Proton reduction catalysis by manganese vinylidene and allenylidene complexes

Article abstract of DOI:10.1016/j.jorganchem.2007.01.055

The present paper reports the unprecedented observation of a catalytic electrochemical proton reduction based on metallocumulene complexes. Manganese phenylvinylidene (η⁵-C₅H₅)(CO)(PPh₃)Mn{double bond, long}C{double bond, long}C(H)Ph (1) and diphenylallenylidene (η⁵-C₅H₅)(CO)₂Mn{double bond, long}C{double bond, long}C{double bond, long}CPh₂ (3) are shown to catalyze the reduction of protons from HBF₄ into dihydrogen in CH₂Cl₂ or CH₃CN media at -1.60 and -0.84 V (in CH₃CN) vs. Fc, respectively. The working potential for 3 (-0.84 V vs. Fc in CH₃CN) is the lowest reported to date for protonic acids reduction in non-aqueous media. The similar catalytic cycles disclosed here include the protonation of 1, 3 into the carbyne cations [(η⁵-C₅H₅)(CO)(PPh₃)Mn{triple bond, long}C-CH₂Ph]BF₄ ([2]BF₄), [(η⁵-C₅H₅)(CO)₂Mn{triple bond, long}C-CH{double bond, long}CPh₂]BF₄ ([4]BF₄) followed by their reduction to the corresponding 19-electron radicals 2^{{radical dot}}, 4^{{radical dot}}, respectively. Both carbyne radicals undergo a rapid homolytic cleavage of the C_β-H bond generating an H-radical producing molecular hydrogen with concomitant recovery of the neutral metallocumulenes thereby completing a catalytic cycle.

Full text of DOI:10.1016/j.jorganchem.2007.01.055

Journal of Organometallic Chemistry 692 (2007) 3207–3211
www.elsevier.com/locate/jorganchem
Proton reduction catalysis by manganese vinylidene
and allenylidene complexes
a
a
a
a
Dmitry A. Valyaev , Mikhail G. Peterleitner , Oleg V. Semeikin , Kamil I. Utegenov ,
a,
b
b
b
Nikolai A. Ustynyuk *, Alix Sournia-Saquet , Noel Lugan , Guy Lavigne
¨
a
A.N. Nesmeyanov Institute of Organoelement Compounds (INEOS), Russian Academy of Sciences, 28 Vavilov Street, 119991 Moscow, Russia
b
Laboratoire de Chimie de Coordination du CNRS, 205 route de Narbonne, 31077 Toulouse Cedex 4, France
Received 18 December 2006; received in revised form 30 January 2007; accepted 30 January 2007
Available online 8 February 2007
Abstract
The present paper reports the unprecedented observation of a catalytic electrochemical proton reduction based on metallocumulene
complexes. Manganese phenylvinylidene (g5-C5H5)(CO)(PPh3)Mn@C@C(H)Ph (1) and diphenylallenylidene (g5-C5H5)(CO)2-
Mn@C@C@CPh2 (3) are shown to catalyze the reduction of protons from HBF4 into dihydrogen in CH2Cl2 or CH3CN media at
ꢀ1.60 and ꢀ0.84 V (in CH3CN) vs. Fc, respectively. The working potential for 3 (ꢀ0.84 V vs. Fc in CH3CN) is the lowest reported
to date for protonic acids reduction in non-aqueous media. The similar catalytic cycles disclosed here include the protonation of 1, 3
into the carbyne cations [(g5-C5H5)(CO)(PPh3)Mn„C–CH2Ph]BF4 ([2]BF4), [(g5-C5H5)(CO)2Mn„C–CH@CPh2]BF4 ([4]BF4) followed
by their reduction to the corresponding 19-electron radicals 2Å, 4Å, respectively. Both carbyne radicals undergo a rapid homolytic cleavage
of the Cb–H bond generating an H-radical producing molecular hydrogen with concomitant recovery of the neutral metallocumulenes
thereby completing a catalytic cycle.
ꢀ 2007 Elsevier B.V. All rights reserved.
Keywords: Proton reduction catalysis; Vinylidene complexes; Allenylidene complexes; Carbyne complexes; Cyclic voltammetry
1. Introduction
acterized by an excessive overvoltage for hydrogen
production.
Hydrogen production by water electrolysis is an impor-
tant challenge of both fundamental and economical signif-
icance in the quest for alternative energy sources. Though
Pt-based catalysts are known to achieve such a reaction,
their replacement by cheaper homogeneous transition
metal catalysts is becoming highly desirable [1]. To date,
only hydrogenases effectively catalyze this process with
TOF up to 103–104 molecules H2 sꢀ1 per site [2]. However,
the known biomimetic models for the active site of [Fe]-
only hydrogenases based on iron dithiolate clusters
(l-SCH2XCH2S-l)Fe2(CO)6 (X = CH2 [3], O [4], NR [5])
are much less active than the native enzymes and are char-
To date, all proposed catalytic cycles for electrochemical
proton reduction involve the formation of transition metal
hydrido intermediates and their subsequent transformation
resulting in dihydrogen elimination (Scheme 1). Depending
on the catalyst structure such processes can proceed via
various catalytic pathways differing in the sequences of
the electrochemical and chemical elemental steps.
For complexes containing polydentate nitrogen ligands
[6] and [Fe]-only hydrogenase model compounds (l-
SCH2XCH2S-l)Fe2(CO)6 (X = CH2, O) [3b,3c,3d,4] of
low basicity the reduction step (c) producing anionic com-
plexes precedes the protonation step (d) that affords termi-
nal hydride species [M]H. The reverse order – protonation
(a) preceding reduction (b) – was encountered for [1.1]fer-
rocenophane [7] and the anionic iron complex [Fe(CO)2
(PMe3)(l-S(CH2)3S-l)Fe(CO)2CN]ꢀ [3a]. In the latter case
*
Corresponding author. Tel.: +7 495 135 5064; fax: +7 495 135 5085.
E-mail address: ustynyuk@ineos.ac.ru (N.A. Ustynyuk).
0022-328X/$ - see front matter ꢀ 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2007.01.055

3208
D.A. Valyaev et al. / Journal of Organometallic Chemistry 692 (2007) 3207–3211
H+
(a)
lytic reactions for 19-e vinylidenes affording 18-e r-alkynyl
[M]
+ e
(g)
products are also known for Co, Rh, and Ru complexes
[10]. As a rule such homolytic Cb–H bond cleavage pro-
cesses are fast on the CV time scale and are supposed to
be favored by the instability of the reactive 19-e metal
moiety.
(c)
+ e
[M]H+
[M]+
(e)
H2
[M]
H+
(d)
Considering that complex 2+ can be restored from 1
upon simple protonation we reasoned that the stepwise
transformation 1 ! 2+ ! 2Å ! 1 (Scheme 2) might emerge
as a viable catalytic cycle for hydrogen production via elec-
trochemical proton reduction.
(b)
H2
+ e
[M]H
(f)
H+
Scheme 1.
This prompted us to investigate the electrochemical
behavior of complex 2+ in dichloromethane and acetoni-
trile solutions in the presence of acid HBF4. In CH2Cl2
the CV of 2+ shows the irreversible one-electron reduction
peak B at ꢀ1.78 V vs. Fc/Fc+ (Fig. 1b) and the oxidation
peak A of the vinylidene complex 1 at ꢀ0.06 V vs. Fc/
Fc+ is observed on the reverse scanning of potential thus
indicating a fast rate for the Cb–H bond homolysis. In ace-
tonitrile the reduction of 2+ proceeds in the same manner
at ꢀ1.60 V vs. Fc/Fc+ but the reduction peak of the acid
proton is observed at ꢀ2.13 V vs. Fc/Fc+ probably due
to partial deprotonation of 2+ with this more basic solvent.
Upon addition of acid (HBF4 Æ Et2O), the catalytic cur-
rent in the reduction peak B of 2+ appeared both in
CH2Cl2 and CH3CN media together with the increasing
intensity of the proton reduction peak C (Fig. 2 for
CH3CN). The oxidation peak of 1 disappears in acidic
solution due to the fast protonation of 1 into 2+. The
increase of acidity induces a cathodic shift of reduction
potentials for 2+ and proton (for 5 equivalents of HBF4
in CH3CN for ꢀ0.07 and 0.15 V, respectively). Such obser-
vations provide reliable evidence for proton reduction
catalysis [3–6]. We can thus conclude that the catalytic pro-
cess follows the CEC (chemical–electrochemical–chemical)
pathway (Scheme 2).
the protonation was found to occur at the Fe–Fe bond to
form a bridging l2-H framework. Some remarkable biomi-
metic catalysts with azadithiolate bridge (l-SCH2N(R)-
CH2S-l)Fe2(CO)6 (R = para-CH2C6H4Br, 2-furyl, CH2-
CH2OCH3, t-Bu, etc.) [5] undergo protonation at the basic
nitrogen atom first, facilitating the subsequent reduction
step. Further protonation at the Fe–Fe bond and transfer
of the second electron causes dihydrogen elimination.
The involvement of cationic intermediates decreases the
working potential for the catalytic hydrogen production
exhibiting the lowest overvoltage values reported so far
[5]. The crucial H2 production step can proceed via the for-
mation of the dihydrogen ligand and its further substitu-
tion (heterolytic way (f), possible for both mono- and
binuclear catalysts [3–6]) or via binuclear reductive elimina-
tion (homolytic way (e), restricted to binuclear models like
[1.1]ferrocenophane [7]). Most of catalytic systems consid-
ered above function only in non-aqueous media with high
working potential for hydrogen production and includes
the intermediate formation of oxygen-sensitive carbonyl-
metallate-anions and metal hydrides, thus remaining far
from the practical applicability.
We are now proposing a new catalytic system for elec-
trochemical proton reduction based on metallocumulene
complexes involving the formation and subsequent activa-
tion of a C–H bond rather than an M–H bond. This new
approach is illustrated for the manganese phenylvinylidene
(g5-C5H5)(CO)(PPh3)Mn@C@C(H)Ph (1) and diphenylall-
enylidene (g5-C5H5)(CO)2Mn@C@C@CPh2 (3) complexes.
Further evidence for the activity of 1 in catalytic hydro-
gen production was obtained by preparative electrolysis of
CH2Cl2 solutions of 1 (2 mM) and HBF4 (10 equiv,
20 mM) at ꢀ1.75 V vs. Fc/Fc+ on Hg-electrode. The initial
rate of the electrolysis of [2]BF4 becomes ꢁ6 times higher
2. Results and discussion
+
+ H+
Mn
Mn
We have reported earlier [8] that the reduction of the
cationic manganese carbyne complex [(g5-C5H5)(CO)-
(PPh3)Mn„C–CH2Ph]BF4 (2+) leads quantitatively to
the neutral vinylidene complex (g5-C5H5)(CO)(PPh3)-
Mn@C@C(H)Ph (1) with concomitant evolution of dihy-
drogen. The reaction can be understood in terms of a
homolytic cleavage of the Cb–H bond in 19-electron (19-
e) carbyne complex [(g5-C5H5)(CO)(PPh3)Mn„C–CH2Ph]
(2) generating an H-radical, which undergoes rapid dimer-
ization into H2. Few examples of related C–H bond activa-
tion processes in 19-e Mn [8a] and Re [9] radicals yielding
hydrogen have been previously reported. Related homo-
C
OC
Ph3P
C
Ph
OC
Ph3P
C
H
CH2Ph
2+
18 e
18 e
1
- 1.60 V
vs. Fc
+ e
.
H
.
Mn
C
OC
very fast
CH2Ph
Ph3P
.
2
19 e
Scheme 2.

D.A. Valyaev et al. / Journal of Organometallic Chemistry 692 (2007) 3207–3211
3209
in CH3CN) is within the range of those found for most
[Fe]-only biomimetic models (ꢀ1.4 to ꢀ1.7 V vs. Fc/Fc+)
[3–5].
The formation of hydrogen atoms followed by their
recombination into dihydrogen constitutes a rather uncom-
mon sequence for hydrogen production. Though the elim-
ination of H might be regarded as thermodynamically
unfavorable process we tentatively propose that reforma-
tion of the stable 18-e metal atom configuration constitutes
the driving force of the C–H homolysis and provides the
appropriate energy balance. The concerted pathway of
dihydrogen formation directly from transition state con-
sisting of two molecules of 2 should also be considered.
The nature of ‘‘real’’ intermediate hydrogen species
remains unclear and further detailed studies are to be done
to gain mechanistic insight into the factors governing these
specific radical processes.
In order to expand the scope of such catalysts, we inves-
tigated in parallel a closely related cycle based on the man-
ganese diphenylallenylidene complex 3 for which a more
positive reduction potential of the diphenylvinylcarbyne
4+ could be expected (Scheme 3) [12].
CV of the diphenylallenylidene complex 3 in CH2Cl2
solution shows two one-electron oxidation steps: the first
one, reversible at +0.35 V vs. Fc/Fc+ (Fig. 3a, peak D1)
and the second one, irreversible at +1.19 V (not shown).
The same pattern is observed for its reduction occurring
at ꢀ1.55 V (Fig. 3a, peaks D2, reversible) and at ꢀ2.20 V
(irreversible, not shown). The vinylcarbyne cation 4+
undergoes the irreversible one-electron reduction (Fig. 3b,
peak E) at ꢀ0.84 V vs. Fc/Fc+ (CH3CN) and ꢀ0.93 V
(CH2Cl2), and all representative peaks of neutral allenylid-
ene 3 are observed thus the Cb–H bond activation process
in 4 (Scheme 3) is also fast on the CV time scale.
Fig. 1. Cyclic voltammograms of complexes 1 (a) and 2+ (b). (GC-
electrode, CH2Cl2, 0.1 M Bu4NPF6, 1 · 10ꢀ3 M, 0.2 V sꢀ1, potentials vs.
Fc/Fc+.)
CVs in acetonitrile in the presence of acid also reveal
characteristic catalytic currents in the reduction peak of
vinylcarbyne 4+ (Fig. 4, peak E, ꢀ0.84 V vs. Fc/Fc+) indic-
ative of the catalytic proton reduction process, as for 1.
The working potential for catalytic proton reduction for
3 is ꢁ0.7 V more positive than for 1 but catalytic currents
in the case of 3 are generally by 30–50% lower thus reveal-
Ph
+ H+
+
Fig. 2. Cyclic voltammogram of complex 2+ in the presence of HBF4.
(GC-electrode, CH3CN, 0.1 M Bu4NPF6, 2 · 10ꢀ3 M, 0.2 V sꢀ1, poten-
tials vs. Fc/Fc+.)
Mn
OC
Mn
C
C
OC
C
OC
Ph
C
Ph
C
C
OC
H
18 e
Ph
4+
3
18 e
after addition of HBF4 and gradually slows down as the
acid concentration decreases. In such bulk experiments
hydrogen bubbles can be clearly seen. During 1 h of elec-
trolysis about 5 F per mol of 1 were passed through the
solution, corresponding to ꢁ5 turnovers. The presence of
[2]BF4 in the concentrated solution after the electrolysis
was detected by IR-spectroscopy. The working potential
for catalytic proton reduction with 1 (ꢀ1.60 V vs. Fc/Fc+
- 0.84 V
vs. Fc
+ e
.
Ph
H
.
Mn
C
fast
C
OC
Ph
C
OC
H
19 e
.
4
Scheme 3.

3210
D.A. Valyaev et al. / Journal of Organometallic Chemistry 692 (2007) 3207–3211
Fig. 5. Cyclic voltammogram of complex 4+ at high scan rates. (Pt-
electrode, CH2Cl2, 0.1 M Bu4NPF6, 1 · 10ꢀ3 M, potentials vs. Fc/Fc+.)
The reduction of the carbyne cations 2+ and 4+ under
CV conditions at higher scan rates revealed some differ-
ences in the rate of homolytic Cb–H bond cleavage in 19-
e radicals. The reduction peak for 2+ is completely irrevers-
ible even at 100 V sꢀ1 so, the life-time of 2Å is very short and
the Cb–H cleavage proceeds at high rate. At the same time
the Ia/Ic ratio for the reduction peak of 4+ becomes ca. 50%
at 10 V sꢀ1 and 100% at 100 V sꢀ1 (Fig. 5) illustrating the
greater life-time of 4Å compared to 2Å.
We propose that the greater stability of 4Å is caused by
the lower electron density on the metal atom and the delo-
calization of the unpaired electron at the Cc atom and phe-
nyl substituents of the ligand. This is probably the main
reason for the lower activity of 3 in proton reduction
catalysis.
Fig. 3. Cyclic voltammograms of complexes 3 (a) and 4+ (b). (GC-
electrode, CH2Cl2, 0.1 M Bu4NPF6, 1 · 10ꢀ3 M, 0.2 V sꢀ1, potentials vs.
Fc/Fc+.)
3. Concluding remarks
ing a reduced activity of this catalyst. However, the com-
plex 3 represents the lowest working potential (ꢀ0.84 V
vs. Fc/Fc+ in CH3CN) for reduction of protonic acids in
non-aqueous media catalyzed by homogeneous transition
metal complexes (the best example reported so far was
ꢀ1.13 V vs. Fc/Fc+ for [(l-SCH2)2N(2-C4H3O)]Fe2(CO)6
in CH3CN [5c]).
In summary, we have shown for the first time that pro-
ton reduction catalysis can be achieved with metallocumu-
lene complexes. The catalytic cycle includes the
protonation of manganese complexes 1, 3 into cationic car-
bynes 2+, 4+ and their further reduction to the correspond-
ing 19-e radicals 2Å, 4Å, respectively. Hydrogen elimination
leads to regeneration of the stable neutral 18-e complexes
1 and 3 which can be easily protonated again. The tendency
of the metal to achieve an 18-e configuration appears as the
main driving force in such homolytic processes and is typ-
ical for odd-electron r,p- and p-complexes [8a,11]. The
presence of PPh3 ligand in 1 increases the stability of the
active species but shifts negatively the working potential.
The advantages of the catalytic systems presented in
Schemes 2 and 3 are the cationic nature of the key interme-
diates 2+ and 4+ facilitating the reduction without forma-
tion of unstable carbonylmetallate or metal hydride
species. It is noteworthy that all factors influencing proton
reduction catalysis – basicity and stability of the catalytic
center, the reduction potential of cationic intermediate –
can be tuned by varying the ligand set giving a good oppor-
tunity for the design of new effective and economically via-
ble catalysts.
Fig. 4. Cyclic voltammogram of complex 4+ in the presence of HBF4.
(GC-electrode, CH3CN, 0.1 Bu4NPF6, 2 · 10ꢀ3 M, 0.2 V sꢀ1, potentials
vs. Fc/Fc+.)

D.A. Valyaev et al. / Journal of Organometallic Chemistry 692 (2007) 3207–3211
3211
4. Experimental
INTAS for YS Fellowship No. 05-109-4753 and Du Pont
for personal Ph.D. scholarship.
All manipulations were carried out under purified Ar
atmosphere using standard Schlenk techniques. Solvents
grade CH2Cl2 and CH3CN were distilled over CaH2 prior
to use. Bu4NPF6 and HBF4 · Et2O was purchased from
Aldrich Chemical Co. Vinylidene (g5-C5H5)(CO)(PPh3)-
Mn@C@C(H)Ph and allenylidene (g5-C5H5)(CO)2Mn@
C@C@CPh2 compounds and their protonated carbyne
derivatives were prepared by published methods [8b,12].
Cyclic voltammetry measurements were carried out
using a Autolab PGSTAT100 instrument controlled by
GPES 4.09 software. Experiments were performed under
Ar atmosphere in CH2Cl2 or CH3CN solutions using
three-electrode cell consisted of a glassy carbon working
electrode (d = 3 mm), a platinum wire (S = 1 cm2) as coun-
ter electrode, and a SCE electrode as a reference. For high
scan rate experiments a platinum (d ꢂ 0.5 mm) working
electrode was used. 0.1 M solution of Bu4NPF6 was used
as the supporting electrolyte. The sample concentrations
used were 1 · 10ꢀ3 M (CH2Cl2) and 2 · 10ꢀ3 M (CH3CN).
All peak potentials are given relative to the ferrocene/ferr-
ocenium couple (E = +0.46 V (CH2Cl2) or E = +0.40 V
(CH3CN) vs. SCE). The number of electrons consumed
was estimated by comparison of the currents of the peaks
observed with those of the one-electron Fc/Fc+ or (g5-
C5Me5)2Fe/(g5-C5Me5)2Fe+ couples at the same concen-
trations. Before each measurement the working electrode
was polished with a polishing machine (Presi P230). For
the proton reduction experiments the diluted solution of
HBF4 · Et2O was prepared (70 lL of HBF4 · Et2O in
1 mL of the appropriate solvent) and added via syringe
directly to the electrochemical cell. The values of the cata-
lytic current in the proton reduction experiments were
strongly dependent on the electrode surface quality (the
values of the catalytic current obtained on the freshly pol-
ished electrode are 2–3 times greater than those for the next
repetitive scan) thus all cyclic voltammograms were repro-
duced several times affording the same results.
References
[1] (a) U. Koelle, New J. Chem. 16 (1992) 157–169;
(b) J.-F. Capon, F. Gloaguen, P. Schollhammer, J. Talarmin, Coord.
Chem. Rev. 249 (2005) 1664–1676;
(c) V. Artero, M. Fontecave, Coord. Chem. Rev. 249 (2005) 1518–
1535.
[2] M.W.W. Adams, Biochim. Biophys. Acta 1020 (1990) 115–145.
[3] (a) F. Gloaguen, J.D. Lawrence, T.B. Rauchfuss, M. Benard, M.M.
Rohmer, Inorg. Chem. 41 (2002) 6573–6582;
(b) S.J. Borg, T. Behrsing, S.P. Best, M. Razavet, X. Liu, C.J.
Pickett, J. Am. Chem. Soc. 126 (2004) 16988–16999;
(c) D. Chong, I.P. Georgakaki, R. Mejia-Rodriguez, J. Sanabria-
Chinchilla, M.P. Soriaga, M.Y. Darensbourg, J. Chem. Soc., Dalton
Trans. (2003) 4158–4163;
(d) C. Tard, X. Liu, S.K. Ibrahim, M. Bruschi, L. De Gioia, S.C.
Davies, X. Yang, L.-S. Wang, G. Sawers, C.J. Pickett, Nature 433
(2005) 610–613.
[4] L.-C. Song, Z.-Y. Yang, H.-Z. Bian, Y. Liu, H.-T. Wang, X.-F. Liu,
Q.-M. Hu, Organometallics 24 (2005) 6126–6135.
[5] (a) S. Ott, M. Kritikos, B. Akermark, L. Sun, R. Lomoth, Angew.
Chem., Int. Ed. 43 (2004) 1006–1009;
(b) H. Li, T.B. Rauchfuss, J. Am. Chem. Soc. 124 (2002) 726–727;
(c) S. Jiang, J. Liu, L. Sun, Inorg. Chem. Commun. 9 (2006) 290–292.
[6] (a) I. Bhugun, D. Lexa, J.-M. Saveant, J. Am. Chem. Soc. 118 (1996)
´
3982–3983;
´
(b) V. Grass, D. Lexa, J.-M. Saveant, J. Am. Chem. Soc. 119 (1997)
7526–7532;
(c) M. Razavet, V. Artero, M. Fontecave, Inorg. Chem. 44 (2006)
4786–4795;
(d) J.-P. Collin, A. Jouaiti, J.-P. Sauvage, Inorg. Chem. 27 (1988)
1986–1990;
(e) P. Connolly, J.H. Espenson, Inorg. Chem. 25 (1986) 2684–2688;
R.M. Kellett, T.G. Spiro, Inorg. Chem. 24 (1985) 2373–2377.
[7] U.T. Mueller–Westerhoff, A. Nazzal, J. Am. Chem. Soc. 106 (1984)
5381–5382.
[8] (a) D.A. Valyaev, O.V. Semeikin, N.A. Ustynyuk, Coord. Chem.
Rev. 248 (2004) 1679–1692;
(b) L.N. Novikova, M.G. Peterleitner, K.A. Sevumyan, O.V. Seme-
ikin, D.A. Valyaev, N.A. Ustynyuk, Appl. Organomet. Chem. 16
(2002) 530–536.
[9] M.A.N.D.A. Lemos, A.J.L. Pombeiro, J. Organomet. Chem. 356
(1988) C79–C82.
Controlled potential electrolysis was carried out in
20 mL cell in CH2Cl2-Bu4NPF6 solution using unstirred
Hg pool electrode (the surface area ca. 10 cm2) at ꢀ1.3 V
vs. SCE (ꢀ1.75 V vs. Fc). The typical concentration of cat-
alyst and acid were 2 mM and 20 mM, respectively.
[10] (a) C. Bianchini, P. Innocenti, A. Meli, M. Peruzzini, F. Zanobini, P.
Zanello, Organometallics 9 (1990) 2514–2522;
(b) C. Bianchini, A. Meli, M. Peruzzini, F. Zanobini, P. Zanello,
Organometallics 9 (1990) 241–250;
(c) C.J. Adams, S.J.A. Pope, Inorg. Chem. 43 (2004) 3492–3499.
[11] (a) D. Astruc, Electron Transfer and Radical Processes in Transition-
Metal Chemistry, VCH, 1995 [Chapter 5];
(b) I.V. Shchirina-Eingorn, L.N. Novikova, N.A. Ustynyuk, L.I.
Denisovich, M.G. Peterleitner, Russ. Chem. Bull. 45 (1996) 1608;
(c) O.V. Gusev, L.N. Morozova, T.A. Peganova, M.G. Peterleitner,
S.M. Peregudova, L.I. Denisovich, P.V. Petrovskii,Yu. F. Oprunenko,
N.A. Ustynyuk, J. Organomet. Chem. 493 (1995) 181–187.
[12] N.E. Kolobova, L.L. Ivanov, O.S. Zhvanko, O.M. Khitrova, A.S.
Batsanov, Yu. T. Struchkov, J. Organomet. Chem. 262 (1984) 39–47.
Acknowledgements
This work was supported by Russian Foundation for
Basic Research (Grant No. 02-03-33180, 05-03-32720)
and by the CNRS. Dmitry A. Valyaev is also grateful to