Half-sandwich molybdenum(III) compounds containing diazadiene ligands and their use in the controlled radical polymerization of styrene

Article abstract of DOI:10.1016/S0022-328X(02)01878-8

The reaction of CpMoCl₂ with diazadiene ligands RN=CH-CH=NR (R₂-dad) affords the corresponding paramagnetic complexes CpMoCl₂(R₂-dad) (R = Ph, 1; p-Tol, 2; C₆H₃Prⁱ-2, 6, 3, and Prⁱ, 4). All compounds have been characterized by EPR spectroscopy and have been investigated by cyclic voltammetry. They display one-electron oxidation and reduction processes, these being reversible or irreversible depending on the nature of R. The irreversibility of the reduction wave is due to a chemical follow-up process which consists of chloride loss from the reduced product. This phenomenon is suppressed in the presence of excess chloride in solution. An X-ray structure of 3 verifies the mononuclear nature of the compound and the chelating mode of the dad ligand, the MoN₂C₂ ring being essentially planar. The N-C and C-C bond distance pattern suggests the important contribution of an enediamido Mo(V) limiting form. In the presence of 1-bromoethylbenzene, complexes 1-4 catalyze the controlled/'living' radical polymerization of styrene. Complex 3 also leads to a controlled/'living' radical polymerization of styrene in the presence of AIBN (α,α-azoisobutyronitrile) as a radical generator. Therefore, this is the second example of a compound which is capable of controlling the styrene radical polymerization under both ATRP and SFRP conditions.

Full text of DOI:10.1016/S0022-328X(02)01878-8

Journal of Organometallic Chemistry 663 (2002) 269Á276
/
www.elsevier.com/locate/jorganchem
Half-sandwich molybdenum(III) compounds containing diazadiene
ligands and their use in the controlled radical polymerization of
styrene
Francois Stoffelbach, Rinaldo Poli ꢀ, Philippe Richard
¸
´ ´ ´
Laboratoire de Synthe`se et d’Electrosynthe`se Organometalliques, Faculte de Science Gabriel, Universite de Bourgogne, 6 Boulevard Gabriel, 21000
Dijon, France
Received 4 June 2002; accepted 4 September 2002
Dedicated to Pascual Royo on the occasion of his 65th birthday
Abstract
The reaction of CpMoCl2 with diazadiene ligands RNÄ
/
CHÃ
/
CHÄ
/
NR (R2Ã
/
dad) affords the corresponding paramagnetic
Ph, 1; p-Tol, 2; C6H3Pri2-2; 6; 3; and Pri, 4). All compounds have been characterized by EPR
complexes CpMoCl2(R2Ãdad) (Rꢁ
/
/
spectroscopy and have been investigated by cyclic voltammetry. They display one-electron oxidation and reduction processes, these
being reversible or irreversible depending on the nature of R. The irreversibility of the reduction wave is due to a chemical follow-up
process which consists of chloride loss from the reduced product. This phenomenon is suppressed in the presence of excess chloride
in solution. An X-ray structure of 3 verifies the mononuclear nature of the compound and the chelating mode of the dad ligand, the
MoN2C2 ring being essentially planar. The NÃ
/
C and CÃ
/
C bond distance pattern suggests the important contribution of an
4 catalyze the controlled/‘living’ radical
enediamido Mo(V) limiting form. In the presence of 1-bromoethylbenzene, complexes 1Á
/
polymerization of styrene. Complex 3 also leads to a controlled/‘living’ radical polymerization of styrene in the presence of AIBN
(a,a-azoisobutyronitrile) as a radical generator. Therefore, this is the second example of a compound which is capable of controlling
the styrene radical polymerization under both ATRP and SFRP conditions.
# 2002 Elsevier Science B.V. All rights reserved.
Keywords: Molybdenum; Cyclopentadienyl; Diazadiene ligands; Paramagnetic organometallics; Controlled radical polymerization
1. Introduction
acidic CO, as in [CpꢀMoCl(PMe3)2(CO)]ꢂ [2], to the
very p-basic OH, as in [CpMo(OH)(PMe3)3]ꢂ [3],
through the p-neutral alkyls, as in CpMo(CH3)2(PMe3)2
[4]. The dichloride family, CpMoCl2L2, includes mem-
Half-sandwich complexes of Mo(III) having a four-
legged piano stool structure and a 17-electron config-
uration now represent a relatively large family of stable
organometallic radicals [1]. Their susceptibility to be
oxidized to 16-electron (generally spin triplet) Mo(IV)
derivatives or else to be reduced to saturated Mo(II)
derivatives depends very strongly on the electronic
properties of the ligands. Nevertheless, this family of
compounds has been found to be compatible with a
broad spectrum of ancillary ligands, from the very p-
bers where L or L2ꢁtertiary phosphine [5], phosphine
/
donors linked as side chains to the cyclopentadienyl
ligand [6], or dienes [7]. Members containing bifunc-
tional ligands such as phosphineÁ
/
oxazoline, Á
/
amide or
thioether [8,9], have also been reported more recently,
Á
/
these being the first examples featuring neutral N, O or
S donors. We have been unable for some time to extend
this family to ligands L2 with two hard nitrogen donors.
We report here the first examples of this class, obtained
using a variety of 1,4-diazadienes, RNÄ
Diazadienes have previously shown their coordinating
ability in the related half-sandwich CpꢀMCl2(RNÄCHÃ
CHÄNR) (MꢁNb, Ta) families [10Á14].
/
CHÃ
/
CHÄ
/
NR.
ꢀ Corresponding author. Tel.:
393720
E-mail address: poli@u-bourgogne.fr (R. Poli).
ꢂ
/
33-380-396881; fax:
ꢂ/33-380-
/
/
/
/
/
0022-328X/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 2 - 3 2 8 X ( 0 2 ) 0 1 8 7 8 - 8

270
F. Stoffelbach et al. / Journal of Organometallic Chemistry 663 (2002) 269Á276
/
A further interest of this work is related to our recent
from P4O10 and distilled under argon. Molecular
weights and molecular weight distributions were mea-
sured with a Gynkoteck P580 size exclusion chromato-
grapher equipped with a refractometer and 2 B Jordi
discovery [15] that complexes CpMoCl2(PMe3)2 and
CpMoCl2(dppe) are able to control the radical poly-
merization of styrene by two different mechanisms
simultaneously, i.e. atom transfer radical polymerization
(ATRP) and stable free radical polymerization (SFRP)
(Scheme 1), an unprecedented phenomenon. Both these
equilibria have the property of equilibrating the reactive
DVB columns (range 1000Á10 00 000) using THF as
/
eluent (1 ml minꢃ1). The instrument was calibrated by
using polystyrene standards. EPR measurements were
carried out at the X-band microwave frequency on a
Bruker ESP300 spectrometer. The spectrometer fre-
quency was calibrated with diphenylpicrylhydrazyl
radical (R+) with a dormant species (MÃ
/
R for SFRP,
R for ATRP), thereby lowering the free radical
concentration and consequently suppressing the bimo-
lecular termination processes. Other families of metal
XÃ
/
(DPPH, gꢁ2.0037). Cyclic voltammograms were re-
/
corded with an EG&G 362 potentiostat connected to a
Macintosh computer through MACLAB hardware/soft-
complexes such as CuX(L3) (Xꢁ
/
Cl or Br; L3ꢁtria-
/
mine, diaminopyridine, etc.) or CpRuCl2(PPh3) have
been shown to be capable of controlling radical poly-
merization by the ATRP mechanism only (see Scheme 1)
[16], while yet other families (organocobalt porphyrin
complexes, for instance [17]) can only operate by the
SFRP mechanism. Yet another novel feature of the half-
sandwich Mo(III) complexes is the transition from
‘living’ to catalyzed chain transfer (CCT) by a simple
modification of the coordination sphere. While the two
phosphine complexes mentioned above undergo ‘living’
polymerization under both ATRP and SFRP condi-
tions, the diene complex CpMoCl2(h4-C4H6) undergoes
‘living’ polymerization only under SFRP conditions and
CCT under ATRP conditions. All these features have
been rationalized on the basis of a global kinetic scheme
and simulation tests [15]. In order to broaden our
understanding of this phenomenon, we have tested our
new diazadiene (dad) complexes in the controlled radical
polymerization of styrene and the related results are also
presented here.
ware. The electrochemical cell was fitted with an AgÁ
AgCl reference electrode, a platinum disk working
electrode and platinum wire counter-electrode.
/
a
[Bu4N]PF6 (ca. 0.1 M) was used as supporting electro-
lyte in THF. All potentials are reported relative to the
ferrocene standard, which was added to each solution
and measured at the end of the experiments. The
elemental analyses were carried out by the analytical
service of LSEO with a Fisons EA 1108 instrument.
[CpMo(m-Cl)2]2 was obtained according to a previously
described synthetic procedures [18], and the N,N-
disubstituted 1,4-diaza-1,3-diene ligands (R2Ã
/
dad, with
Ph, p-Tol, C6H3Pri2-2; 6; and Pr) were prepared as
described in the literarure [19Á22]. 1-Bromopheny-
i
Rꢁ
/
/
lethane was purchased from Aldrich Chemical Co and
degassed prior to use. AIBN (JANSSEN) was recrys-
tallized twice from MeOH before use.
2.2. Synthesis of CpMoCl2(Ph2Ã
/
dad) (1)
[CpMo(m-Cl)2]2 (0.204 g, 0.44 mmol) and Ph2Ã
/dad
(0.183 g, 0.88 mmol) were suspended in 20 ml of CH2Cl2
at room temperature (r.t.). The mixture was then stirred
for several hours (ca. 20). The solution was filtered
through Celite to remove a small amount of residual
solid and concentrated under reduced pressure to ca. 4
ml. Addition of 10 ml of C5H12 gave complex 1 as a red
brown microcrystalline solid, which was washed with
2. Experimental
2.1. General
All manipulations were carried out under an atmo-
sphere of dry and oxygen-free argon with standard
Schlenk techniques. Styrene was washed with an aqu-
eous NaOH solution (10%), neutralized with water,
dried over MgSO4 and then distilled at 25 8C under
reduced pressure. Toluene and C5H12 were purified by
reflux over sodium benzophenone ketyl and distilled
under argon prior to use. Dichloromethane was purified
C5H12 (3ꢄ5 ml) and dried in vacuo. Yield: 0.2341 g
/
(60%). Anal. Calc. for C19H17Cl2MoN2 (440.20): C,
51.84; H, 3.89; N, 6.36. Found: C, 52.46; H, 4.06; N,
6.46%. EPR (CH2Cl2): gꢁ
voltammetry (THF): irreversible oxidation with Ep,a
0.01 V and reversible reduction at E1/2 1.12 V
(DEpꢁ71 mV). The ferrocene peak showed DEpꢁ76
mV.
/1.968 (aMo
ꢁ/38.7 G). Cyclic
ꢁ
/
ꢃ
/
ꢁ
/
ꢃ
/
/
/
2.3. Synthesis of CpMoCl2(p-Tol2Ã
/
dad) (2)
[CpMo(m-Cl)2]2 (0.195 g, 0.42 mmol) and p-Tol2Ã
/dad
(0.200 g, 0.85 mmol) were suspended in 8 ml of
C6H5CH3 at r.t. The mixture was then stirred for several
hours (ca. 20) at 70 8C. Dichloromethane (20 ml) was
Scheme 1.

F. Stoffelbach et al. / Journal of Organometallic Chemistry 663 (2002) 269Á
/
276
271
added at r.t. and the solution was filtered through Celite
to remove a small amount of residual solid and
evaporated to dryness under reduced pressure to give
complex 2 as a red brown microcrystalline solid. Yield:
0.276 g (70%). Anal. Calc. for C21H21Cl2MoN2 (468.26):
C, 53.87; H, 4.52; N, 5.98. Found: C, 52.92; H, 4.64; N,
plex was added to a 50 ml Schlenk tube equipped with a
stirring bar. Styrene and 1-bromophenylethane were
added to the reaction flask by a syringe. The Schlenk
tube was then immersed in an oil bath heated at 90 8C.
Aliquots were withdrawn periodically for a reaction
monitoring by GPC.
5.56%. EPR (C6H5CH3): gꢁ
Cyclic voltammetry (THF): irreversible oxidation at
Ep,a 0.00 and irreversible reduction at Ep,a 0.99
V. The reduction wave is observed at Ep,a 0.89 V in
CH2Cl2 and becomes reversible after addition of Clꢃ
/
1.969 (aMo
ꢁ
/
38.9 G).
2.7. SFRP polymerization
ꢁ
/
ꢁ
/
ꢃ
/
ꢁ
/
ꢃ
/
SFRP polymerizations were conducted as follows: the
Mo complex and AIBN (a,a-azoisobutyronitrile) were
added to a 50 ml Schlenk tube equipped with a stirring
bar. Styrene was added by a syringe. The Schlenk tube
was immersed in an oil bath heated at 100 8C. Aliquots
were withdrawn periodically for a reaction monitoring
by GPC.
while it shifts to E1/2
Section 3.
ꢁ
/
ꢃ
/
0.895 V (Ep,a
ꢁ
/
ꢃ0.93 V), see
/
2.4. Synthesis of CpMoCl2[(C6H3Pri2-2,6)2Á
dad] (3)
/
[CpMo(m-Cl)2]2 (0.3013 g, 0.65 mmol) and
(C6H3Pri2-2; 6)2Ãdad (0.506 g, 1.34 mmol) were sus-
pended in 10 ml of C6H5CH3 at r.t. The mixture was
then stirred for 4 days at 70 8C. The solution was
filtered through Celite and dried in vacuo to give an oily
residue. Addition of 5 ml of C5H12 gave complex 3 as a
2.8. X-ray analysis of compound 3
Intensity data were collected on a Nonius Kappa
CCD at 110 K. The structure was solved via a Patterson
search program and refined with full-matrix least-
squares methods based on F2
(SHELXL-97) [23] with
the aid of the WINGX program [24]. All non-hydrogen
atoms were refined with anisotropic thermal parameters.
Hydrogen atoms were included in their calculated
positions or found in the final difference Fourier maps
and refined with a riding model. The crystal data and
refinement parameters are collected in Table 1.
redÁ
with C5H12 (2ꢄ
/
brown microcrystalline solid, which was washed
5 ml) and dried in vacuo. Yield: 0.502 g
/
(70%). Anal. Calc. for C31H41Cl2MoN2 (608.52): C,
61.19; H, 6.79; N, 4.60. Found: C, 61.52; H, 6.83; N,
4.68%. EPR (C6H5CH3): gꢁ
/
1.989 (aMo
ꢁ38.0 G).
/
Recrystallization by slow cooling of a hot heptane
solution afforded crystals of 3, which were used for
the X-ray analysis. Cyclic voltammetry (THF): irrever-
sible oxidation at Ep,a
ꢁ0.33 V and irreversible reduc-
/
tion at Ep,c 1.01 V.
ꢁ
/
ꢃ
/
3. Results and discussion
2.5. Synthesis of CpMoCl2(Pri2Á
dad) (4)
/
3.1. Synthesis and EPR characterization
[CpMo(m-Cl)2]2 (0.4057 g, 0.87 mmol) and Pri2Ãdad
(0.378 g, 2.69 mmol) were suspended in 20 ml of
C6H5CH3 at r.t. The mixture was then stirred for 2
days at 70 8C. The resulting solution was evaporated to
dryness under reduced pressure and the residue was
further dried under vacuum at 70 8C to sublime the
residual Pri2Ãdad: The product was extracted with 40 ml
of CH2Cl2, the resulting solution was filtered through
Celite and then evaporated to dryness to yield a dark red
microcrystalline solid. Yield: 0.4172 g (64%). Anal. Calc.
for C13H21Cl2MoN2 (372.17): C, 41.95; H, 5.69; N, 7.53.
The synthesis of the title compounds is accomplished
by the straightforward addition of the appropriate
diazadiene ligand RNÄ
/
CHÃ
/
CHÄ
/
NR (R2Ã
/
dad; Rꢁ
/
Ph, p-Tol, C6H3Pr2i -2; 6; and Pri) to [CpMoCl2]2, see
Eq. (1). The resulting complexes have been isolated as
stable crystalline solids in good yields and fully char-
acterized. They are air sensitive in the solid state and in
solution but can be easily handled under an inert
atmosphere.
Found: C, 42.29; H, 5.80; N, 7.69%. EPR (CH2Cl2): gꢁ
1.975 (aMo 42.8 G). Cyclic voltammetry (THF): re-
versible oxidation at E1/2 90 mV) and
0.01 V (DEpꢁ
reversible reduction at E1/2 80 mV).
1.10 V (DEpꢁ
The ferrocene peak showed DEpꢁ110 mV.
/
ꢁ
/
ꢁ
/
/
ꢁ
/
ꢃ
/
/
/
2.6. ATRP polymerizations
All ATRP polymerization reactions were performed
following the same experimental procedure. The com-
ð1Þ

272
F. Stoffelbach et al. / Journal of Organometallic Chemistry 663 (2002) 269Á276
/
Table 1
Crystal data and structure refinement for compound 3
Empirical formula
M
C31H41Cl2MoN2
608.50
110(2)
Temperature (K)
˚
l (A)
0.71073
Monoclinic
P21/n
Crystal system
Space group
Unit cell dimensions
˚
a (A)
10.35810(10)
21.56210(10)
13.2046(2)
90
˚
b (A)
˚
c (A)
a (8)
b (8)
g (8)
98.0140(10)
90
3
˚
V (A )
Z
F(000)
2920.35(5)
4
1268
Fig. 1. An ORTEP view of the geometry of compound 3.
Dcalc (g cmꢃ3
)
1.384
Diffractometer
Scan type
EnrafÁNonius KappaCCD
Mixture of f rotations and v
scans
/
m (mmꢃ1
)
0.654
0.38ꢄ0.38ꢄ0.38
0.65
Crystal size (mm3)
ꢃ1
˚
sin(u)/l max (A
Index ranges
)
ꢃ135h 512, ꢃ275k 5; 27,
ꢃ165l 517
SCALEPACK
Absorption correction
RCꢁReflections collected
IRCꢁIndependent RC
IRCGTꢁIRC and [I ꢀ2s(I)]
Refinement method
17 529
6559 [R(int)ꢁ0.0231]
5762
Full-matrix least-squares on F2
6559/0/325
R1 aꢁ0.0272, wR2 bꢁ0.0624
R1 aꢁ0.0339, wR2 bꢁ0.0650
1.032
Data/restraints/parameters
R for IRCGT
R for IRC
c
Goodness-of-fit
Largest difference peak and hole 0.410 and ꢃ0.660
ꢃ3
˚
(e A
)
a
R1ꢁa (jjFojꢃjFcjj)/a jFo j.
wR2ꢁ
/
[a w(Fo2 ꢃFc2)2=[a w(Fo2)2]1=2
where
(0:0270ꢄP)2 ꢂ1:80ꢄP] where Pꢁ(Max(Fo2; 0)ꢂ2ꢄFc2)=3:
wꢁ1=[s2(Fo2)ꢂ
b
/
Goodness-of-fitꢁ
/
[a w(Fo2 ꢃFc2)2=(No ꢃNv)]1=2
:/
c
Like the previously characterized members of the
same family, these compounds exhibit sharp EPR
signals in homogeneous solutions at room temperature,
in agreement with their formulation as mononuclear
metalloradicals and with the absence of low-lying
excited states. The g parameter for all compounds is
quite close the spin-only value and the Mo hyperfine
splittings are around 40 G, the alkyl substituted member
(4) having a slightly greater coupling than the aryl
Fig. 2. A side view of the geometry of compound 3, emphasizing the
perpendicular orientation of the dad aryl substituent relative to the
coordination plane.
feature is the arrangement adopted by the five-mem-
bered ring containing the metal atom and the diazadiene
ligand. This ligand system may adopt a variety of
coordination modes and examples are known in the
literature which can be most closely described as either
A, B or C [25]. For compound 3, the five-membered ring
substituted ones (1Á
3.2. Structural characterization of compound 3
The mononuclear nature of complexes 1Á4 is verified
/
3).
is essentially planar (Rms deviation of fitted atomsꢁ
/
˚
0.07; highest deviation from the plane: 0.0842 A for
atom N1). The fit of the N2C2 plane is even better (Rms
/
in one case (compound 3) by a single crystal X-ray
analysis. Two different views of the molecule are shown
in Figs. 1 and 2. Selected bond distances and angles are
listed in Table 2. The most interesting geometrical
deviation of fitted atomsꢁ0.002), this plane having a
/
small dihedral angle with the MoN2 plane [11.42(9)8], as
can be easily appreciated from Fig. 2. This excludes the

F. Stoffelbach et al. / Journal of Organometallic Chemistry 663 (2002) 269Á
/
276
273
Table 2
description of the coordination geometry as B. Analo-
Cl distances are shorter than those
a
˚
Bond lengths (A) and angles (8) for compound 3
gously, the MoÃ
/
˚
˚
found for CpMoCl2(PMe3)2 (2.468(2) and 2.474(2) A)
4
Bond lengths
[31] and CpMoCl2(h -C4H6) (2.435(1) and 2.433(1) A)
[26], in further agreement with the assignment of a
MoÃCNT
MoÃN(1)
MoÃN(2)
MoÃCl(1)
2.035(1)
MoÃCl(2)
N(1)ÃC(6)
N(2)ÃC(7)
C(6)ÃC(7)
2.3893(4)
1.351(2)
1.353(2)
1.363(2)
2.0806(14)
2.0925(14)
2.4098(4)
higher formal oxidation state to the metal in compound
3.
Bond angles
CNTÃMoÃN(1)
CNTÃMoÃN(2)
CNTÃMoÃCl(1)
CNTÃMoÃCl(2)
Cl(1)ÃMoÃCl(2)
N(1)ÃMoÃN(2)
N(1)ÃMoÃCl(1)
118.2(2)
117.5(1)
104.9(1)
108.5(1)
84.395(16)
74.16(5)
85.02(4)
N(2)ÃMoÃCl(2)
N(1)ÃMoÃCl(2)
N(2)ÃMoÃCl(1)
C(6)ÃN(1)ÃMo
C(7)ÃN(2)ÃMo
C(6)ÃC(7)ÃN(1)
C(7)ÃC(6)ÃN(2)
83.84(4)
133.32(4)
137.53(4)
116.75(11)
116.77(11)
115.67(15)
114.81(15)
3.3. Electrochemical characterization
The electrochemical properties of compounds 1Á4
/
reflect the degree of electron donation provided by the
diazadiene ligands with respect to other ligands in
similar compounds (see Table 3). All the compounds
exhibit two one-electron processes, one reduction and
one oxidation.
a
CNTꢁcyclopentadienyl ring centroid.
The potentials associated to the one-electron oxida-
arrangement C, which is characterized by a rather severe
folding around the NÃN hinge.
tion process are more positive in the order (phosphineB
diazadieneBalkyneBdiene), whereas those associated
to the one-electron reduction process are less negative in
the order (phosphineBdiene, alkyneBdiazadiene).
/
/
/
/
/
/
While both trends agree in showing that the phosphine
ligands are the best electron donors, the inversion
between the diene and alkyne on one side and the
diazadiene on the other side is unexpected. This inver-
sion can be reconciled by the fact that the molecular
orbital involved in the redox process is not necessarily
the same for the oxidation and the reduction. The
different ligands may have a different effect on each
orbital via a different type of interaction. In fact, while
the reduction most probably involves the singly occu-
pied orbital, yielding a closed-shell 18-electron config-
uration, the oxidation process may involve the second
highest occupied MO, at least in the diazadiene case. A
few phosphine-containing oxidation products have been
isolated and shown to possess a spin triplet ground state.
For the alkyne and diene complexes, the spin state of the
oxidation products is unknown since these are unstable
complexes.
Curiously, the oxidation occurs approximately at the
same potential for the aryl substituted (1 and 2) and
alkyl substituted (4) complexes, for which an electronic
difference may be expected, whereas the potential is
significantly more positive for the aryl-substituted com-
plex 3. These trends do not have a straightforward
rationalization, but we note that the steric bulk of the
C6H3Pri2 substituents may weaken the donor properties
of the dad ligand relative to the other dad ligands,
making the metal electron-poorer and consequently
more difficult to oxidize.
It is also interesting to observe that the NÃ
C distances of the dad ligand are considerably length-
ened and shortened, respectively, when compared with
/
C and CÃ
/
those of free dad molecules such as s-trans-Cy2dad (NÃ
/
˚
˚
C: 1.457(2) A) and also with those of
complexes to which the bonding mode A can be clearly
C: 1.258(2) A; CÃ
/
˚
˚
C: 1.40Á1.46 A) [25].
assigned (NÃ
/
C: 1.26Á
/
1.30 A; CÃ
/
/
The observed bonding parameters are more consistent
with a bonding description where the arrangement B
plays a significant role. This limiting form corresponds
to an enediamido complex of Mo(V). There is an
interesting similarity between this situation and that of
the analogous diene systems [26Á28] for which con-
/
tribution of an h4-diene Mo(III) formulation as well as
an enediyl Mo(V) formulation is invoked, though the
former seems to predominate [27]. Another interesting
point of discussion is the remarkable structural change
undergone by the CpMCl2(dad) system on going from
Nb and Ta (for which the C-type arrangement is always
observed [10Á14]) to molybdenum. It seems reasonable
/
to attribute this change to the presence of one additional
electron on the Group 6 metal. Therefore, the six-
electron coordination mode for the ligand in C would
lead to an unfavorable 19-electron configuration.
Complex 3 appears to be the first crystallographically
characterized diazadiene complex of molybdenum(III).
The MoÃ
/
N distances are considerably shorter than in
The reduction waves are found in the narrow range
the structurally related half-sandwich Mo(II) complex
[CpMo(H2Biim)(CO)2]ꢂ
˚
between ꢃ
/
0.99 and ꢃ1.12 V for the entire family of
/
(H2Biimꢁbisimidazole;
/
complexes. We should also mention the possibility that
the reduction processes are ligand-based rather than
metal-based, since we find that the free ligands can also
be reversibly reduced in THF. If this is the case,
2.17(2) and 2.21(2) A) [29], and also shorter than the
higher oxidation state MoO2Cl2(But2Ãdad) (2.399(2) and
2.388(2) A) [30]. This is supporting evidence for the
˚

274
F. Stoffelbach et al. / Journal of Organometallic Chemistry 663 (2002) 269Á276
/
Table 3
Electrochemical properties of compounds 1Á4 and comparison with representative similar compounds
/
a
a
Compound
Reduction
Oxidation
Solv.
Reference
b
CpMoCl2(h4-C4H6)
CpMoCl2(PhCÅCPh)
CpMoCl2(EtCÅCEt)
CpMoCl2(PhNCHCHNPh) (1)
ꢃ1.11
ꢃ1.14
ꢃ1.33
ꢃ1.12
ꢃ0.99
ꢃ1.01
ꢂ0.73
ꢂ0.32
ꢂ0.16
ꢃ0.01
0.00
THF
THF
[32]
[33]
b
b
THF
THF
[33]
b
This work
This work
This work
This work
[34]
c
b
CpMoCl2(p-TolNCHCHNÃp-Tol) (2)
CpMoCl2(ArNCHCHNAr) (3)
CpMoCl2(Pri NCHCHNPri ) (4)
CpMoCl2(dppe)
CpMoCl2(PMe3)2
THF
THF
d
c
b
ꢂ0.33
ꢃ1.10
ꢃ1.91
ꢂ0.01
ꢃ0.33
ꢃ0.52
THF
CH2Cl2
CH2Cl2
c,e
[31]
a
Potentials are given relative to ferroceneÁ
Ep,a of an irreversible anodic wave.
Ep,c of an irreversible cathodic wave.
/
ferrocenium.
b
c
d
e
Arꢁ
/
2; 6ꢃC6H3Pri2:/
In THF (no reduction process is observed in CH2Cl2 before the solvent discharge).
however, coordination would cause a rather high
up process, as proposed in Scheme 2. The chloride loss
must be a reversible process, whereas the product of
chloride loss must be electrochemically inactive or
undergo an oxidation process at a more positive
potential.
positive shift (ꢀ
/
1 V) since the measured E1/2 are ꢃ
/
2.23 V for (C6H3Pri2)2Ãdad and ꢃ
/
2.02 V for p-Tol2Ã
/
dad. A greater insight into this problem could only be
obtained by theoretical studies or by isolation and
spectroscopic/magnetic characterization of the reduction
product.
3.4. Controlled radical polymerization of styrene
The reduction process shows electrochemical reversi-
bility in THF for compounds 1 and 4, whereas the
return oxidation wave has a lower intensity for com-
pounds 2 and 3, indicating that the reduction is followed
by a relatively fast follow-up chemical process. The
nature of this process has been clarified by more detailed
studies in the presence of variable amounts of added
Clꢃ salts, which were carried out in CH2Cl2 for
solubility reasons. The general shape of the voltammo-
grams and the potential values in the absence of chloride
did not vary greatly on going from THF to CH2Cl2,
except that the processes were less chemically reversible
in CH2Cl2. Even the reduction of compound 1 lost
reversibility when carried out in this solvent. The
reduction wave becomes, however, more reversible in
the presence of excess free chloride while the reduction
potential shifts toward more negative values, as shown
in Fig. 3 for the representative case of compound 2. This
behavior is consistent with chloride dissociation from
the reduced 18-electron product as the chemical follow-
Complexes 1Á4 are all active catalysts for the
/
controlled radical polymerization of styrene under
ATRP conditions (see Scheme 1) using (1-bro-
moethyl)benzene (BEB) as initiator. The controlled
nature of the polymerization processes is indicated by
the slow and linear growth of the average molecular
weight with conversion, by the good correspondence
between the experimental and the theoretical number-
average molecular weight (Mn) values, and by the
relatively low polydispersity indexes (PDIꢁ
/Mw/Mn
where Mwꢁweight-average molecular weight). An ex-
/
ample is shown in Fig. 4 for the polymerization with
complex 4, while all measurements are available as
Section 5. The experimental Mn values are slightly lower
than the theoretical ones when using compounds 1, 2
and 3, and slightly larger when using compound 4, but
the match is overall quite good.
The polymerization kinetics are first order in mono-
mer and provide the following values for the apparent
propagation constant in sꢃ1: 1.42ꢄ
/
10ꢃ4 for 1, 0.22ꢄ
/
10ꢃ4 for 2, 1.20ꢄ
/
10ꢃ4 for 3 and 0.61ꢄ
/
10ꢃ4 for 4, all
in bulk at 90 8C. This apparent rate constant is given by
Fig. 3. Cyclic voltammetric response for a solution of compound 2
10ꢃ3 M) in CH2Cl2 in the presence of added tetrabutylamm-
(1.9ꢄ
/
monium chloride. Clꢃ/Moꢁ
/
0 (a); 23 (b); 70 (c).
Scheme 2.

F. Stoffelbach et al. / Journal of Organometallic Chemistry 663 (2002) 269Á
/
276
275
Fig. 4. Mn and PDI vs. conversion for polystyrene obtained by ATRP
(bulk, 90 8C) in the presence of complex 4 and BEB (4/BEB/styreneꢁ
1/1/390). Diamonds: experimental Mn values; squares: theoretical Mn
Fig. 5. Mn and PDI vs. conversion for polystyrene obtained by SFRP
(bulk, 100 8C) in the presence of complex 3 and AIBN (3/AIBN/
/
styreneꢁ1/0.5/347). Black spheres: experimental Mn values; open
/
values; circles: PDI values.
squares: theoretical Mn values. Open triangles: PDI values.
the expression in Eq. (2) (kpꢁ
/
propagation rate con-
to control both the ATRP and the SFRP of styrene, like
related phosphine-containing half-sandwich Mo(III)
complexes [15]. The fact that the ATRP equilibrium
must forcedly be shifted towards M and RX means that
the M concentration will be significant under dynamic
equilibrium conditions. In turn, this condition favors the
engagement of the SFRP equilibrium for trapping the
reactive radical. Therefore, both the ATRP and SFRP
stant; KATRP equilibrium constant of the ATRP
ꢁ
/
equilibrium in Scheme 1). Since kp is the same for all
cases, these values should reflect the relative position of
the ATRP equilibrium. However, while the concentra-
tions of catalyst (M) and initiator (RX) have been
chosen identical in the different experiments, that of the
atom transferred complex (MÃ
/
X) is highly dependent
on the amount of bimolecular terminations that occur
during the initial phase of the polymerization process.
Therefore, no conclusions concerning the relative values
of the KATRP can be made in the absence of more
detailed kinetic investigations.
dormant species (RÃ
neously when the experiment is carried out under ATRP
conditions, as discussed earlier [15].
In our previous controlled radical polymerization
study [15], it was established that the system ability to
afford controlled polymerization by either the stable free
radical or the atom transfer mechanism is closely related
/
X and MÃ
/
R) are present simulta-
ð2Þ
to the strength of the MoÃ
/R bond in the first case and
As phosphine containing half-sandwich complexes of
Mo(III) have provided the first example of systems that
are capable of controlling the radical polymerization by
both ATRP and SFRP mechanisms simultaneously,
whereas the analogous butadiene system leads to con-
trolled polymerization only under SFRP conditions (see
Section 1), it was of interest to verify whether the
diazadiene systems are capable of functioning as rever-
sible spin traps for SFRP leading to a ‘living’ polymer-
ization process.
Indeed, complex 3 leads to the controlled/‘living’
polymerization of styrene when thermolyzed in situ
with AIBN, see Fig. 5. Like for the ATRP polymeriza-
tion described above, the polymer Mn grows slowly and
linearly with conversion, it is close to the theoretical
value (under the assumption of a radical production
the MoÃX and RÃX bonds in the other case. It was
/
/
further established that these strengths do not depend
on the nature of the ligands in a dramatic way. In line
with these arguments, the related diazadiene complexes
reported here are also able to undergo controlled
polymerization by both mechanisms. While the radical
polymerization was found to be ‘living’ with both
phosphine and h4-diene complexes under SFRP condi-
tions, however, only the phosphine systems gave ‘living’
polymerization under ATRP conditions. The diene
complex leads to catalyzed chain transfer. This differ-
ence was attributed to the different steric protection
assured by the different ligands. While the high con-
centration of the Mo(III) complex (which is the chain
transfer catalyst) under ATRP conditions should always
favor the transfer process, the bulkier phosphine ligand
protect the metal center from the close approach of the
reactive radical, thereby slowing the rate-determining
event of the chain transfer process, namely the H atom
transfer to form a Mo(IV) hydride complex. Under the
validity of this hypothesis, the results reported here for
the diazadiene derivatives indicate that the bulk increase
efficiency fꢁ1 for the AIBN thermolysis), and the PDI
/
values are relatively low. The other complexes (1, 2 and
4) have not been tested under the same conditions, but
there is little doubt that they would also be capable of
controlling the SFRP of styrene.
The results shown above indicate that at least one
half-sandwich diazadiene molybdenum complex is able
on going from CH2Ä
/
CHÃ
/
CHÄ
/
CH2 to RNÄ
/
CHÃ
/
CHÄ
/

276
F. Stoffelbach et al. / Journal of Organometallic Chemistry 663 (2002) 269Á276
/
[3] J.C. Fettinger, H.-B. Kraatz, R. Poli, E.A. Quadrelli, J. Chem.
Soc. Dalton Trans. (1999) 497.
NR is already sufficient to slow down the H atom
transfer process.
[4] E. Le Grognec, R. Poli, P. Richard, J. Chem. Soc. Dalton Trans.
(2001) 2251.
[5] P.D. Grebenik, M.L.H. Green, A. Izquierdo, V.S.B. Mtetwa, K.
Prout, J. Chem. Soc. Dalton Trans. (1987) 9.
[6] M.D. Fryzuk, L. Jafarpour, S.J. Rettig, Organometallics 18
(1999) 4050.
4. Conclusion
Members of the 17-electron CpMoCl2L2 family where
both L ligands are nitrogen donors have been synthe-
sized for the first time. Back-bonding to the diazadiene
ligand is substantial, yielding a structure where the
Mo(V)-enediamido limiting form contributes in a sub-
stantial way. The electrons of the central ene function,
however, do not participate in the bonding contrary to
the Group 5 analogues. Like previously investigated 17-
electron half-sandwich complexes of molybdenum, these
compounds are capable of controlling the radical
polymerization of styrene to yield ‘living’ polymers
under both SFRP and ATRP conditions. We are now
extending the application of these and other molybde-
num complexes to the polymerization of other mono-
mers.
[7] J.L. Davidson, K. Davidson, W.E. Lindsell, J. Chem. Soc. Chem.
Commun. (1983) 452.
[8] D. Morales, R. Poli, P. Richard, J. Andrieu, E. Collange, J.
Chem. Soc. Dalton Trans. (1999) 867.
[9] J. Andrieu, P. Braunstein, J.-M. Camus, D. Morales, F. Naud, R.
Poli, P. Richard, J. Chem. Soc. Dalton Trans. (2000) 2577.
[10] H. Kawaguchi, Y. Yamamoto, K. Asaoka, K. Tatsumi, Organo-
metallics 17 (1998) 4380.
[11] K. Mashima, Y. Matsuo, K. Tani, Organometallics 18 (1999)
1471.
[12] A. Nakamura, K. Mashima, J. Organometal. Chem. 621 (2001)
224.
[13] Y. Matsuo, K. Mashima, K. Tani, Angew. Chem. Int. Ed. Engl.
40 (2001) 960.
[14] Y. Matsuo, K. Mashima, K. Tani, Organometallics 21 (2002) 138.
[15] E. Le Grognec, J. Claverie, R. Poli, J. Am. Chem. Soc. 123 (2001)
9513.
[16] K. Matyjaszewski, J.H. Xia, Chem. Rev. 101 (2001) 2921.
[17] B. Wayland, G. Poszmick, S. Mukerjee, J. Am. Chem. Soc. 116
(1994) 7943.
5. Supplementary material
[18] F. Abugideiri, R. Poli, in: W.A. Herrmann (Ed.), Bis{m-
dichloro(h-pentamethylcyclopentadienyl)molybdenum}
Mo)-{Mo(h-C5Me5)(m-Cl)2}2, Stuttgart, 1997.
[19] J.M. Kliegman, R.K. Barnes, J. Org. Chem. 35 (1970) 3140.
[20] J.M. Kliegman, R.K. Barnes, Tetrahedron 26 (1970) 2555.
[21] H. tom Dieck, I.W. Renk, Chem. Ber. 104 (1971) 92.
(MoÃ
/
Crystallographic data for the structural analysis of
compound 3 have been deposited with the Cambridge
Crystallographic Data Centre, CCDC no. 189193.
Copies of this information may be obtained free of
charge from The Director, CCDC, 12 Union Road,
[22] L. Jafarpour, E.D. Stevens, S.P. Nolan, J. Organomet. Chem. 606
(2000) 49.
Cambridge CB2 1EZ, UK (Fax: ꢂ44-1223-336033; e-
/
[23] G.M. Sheldrick, ‘SHELX97 (Includes SHELXS97 and SHELXL97),
Release 97-2, Programs for Crystal Structure Analysis’, Univer-
mail: deposit@ccdc.cam.ac.uk or www: http://
www.ccdc.cam.ac.uk). Tables of conversions, experi-
mental and theoretical Mn, and PDI vs. time for the
ATRP and SFRP experiments (2 pages) are available
from the principal author.
sity of Gottingen, 1998.
¨
[24] L.J. Farrugia, J. Appl. Crystallogr. 32 (1999) 837.
[25] G. Van Koten, K. Vrieze, Adv. Organomet. Chem. 21 (1982) 151
(and references therein).
[26] J.L. Davidson, K. Davidson, W.E. Lindsell, N.W. Murrall, A.J.
Welch, J. Chem. Soc. Dalton Trans. (1986) 1677.
[27] L.-S. Wang, J.C. Fettinger, R. Poli, J. Am. Chem. Soc. 119 (1997)
4453.
Acknowledgements
[28] R. Poli, L.-S. Wang, J. Am. Chem. Soc. 120 (1998) 2831.
[29] M.G.B. Drew, V. Felix, I.S. Goncalves, F.E. Kuhn, A.D. Lopes,
C.C. Romao, Polyhedron 17 (1998) 1091.
[30] K. Dreisch, C. Andersson, C. Stalhandske, Polyhedron 12 (1993)
303.
We are grateful to the French Ministery of Research
and to the CNRS for funding, to Dr Mikhail Voro-
tyntsev for helpful discussion, and to B. Rebie`re for his
technical assistance.
[31] S.T. Krueger, R. Poli, A.L. Rheingold, D.L. Staley, Inorg. Chem.
28 (1989) 4599.
[32] E. Le Grognec, R. Poli, P. Richard, Organometallics 19 (2000)
3842.
References
[33] E. Le Grognec, R. Poli, P. Richard, J. Chem. Soc. Dalton Trans.
(2000) 1499.
[1] R. Poli, J. Coord. Chem. Sect. B 29 (1993) 121.
[2] J.C. Fettinger, D.W. Keogh, R. Poli, J. Am. Chem. Soc. 118
(1996) 3617.
[34] S.T. Krueger, B.E. Owens, R. Poli, Inorg. Chem. 29 (1990)
2001.