Fig. 2 Gas chromatogram of the head space after 3 h in anaerobic (a)
and aerobic conditions (b) for a DCE solution containing 5 mM
DMFc and 10 mM HTB.
those obtained in biphasic systems,9 whereas some small
amount of H2 can still be observed in aerobic conditions
(Fig. 2b) suggesting that proton and oxygen reductions do
compete, but that hydrogen production cannot account for the
low H2O2 generation yield. It must be stressed that the
solution turns green within minutes in the presence of oxygen
whereas it takes a couple of hours in anaerobic conditions.
Previous studies at liquid–liquid interfaces suggested that in
both cases the initial step is the protonation of DMFc leading
to the formation of DMFcH+. The protonation of ferrocene
was reported to result in a tilting of the rings thus facilitating
the binding of O2 at the Fe atom directly, which is then
followed by oxygen reduction.2 Considering that insertion of
triplet O2 is spin-forbidden, Fomin proposed the formation of
an hydrogen-bond intermediate, O2 being sandwiched between
two protonated ferrocenes followed by the generation of H2O2
by concerted breakdown of two Fe–H and formation of two
H–O bonds.10 The present DFT computations do not support
the hypothesis that triplet molecular oxygen O2 coordinates to
the iron atom through a spin-forbidden mechanism or inserts
into the Fe–H bond. Instead, the present computations show
that O2 can approach the activated hydride directly via a
delocalized triplet (diradical) transition state [DMFcꢂꢂꢂHꢂꢂꢂOO]+
(Scheme 1) with an activation barrier of 15 kcal molꢀ1 in the
gas phase and 14.6 kcal molꢀ1 in the solvent. This process
leads to the formation of decamethylferrocenium DMFc+ and
a hydrogen peroxyl radical. The generation of H2O2 from the
latter is then expected to proceed rapidly. Note that other
possible mechanistic routes occurring via either a superoxoiron
[DMFc–O2] (i.e. protonation last, spin-forbidden) or a super-
oxide intermediate [DMFc–O2H]+ (i.e. insertion into the
Fe–H bond) were found to have considerably higher activation
energy barriers than the mechanism proposed above (see
details in Supplementary Dataw).
Scheme 1 Proposed mechanism for the reduction of O2 into HO2 by
DMFc. Gas phase zero-point corrected energies computed at the
BP86/DZP level are given in red. COSMO corrected energies
(e = 10.36) computed at the same level are given in blue. The unpaired
spin density is displayed in green. The spin quantum number (i.e. s) is
given as a superscript for each compound. Further details can be
found in the Supplementary Data.w
that protonated metallocenes are highly reactive in reduction
reactions useful to energy research such as hydrogen
production, oxygen or even carbon dioxide reduction.
This work was supported by EPFL, the Swiss National
Science Foundation (FNRS 200020-116588) and the COST
Action, D36/007/06. C. C. acknowledges the Sandoz Family
Foundation. I. H. also gratefully acknowledges the Scientific
and Technological Research Council of Turkey (TUBITAK)
under the 2212-PhD Scholarship Programme. Z. S. is also
grateful to the Grant Agency of the Czech Republic (grant no.
203/07/1257). We also thank Prof. C. Amatore (ENS, Paris)
for helpful discussions.
Notes and references
y HTB was prepared by shaking x mM LiTB and 10 mM HCl in the
water phase with pure DCE for 1 h. The concentration of HTB
extracted to DCE was approximately x mM, as determined by the
pH change of the water phase before and after shaking.
1 R. Prins and A. G. T. G. Kortbeek, J. Organomet. Chem., 1971, 33,
C33–C34.
2 T. E. Bitterwolf and A. C. Ling, J. Organomet. Chem., 1972, 40,
C29–C32.
3 S. Fukuzumi, S. Mochizuki and T. Tanaka, Chem. Lett., 1989,
27–30.
4 S. Fukuzumi, S. Mochizuki and T. Tanaka, Inorg. Chem., 1989, 28,
2459–2465.
5 S. Fukuzumi, K. Okamoto, C. P. Gros and R. Guilard, J. Am.
Chem. Soc., 2004, 126, 10441–10449.
6 V. J. Cunnane, G. Geblewicz and D. J. Schiffrin,
Electrochim. Acta, 1995, 40, 3005–3014.
7 B. Su, R. P. Nia, F. Li, M. Hojeij, M. Prudent, C. Corminboeuf,
Z. Samec and H. H. Girault, Angew. Chem., Int. Ed., 2008, 47,
4675–4678.
In the case of proton reduction, hydrogen evolution can
proceed via a proton attack on DMFcH+ or by a bimolecular
route involving two DMFcH+. The present experimental
results confirm that hydrogen production is a much slower
process than oxygen reduction, which could be explained by
the bimolecular pathway.
In conclusion, oxygen/proton reductions by DMFc in
1,2-dichloroethane have been observed experimentally. DFT
computations suggest a reaction pathway involving protonated
DMFc, DMFcH+, as an intermediate species, which reacts
with oxygen to produce hydrogen peroxide. This study shows
8 F. Li, B. Su, F. C. Salazar, R. P. Nia and H. H. Girault, Electro-
chem. Commun., 2009, 11, 473–476.
9 I. Hatay, B. Su, F. Li, R. Partovi-Nia, H. Vrubel, X. Hu, M. Ersoz
and H. H. Girault, Angew. Chem., Int. Ed., 2009, 48,
5139–5142.
10 V. M. Fomin, Russ. J. Gen. Chem., 2007, 77, 954–960.
ꢁc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 2918–2919 | 2919