Evidence for an alternative to the oxygen rebound mechanism in C-H bond activation by non-heme FeIVO complexes

Article abstract of DOI:10.1021/ja308290r

The hydroxylation of alkanes by heme Fe^IVO species occurs via the hydrogen abstraction/oxygen rebound mechanism. It has been assumed that non-heme Fe^IVO species follow the heme Fe^IVO paradigm in C-H bond activation reactions. Herein we report theoretical and experimental evidence that C-H bond activation of alkanes by synthetic non-heme Fe ^IVO complexes follows an alternative mechanism. Theoretical calculations predicted that dissociation of the substrate radical formed via hydrogen abstraction from the alkane is more favorable than the oxygen rebound and desaturation processes. This theoretical prediction was verified by experimental results obtained by analyzing iron and organic products formed in the C-H bond activation of substrates by non-heme Fe^IVO complexes. The difference in the behaviors of heme and non-heme Fe^IVO species is ascribed to differences in structural preference and exchange-enhanced reactivity. Thus, the general consensus that C-H bond activation by high-valent metal-oxo species, including non-heme Fe^IVO, occurs via the conventional hydrogen abstraction/oxygen rebound mechanism should be viewed with caution.

Full text of DOI:10.1021/ja308290r

Communication
pubs.acs.org/JACS
Evidence for an Alternative to the Oxygen Rebound Mechanism in
C−H Bond Activation by Non-Heme FeIVO Complexes
Kyung-Bin Cho,†,§ Xiujuan Wu,†,§ Yong-Min Lee,† Yoon Hye Kwon,† Sason Shaik,*,‡
and Wonwoo Nam*,†
†Department of Bioinspired Science and Department of Chemistry and Nano Science, Ewha Womans University, Seoul 120-750,
Korea
‡Institute of Chemistry and The Lise Meitner-Minerva Center for Computational Quantum Chemistry, The Hebrew University of
Jerusalem, 91904 Jerusalem, Israel
S
* Supporting Information
Scheme 1
ABSTRACT: The hydroxylation of alkanes by heme
FeIVO species occurs via the hydrogen abstraction/oxygen
rebound mechanism. It has been assumed that non-heme
FeIVO species follow the heme FeIVO paradigm in C−H
bond activation reactions. Herein we report theoretical and
experimental evidence that C−H bond activation of
alkanes by synthetic non-heme FeIVO complexes follows
an alternative mechanism. Theoretical calculations pre-
dicted that dissociation of the substrate radical formed via
hydrogen abstraction from the alkane is more favorable
than the oxygen rebound and desaturation processes. This
theoretical prediction was verified by experimental results
obtained by analyzing iron and organic products formed in
the C−H bond activation of substrates by non-heme
FeIVO complexes. The difference in the behaviors of heme
and non-heme FeIVO species is ascribed to differences in
structural preference and exchange-enhanced reactivity.
Thus, the general consensus that C−H bond activation by
high-valent metal−oxo species, including non-heme FeIVO,
occurs via the conventional hydrogen abstraction/oxygen
rebound mechanism should be viewed with caution.
products (Scheme 1B, pathways a and c).8,9 In these
hydroxylation/desaturation reactions, FeII species are presumed
to be the end product.5−9 Indeed, theoretical calculations have
found a small barrier for the oxygen rebound and desaturation
reactions by the FeIII−OH species.7,9 Moreover, since hydroxy-
lated/desaturated products are usually detected in experiments,
there seemed to be no doubt that non-heme FeIVO species follow
the heme FeIVO paradigm (e.g., the hydrogen abstraction/
oxygen rebound mechanism). However, several recent reports
do not fit into this picture. For example, it has been shown that
two FeIVO molecules containing the Me,HPytacn ligand, [FeIVO-
igh-valent FeIVO species are ubiquitous in nature and
perform a wide range of important chemical and biological
H
reactions.1 In heme enzymes such as cytochrome P450, FeIVO
porphyrin π-cation radical intermediates, called compound I
(Cpd I), are known to perform a variety of oxygenation reactions,
including alkane hydroxylation.2 It has been shown in both
experimental and theoretical studies that alkane hydroxylation by
Cpd I is initiated by a rate-determining hydrogen abstraction step
by the (Porp•+)FeIVO species (Scheme 1A, pathway a), which is
followed by an oxygen rebound step between the resulting
(Porp)FeIVOH and substrate radical species (Scheme 1A,
pathway b).2 The oxygen rebound mechanism has been strongly
supported and is well-established in enzymatic reactions using a
special group of substrates termed “radical clocks”2c as well as in
biomimetic iron−oxo porphyrin reactions.3,4
(
Me,HPytacn)(S)]2+,10 are required to desaturate one molecule of
9,10-dihydroanthracene (DHA), thereby yielding two molecules
of the corresponding FeIII species and one molecule of
anthracene.11 In the hydroxylation of alkanes by [(Bn-TPEN)-
FeIVO]2+ (1) and [(N4Py)FeIVO]2+ (2),10 only low yields of FeII
products but significant amounts of organic products generated
by O2 trapping of alkyl radicals (·CR3) were detected.12 The
In non-heme iron enzymes and models, C−H bond activation
of alkanes by non-heme FeIVO species has been proposed to give
alcohol products via the hydrogen abstraction/oxygen rebound
mechanism (Scheme 1B, pathways a and b).5−7 In addition, non-
heme FeIVO intermediates occasionally produce desaturated
Received: August 21, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja308290r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
latter result implies that the substrate radical appears to escape
from the cage (Scheme 1B, pathway d) and hence reacts with O2
instead of rebounding (Scheme 1B, pathway g). Calculations on
C−H activation of cyclohexane (c-Hex) by 2 indicate that while
the rebound barrier is as low as 1.9 kcal/mol (ΔG in solvent), the
dissociation pathway is much more preferred on the free energy
scale.9b We also recently showed that analogous reactions with
non-heme MnIVO complexes lead to final products containing
MnIII, not MnII, despite the formation of significant amounts of
hydroxylated products.13 An alternative mechanism for the
MnIVO reactions was therefore proposed, in which a substrate
radical dissociates and reacts with a second MnIVO molecule to
give the corresponding MnIII species and hydroxylated products
(analogous to Scheme 1B, pathways e and f). Hence, there is
mounting evidence in the literature that the oxygen rebound and
desaturation processes may not be universal in C−H bond
activation reactions by high-valent metal−oxo intermediates.
Herein we report a combination of theoretical and
experimental evidence that hydroxylation of alkanes by synthetic
non-heme FeIVO complexes does not follow the conventional
oxygen rebound and desaturation mechanisms (Scheme 1B,
pathways b and c). Instead, dissociation of the substrate radical
formed via hydrogen abstraction from the alkane C−H bond by
FeIVO species (Scheme 1B, pathway d) is shown to be more
favorable than the oxygen rebound and desaturation processes in
non-heme iron model reactions.
Theoretical Calculations. We investigated the C−H bond
activation of c-Hex by the two non-heme FeIVO complexes 1 and
2 using density functional theory (DFT).14 We calculated the
energy barriers for the rebound, desaturation, and dissociative
processes (Scheme 1B, pathways b−d). Although similar
calculations for 2 using a different computational procedure
were presented previously,9b the ones presented here were
performed to ensure that the methodology would be comparable
to that used for 1. The values discussed here are electronic
energies only, calculated at the B3LYP/LACV3P*+//LACVP
level including CPCM solvation. The calculated free energies
(ΔG) are reported in the Supporting Information (SI) only, as
those values are deemed less reliable in solvent-optimized
systems15 (see the SI for more details on methodology).
Figure 1 shows the calculated energy profiles of the reactions.
Apart from the initial reactant state, which has an S = 1 ground
state, we found that singlet and triplet energies along the reaction
path were high enough to not matter for the current study (see
Tables S1−S4 in the SI). Hence, the energy values mentioned
below are for the quintet (S = 2) states (relative to the triplet
reactant state), unless stated otherwise. The lower energies on
the quintet surface follow well-established patterns shown by
many non-heme synthetic FeIVO species.7e,16 This is due to
exchange-enhanced reactivity (EER), where the added electron
from the substrate goes into the σz*2 orbital of the FeIVO moiety,
strengthening the stabilizing exchange interactions with the other
unpaired electrons.17 This of course assumes a spin-inversion
probability of unity, which may or may not be correct;18 however,
our final conclusions would not change even if singlet and triplet
energies were used (see Tables S3 and S4).
The relative energies for the C−H activation transition states
(TSC−H) were found to be the highest ones in the whole reaction
landscape (Figure 1), in accord with earlier findings that C−H
bond activation by the iron−oxo complex is the rate-determining
step.6,7a The barriers for 1 (14.2 kcal/mol) and 2 (19.1 kcal/mol)
indicate that the former is more reactive, in agreement with
previous experimental results.6,18 The rebound transition states
Figure 1. B3LYP/LACV3P*+//LACVP-calculated energies for the
reactions of (top) 1 and (bottom) 2 with c-Hex. Only the quintet
surfaces are shown, with all values in kcal/mol relative to the triplet
reactant state. Singlet and triplet values are given in the SI.
(TSreb) were found to be at 8.1 and 13.2 kcal/mol for 1 and 2,
respectively, and the desaturation transition states (TSdes) were
found to be quite comparable to the rebound barriers (i.e., 8.1
and 12.5 kcal/mol, respectively). Hence, a competition between
these two follow-up reactions would be expected to occur. What
is interesting here is that both TSreb and TSdes lie higher than the
system energy when the substrate radical has dissociated from
the FeIIIOH complex (i.e., Edis = 6.7 and 12.0 kcal/mol for 1 and
2, respectively). While the electronic energy differs from the
other barriers by less than 2 kcal/mol, the entropy increase
during dissociation should increase this difference further. The
dissociation of the FeIIIOH intermediate from the substrate
radical is analogous to the proposed MnIIIOH case13 and is in
agreement with earlier experimental12 and theoretical findings.9b
Hence, it is plausible to propose that dissociation of the substrate
radical could be competitive with or even preferable to the
rebound and desaturation pathways. To test this proposal, we
performed the required C−H activation experiments with the
synthetic non-heme FeIVO complexes and analyzed the iron (i.e.,
FeII vs FeIII) and organic products formed in these reactions.
Experimental Evidence. The C−H bond activation of hydro-
carbons by 1 and 2 was performed with various substrates such as
DHA, 1,4-cyclohexadiene (CHD), triphenylmethane (TPM),
ethylbenzene (EB), and c-Hex. Upon addition of EB to 1 at 25
°C, 1 disappeared completely, but [FeII(Bn-TPEN)]2+ was
formed in only ∼10% yield (Figure 2a; see Figure S1 in the SI for
UV−vis spectra of [FeII(Bn-TPEN)]2+ and 1). Interestingly,
addition of ferrocene (Fc) to the resulting solution caused the full
formation of [FeII(BnTPEN)]2+ and the formation of Fc+ in
∼90% yield (Figure 2a). Spectroscopic analyses of the reaction
solutions were done with electron paramagnetic resonance
(EPR) spectroscopy and electrospray ionization mass spectrom-
B
dx.doi.org/10.1021/ja308290r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
activation reaction (i.e., in CH3CN at 25 °C), the reactants
remained intact and the formation of an FeIII species (i.e.,
[FeIII(OH)(N4Py)]2+) was not observed (Figure S6). This result
demonstrates that the FeIII species is formed via the dissociation
mechanism (Scheme 1B, pathway d) rather than a compro-
portionation reaction between FeIVO and FeII species. A similar
conclusion was reached in the C−H bond activation reaction of
hydrocarbons by non-heme MnIVO species.13,21
Product analysis revealed that products formed in the C−H
bond activation reactions by non-heme FeIVO complexes
depended on the reaction conditions, such as the presence of
O2 or an alkyl radical scavenger (i.e., CCl3Br) (see the
Experimental Section in the SI). For example, cyclohexanol
(25
2%) and cyclohexanone (11
2%) were obtained as
products in the reaction of 1 (2.0 mM) and c-Hex (1.0 × 103
mM) under an Ar atmosphere in CH3CN at 25 °C. An isotope
experiment with 18O-labeled 1 (1-18O) showed that the oxygens
in the cyclohexanol and cyclohexanone products were derived
from 1 (Figure S7). When the identical reaction was carried out
with 1-18O in the presence of 16O2, the product distribution was
different [i.e., cyclohexanol (10 2%) and cyclohexanone (21
3%)] and the cyclohexanol product contained both 16O (77
Figure 2. (a) UV−vis spectral changes observed in the reaction of 1 (2.0
mM) and EB (1.0 × 102 mM) in CH3CN at 25 °C followed by addition
of Fc (10 mM) to the resulting solution (1, blue line; 1 + EB, black line;
1 + EB + Fc, red line). The inset is an expansion of the 550−900 nm
region showing the spectral changes of Fc+ at 615 nm and 1 at 740 nm.
The amounts of [FeII(Bn-TPEN)]2+ at 390 nm and Fc+ at 615 nm were
calculated using ε values of 9200 and 400 M−1 cm−1 for [FeII(Bn-
TPEN)]2+ and Fc+, respectively. (b) X-band EPR spectra recorded at 5
K with the reaction solutions of 1 + EB (black solid line) and 1 + EB + Fc
(red dashed line). (c) ESI-MS spectra recorded with the reaction
solutions of (left) 1 + EB and (right) 1 + EB + Fc.
6%) and 18O (23
3%) (Figure S7). In addition, when we
carried out hydroxylations of c-Hex using 1 and 2 in the presence
of CCl3Br under an Ar atmosphere, we observed the formation of
bromocyclohexane as the sole product, indicating that a
cyclohexanyl radical was trapped by CCl3Br.3c,22 The exper-
imental evidence provided here strongly supports the proposal
that C−H bond activation by non-heme FeIVO complexes
prefers the dissociation process (Scheme 1B, pathway d) rather
than the oxygen rebound and desaturation processes (Scheme
1B, pathways b and c).
Comparison to Heme Systems. As shown above, the substrate
dependence of the second barrier determines the precise nature
of the products formed.9a As our TSreb is less than 2 kcal/mol
higher than Edis (disregarding entropy), it is not excluded that
certain substrates will more easily rebound/desaturate than
others. With a given substrate and otherwise identical conditions,
however, we can make some interesting comparisons between
heme and non-heme FeIVO systems. In our example using c-Hex,
calculations showed that the rebound reaction occurs over a very
low TSreb in the low-spin heme case,23 with which few other
reactions, including the dissociation reaction, could compete. In
the non-heme case, the rebound reaction would occur in the
high-spin state with a barrier that is small but nevertheless larger
than the dissociation barrier. The question is therefore “Why
does TSreb lie lower in the heme case?”
There may be several factors that could produce an energy
difference of this scale. Among these, we observed two plausible
ones in our calculations. Figure 3 shows the valence-electron
orbitals in the heme FeIVOH intermediate stage versus the non-
heme FeIIIOH stage. In heme, the low-spin configuration is
preferred, and the second electron is transferred to a low-lying
orbital (πy*z). Therefore, this process would not be expected to
contribute much to any barrier.2b,23 In non-heme cases, the
situation is somewhat similar, with the second electron being
transferred to π*yz as well, even though the system is in a high-spin
state.9a This has one drawback compared with the heme case.
The number of favorable exchange interactions between the five
unpaired electrons in a non-heme high-spin FeIIIOH config-
uration is diminished by 40% upon loss of the radical at πy*z,
presumably contributing to a barrier. The other factor is steric
effects due to substrate and/or ligand bulkiness. In the high-spin
etry (ESI-MS). An X-band EPR spectrum of the reaction solution
of 1 and EB exhibited signals at g = 2.38, 2.19, and 1.96 (Figure
2b), characteristic of a low-spin (S = 1/2) FeIII species, whereas
the 1H NMR Evans method19 at room temperature suggested an
S = 5/2 spin state. Upon addition of Fc to the resulting solution,
the EPR spectrum became silent (Figure 2b), indicating
reduction of the FeIII species to an FeII species. In ESI-MS
experiments, we observed ion peaks at m/z 248.0 and 289.0 in
the reaction solution of 1 and EB (Figure 2c, left panel); the
peaks correspond to [FeIII(OH)(Bn-TPEN)]2+ (calcd m/z
248.0) and [FeIII(OH)(Bn-TPEN)(CH3CN)2]2+ (calcd m/z
289.0). Reduction of the FeIII species by Fc afforded an ESI-MS
spectrum showing a prominent ion peak at m/z 239.6,
corresponding to [FeII(Bn-TPEN)]2+ (calcd m/z 239.5) (Figure
2c, right panel). On the basis of these analyses of the reaction
solutions, we conclude that an FeIII species, not an FeII species,
was formed as the major product in the oxidation of EB by 1. It
should also be mentioned that the formation of FeIII species was a
common feature in the C−H bond activation of hydrocarbons by
non-heme FeIVO complexes, irrespective of the iron complex
(i.e., 1 or 2) and the organic substrate (i.e., DHA, CHD, TPM,
EB, or c-Hex) (see Figures S2 and S3 for data from the 1 + c-Hex
and 2 + EB reactions, respectively).20 Furthermore, it is worth
noting that the amounts of FeII and FeIII products varied slightly
depending on the substrate (Table S13) but not on the FeIVO
complex and substrate concentrations (see Figures S4 and S5 for
the reactions of EB with 1 and 2, respectively). Finally, we ruled
out the possibility that the FeIII species was produced by a
comproportionation reaction between FeIVO and FeII species.
For example, when we mixed equal amounts of 2 and
[FeII(N4Py)]2+ under the conditions of the C−H bond
C
dx.doi.org/10.1021/ja308290r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
REFERENCES
■
(1) Nam, W. Acc. Chem. Res. 2007, 40, 465 and references therein.
(2) (a) Ortiz de Montellano, P. R. Cytochrome P450: Structure,
Mechanism, and Biochemistry; Kluwer/Plenum: New York, 2005.
(b) Shaik, S.; Kumar, D.; de Visser, S. P.; Altun, A.; Thiel, W. Chem.
Rev. 2005, 105, 2279. (c) Ortiz de Montellano, P. R. Chem. Rev. 2010,
110, 932. (d) Rittle, J.; Green, M. T. Science 2010, 330, 933.
(3) (a) Groves, J. T.; McClusky, G. A. J. Am. Chem. Soc. 1976, 98, 859.
(b) Groves, J. T.; Van Der Puy, M. J. Am. Chem. Soc. 1976, 98, 5290.
(c) Groves, J. T. J. Chem. Educ. 1985, 62, 928. (d) Groves, J. T. Proc. Natl.
Acad. Sci. U.S.A. 2003, 100, 3569.
(4) (a) Ogliaro, F.; Harris, N.; Cohen, S.; Filatov, M.; de Visser, S. P.;
Shaik, S. J. Am. Chem. Soc. 2000, 122, 8977. (b) Guallar, V.; Baik, M.-H.;
Lippard, S. J.; Friesner, R. A. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 6998.
(c) Kamachi, T.; Yoshizawa, K. J. Am. Chem. Soc. 2003, 125, 4652.
(5) (a) Krebs, C.; Galonic Fujimori, D.; Walsh, C. T.; Bollinger, J. M.,
́
Jr. Acc. Chem. Res. 2007, 40, 484. (b) Que, L., Jr. Acc. Chem. Res. 2007, 40,
493. (c) Nam, W. Acc. Chem. Res. 2007, 40, 522.
Figure 3. Valence-electron orbitals involved in electron shifts during the
rebound step for (top) heme (FeIVOH) and (bottom) non-heme
(FeIIIOH) intermediates. Blue arrows indicate the energetically
preferred reactions, while red arrows generally lead to higher barriers.
Green spin arrows indicate electrons nascent from the substrate. See the
text for a detailed discussion.
(6) (a) Kaizer, J.; Klinker, E. J.; Oh, N. Y.; Rohde, J.-U.; Song, W. J.;
Stubna, A.; Kim, J.; Munck, E.; Nam, W.; Que, L., Jr. J. Am. Chem. Soc.
̈
2004, 126, 472. (b) Seo, M. S.; Kim, N. H.; Cho, K.-B.; So, J. E.; Park, S.
́
K.; Clemancey, M.; Garcia-Serres, R.; Latour, J.-M.; Shaik, S.; Nam, W.
Chem. Sci. 2011, 2, 1039.
(7) (a) Hirao, H.; Kumar, D.; Que, L., Jr.; Shaik, S. J. Am. Chem. Soc.
2006, 128, 8590. (b) de Visser, S. P. J. Am. Chem. Soc. 2006, 128, 9813.
(c) Comba, P.; Maurer, M.; Vadivelu, P. J. Phys. Chem. A 2008, 112,
13028. (d) Hirao, H.; Que, L., Jr.; Nam, W.; Shaik, S. Chem.Eur. J.
2008, 14, 1740. (e) Geng, C.; Ye, S.; Neese, F. Angew. Chem., Int. Ed.
2010, 49, 5717.
(8) (a) Broun, P.; Shanklin, J.; Whittle, E.; Somerville, C. Science 1998,
282, 1315. (b) Whittle, E. J.; Tremblay, A. E.; Buist, P. H.; Shanklin, J.
Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 14738. (c) Mukherjee, A.;
non-heme case, the Fe−O−C angle is very roughly 180° as a
7a
result of the selection rules of TSC−H
,
but the substrate must
relocate to assume a more bent angle in order to interact with the
πy*z orbital in the second electron movement step (this relocation
is not necessary for the low-spin heme case). Hence, we suggest
that a difference in EER and steric effects are the two main factors
responsible for the difference between heme and non-heme
cases.
In conclusion, theoretical calculations have predicted that
dissociation of the substrate radicals formed by hydrogen
abstraction from alkane C−H bonds by non-heme FeIVO
complexes is more favorable than the oxygen rebound and
desaturation processes. This is in contrast to the case of heme
FeIVO, where the low-spin, reorientation-free, and EER-
independent reactions lead to a minimal rebound barrier. This
theoretical prediction has been verified by experimental results
obtained by analyzing the products formed in C−H bond
activation of various substrates by synthetic non-heme FeIVO
complexes. Thus, the general view that C−H bond activation by
high-valent metal−oxo species universally occurs via the
hydrogen abstraction/oxygen rebound mechanism should be
viewed with caution.
Martinho, M.; Bominaar, E. L.; Munck, E.; Que, L., Jr. Angew. Chem., Int.
̈
Ed. 2009, 48, 1780. (d) Bigi, M. A.; Reed, S. A.; White, M. C. Nat. Chem.
2011, 3, 216.
(9) (a) Usharani, D.; Janardanan, D.; Shaik, S. J. Am. Chem. Soc. 2011,
133, 176. (b) Janardanan, D.; Usharani, D.; Chen, H.; Shaik, S. J. Phys.
Chem. Lett. 2011, 2, 2610.
(10) Abbreviations: Me,HPytacn = 1-(2′-pyridylmethyl)-4,7-dimethyl-
1,4,7-triazacyclononane; Bn-TPEN = N-benzyl-N,N′,N′-tris(2-pyridyl-
methyl)ethane-1,2-diamine; N4Py = N,N-bis(2-pyridylmethyl)-N-
bis(2-pyridyl)methylamine.
(11) Company, A.; Prat, I.; Frisch, J. R.; Mas-Balleste,
́
R.; Guell, M.;
̈
Juhasz, G.; Ribas, X.; Munck, E.; Luis, J. M.; Que, L., Jr.; Costas, M.
́
̈
Chem.Eur. J. 2011, 17, 1622.
(12) Klinker, E. J. Ph.D. Thesis, University of Minnesota, Minneapolis,
MN, 2007.
(13) Wu, X.; Seo, M. S.; Davis, K. M.; Lee, Y.-M.; Chen, J.; Cho, K.-B.;
Pushkar, Y. N.; Nam, W. J. Am. Chem. Soc. 2011, 133, 20088.
(14) Kohn, W.; Sham, L. J. Phys. Rev. 1965, 140, A1133.
(15) (a) Ho, J.; Klamt, A.; Coote, M. L. J. Phys. Chem. A 2010, 114,
13442. (b) Cho, K.-B.; Chen, H.; Janardanan, D.; de Visser, S. P.; Shaik,
S.; Nam, W. Chem. Commun. 2012, 48, 2189.
(16) Ye, S.; Neese, F. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 1228.
(17) Shaik, S.; Chen, H.; Janardanan, D. Nat. Chem. 2011, 3, 19.
(18) Cho, K.-B.; Kim, E. J.; Seo, M. S.; Shaik, S.; Nam, W. Chem.Eur.
J. 2012, 18, 10444.
ASSOCIATED CONTENT
* Supporting Information
Theoretical and experimental methods and additional data. This
material is available free of charge via the Internet at http://pubs.
acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
sason@yfaat.ch.huji.ac.il; wwnam@ewha.ac.kr
■
(19) Evans, D. F.; Jakubovic, D. A. J. Chem. Soc., Dalton Trans. 1988,
2927. The measured magnetic moment was 5.7μB, implying an S = 5/2
state.
Notes
(20) The reactions of 2 with DHA and CHD reached completion
within several seconds. In addition, the observed rate constant for the
reaction of 2 with EB was 2.2 × 10−3 s−1 (Figure S3).
(21) Cho, K.-B.; Shaik, S.; Nam, W. J. Phys. Chem. Lett. 2012, 3, 2851.
(22) Groves, J. T.; Nemo, T. E. J. Am. Chem. Soc. 1983, 105, 6243.
(23) Lai, W.; Chen, H.; Cohen, S.; Shaik, S. J. Phys. Chem. Lett. 2011, 2,
2229.
The authors declare no competing financial interest.
§K.-B.C. and X.W. contributed equally.
ACKNOWLEDGMENTS
■
The research at EWU was supported by NRF/MEST of Korea
through CRI, GRL (2010-00353), and WCU (R31-2008-000-
10010-0) (to W.N.). S.S. acknowledges Israel Science
Foundation Grant ISF 53/09 for financial support.
D
dx.doi.org/10.1021/ja308290r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX