the very high activity of
complex 3.
À0.40
To summarize, the present study has provided a consistent
experimental picture for the occurrence of an FeIIIFeIV imido
species 3 formed upon reaction of 1 with the nitrene donor
1s Ob
0.07
[a] Spin densities (1s) of Fe atom, NTs, and the phenoxide bridging
oxygen (noted Ob), and total bonding energies (eV) computed at the
=
ArI NTs and its very high reactivity in HC-abstraction and
III,IV
B3LYP level. Values are given for the broken symmetry state of 3 (Fe2
binuclear complex). Configurations noted FeIV-bNTs and FeIV-aNTs
correspond to partial charge transfer configurations (see Supporting
Information).
NTs-transfer reactions. This high reactivity that is comparable
to the most reactive oxo transfer agents precludes its
interception by usual methods responsive to the millisecond
Angew. Chem. Int. Ed. 2014, 53, 1580 –1584
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1583

.
Angewandte
Communications
[11] Y. Liu, X. Guan, E. L.-M. Wong, P. Liu, J.-S. Huang, C.-M. Che,
J. Am. Chem. Soc. 2013, 135, 7194.
time range. To obtain valuable information on the molecular
and electronic structures of 3, extensive DFT calculations
were performed and calibrated on the structures and Mçss-
bauer parameters of several related diiron complexes so as to
define a fully reliable methodology. This procedure allowed
[12] F. Avenier, J.-M. Latour, Chem. Commun. 2004, 1544.
[13] F. Avenier, E. Gourꢂ, P. Dubourdeaux, O. Sꢂnꢃque, J.-L. Oddou,
J. Pꢂcaut, S. Chardon-Noblat, A. Deronzier, J.-M. Latour,
Angew. Chem. 2008, 120, 727; Angew. Chem. Int. Ed. 2008, 47,
715.
[14] R. L. Lucas, D. R. Powell, A. S. Borovik, J. Am. Chem. Soc. 2005,
127, 11596.
[15] R. H. Perry, T. J. Cahill III, J. L. Roizen, J. Du Bois, R. N. Zare,
Proc. Natl. Acad. Sci. USA 2012, 109, 18295.
À
us to propose for 3 structural parameters (in particular Fe NTs
bond lengths) that are perfectly consistent with those
published for similar species. Moreover, it revealed the
strong radical character of NTs in very-low-lying electronic
configurations originating from both the S = 1 and S = 2 states
of the FeIV ion that are likely to confer the high reactivity of 3.
Whether this reactivity is associated to the particular environ-
ment of the Fe ion or is due to the binuclear nature of the
complex is currently being investigated.
[16] Z. Takats, J. M. Wiseman, B. Gologan, R. G. Cooks, Science
2004, 306, 471.
[17] A. B. Costa, R. G. Cooks, Chem. Phys. Lett. 2008, 464, 1.
[18] DESI normally samples species generated in solution and can be
used to identify solution species through MS techniques.
However, owing to solvent evaporation the possibility that
species detected in DESI conditions are the same as in bulk
solution may not be fully warranted. Nevertheless, the strong
consistency of our DESI-MS results with those derived from all
other bulk solution techniques make this possibility highly
unlikely in the present cases.
Received: August 23, 2013
Revised: October 14, 2013
Published online: January 13, 2014
Keywords: diiron complexes · mass spectrometry ·
.
[19] Y. Goto, T. Matsui, S. Ozaki, Y. Watanabe, S. Fukuzumi, J. Am.
Chem. Soc. 1999, 121, 9497.
mixed-valent compounds · nitrene transfer · radicals
[20] E. V. Kudrik, P. Afanasiev, L. X. Alvarez, P. Dubourdeaux, M.
Clꢂmancey, J.-M. Latour, G. Blondin, D. Bouchu, F. Albrieux,
S. E. Nefedov, A. B. Sorokin, Nat. Chem. 2012, 4, 1024.
[21] W. Kanda, W. Moneta, M. Bardet, E. Bernard, N. Debaecker, J.
Laugier, A. Bousseksou, S. Chardon-Noblat, J.-M. Latour,
Angew. Chem. 1995, 107, 625; Angew. Chem. Int. Ed. Engl.
1995, 34, 588.
[1] G. Dequirez, V. Pons, P. Dauban, Angew. Chem. 2012, 124, 7498;
Angew. Chem. Int. Ed. 2012, 51, 7384.
[2] J. Roizen, M. Harvey, J. DuBois, Acc. Chem. Res. 2012, 45, 911.
[3] A. Correa, O. Garcia Mancheno, C. Bolm, Chem. Soc. Rev. 2008,
37, 1108.
[4] P. L. Holland, Acc. Chem. Res. 2008, 41, 905.
[5] C. T. Saouma, J. C. Peters, Coord. Chem. Rev. 2011, 255, 920.
[6] E. R. King, E. T. Hennessy, T. A. Betley, J. Am. Chem. Soc. 2011,
133, 4917.
[7] J. P. Mahy, P. Battioni, D. Mansuy, J. Fischer, R. Weiss, J.
Mispelter, I. Morgenstern-Badarau, P. Gans, J. Am. Chem. Soc.
1984, 106, 1699.
[8] M. P. Jensen, M. P. Mehn, L. Que, Jr., Angew. Chem. 2003, 115,
4493; Angew. Chem. Int. Ed. 2003, 42, 4357.
[9] E. J. Klinker, T. A. Jackson, M. P. Jensen, A. Stubna, G. Juhasz,
E. L. Bominaar, E. Mꢁnck, L. Que, Jr., Angew. Chem. 2006, 118,
7554; Angew. Chem. Int. Ed. 2006, 45, 7394.
[22] S. Chardon-Noblat, O. Horner, B. Chabut, F. Avenier, N.
Debaecker, P. Jones, J. Pꢂcaut, L. Dubois, C. Jeandey, J.-L.
Oddou, A. Deronzier, J.-M. Latour, Inorg. Chem. 2004, 43, 1638.
[23] M. Jaccob, G. Rajaraman, Dalton Trans. 2012, 41, 10430.
[24] M. Srnec, S. D. Wong, J. England, L. Que, Jr., E. I. Solomon,
Proc. Natl. Acad. Sci. USA 2012, 109, 14326.
[25] S. Shaik, H. Hirao, D. Kumar, Acc. Chem. Res. 2007, 40, 532.
[26] S. Ye, F. Neese, Proc. Natl. Acad. Sci. USA 2011, 108, 1228.
[27] S. P. de Visser, J. U. Rohde, Y. M. Lee, J. Cho, W. Nam, Coord.
Chem. Rev. 2013, 257, 381.
[28] Y. Moreau, H. Chen, E. Derat, H. J. Hirao, C. Bolm, S. Shaik, J.
Phys. Chem. B 2007, 111, 10288.
[10] P. Leeladee, G. N. L. Jameson, M. A. Siegler, D. Kumar, S. P.
de Visser, D. P. Goldberg, Inorg. Chem. 2013, 52, 4668.
1584
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1580 –1584