Cyclopentadienyl chromium β-diketiminate complexes: Initiators, ligand steric effects, and deactivation processes in the controlled radical polymerization of vinyl acetate

Article abstract of DOI:10.1021/om900869p

The new compounds CpCr(nacnac^Ar,Ar') with nacnac ^Ar,Ar' = Ar-N=C(Me)=CHC(Me)=N-Ar' (Ar = Ar' = C₆H ₂Me₃-2,4,6 or mes, 2; C₆H₃Et ₂-2,6 or dep, 3; Ar = C₆H₃Me₂-2,6 or xyl and Ar' = C₆H₃iPr₂-2,6 or dipp, 4) have been synthesized and used in polymerization experiments in addition to the previously known analogues with Ar = Ar' = xyl, 1, or dipp, 5. The compounds were used as moderators for the polymerization of vinyl acetate (VAc) initiated by V-70, according to an OMRP mechanism. The alkylchromium(III) thermal initiator CpCr(nacnac^xyl,xyl)(CH₂CMe₃) (8) was synthesized from CpCr(nacnac^xyl,xyl)(OTs) (7) and Mg(CH ₂CMe₃)₂(dioxane), while 7 was obtained from CpCr(nacnac^xyl,xyl)Cl (6) and AgOTs. The polymerizations carried out with (l-5)/V-70/VAc at elevated temperatures yielded rapid deactivation, suggestive of irreversible radical trapping. On the other hand, room-temperature polymerizations carried out with 6/VAc proceeded, albeit slowly, to greater conversions. A labilizing effect of the Ar/Ar' steric bulk is suggested by QM/MM calculations of Cr'-C BDE for models of the OMRP dormant species with 1,5, and the parent system where Ar = Ar' = Ph. Thermal deactivation of the nacnac ^xyl'xyl system has been evidenced, with formation of the acetate complex CpCr(nacnac^xyl'xyl)(OAc), 9, as confirmed by an UV-vis study and by independent synthesis. This product is proposed to form by β-acetate transfer from the growing radical chain, triggered by a head-head monomer insertion. Compounds 2, 3, 6, 7, 8, and 9 have been structurally characterized by X-ray diffraction methods.

Full text of DOI:10.1021/om900869p

Organometallics 2010, 29, 167–176 167
DOI: 10.1021/om900869p
Cyclopentadienyl Chromium β-Diketiminate Complexes: Initiators,
Ligand Steric Effects, and Deactivation Processes in the Controlled
Radical Polymerization of Vinyl Acetate
Yohan Champouret,† K. Cory MacLeod,‡ Ulrich Baisch,† Brian O. Patrick,§
Kevin M. Smith,*,‡ and Rinaldo Poli*,†,^
†
ꢀ
CNRS, LCC (Laboratoire de Chimie de Coordination), Universite de Toulouse, UPS, INPT, 205 Route de
Narbonne, F-31077 Toulouse, France, ‡Department of Chemistry, University of British Columbia Okanagan,
3333 University Way, Kelowna, British Columbia, Canada V1 V 1 V7, §Department of Chemistry, University
of British Columbia, Vancouver, British Columbia, Canada V6T 1Z1, and ^Institut Universitaire de France,
103 Boulevard Saint-Michel, 75005 Paris, France
Received October 7, 2009
0
0
The new compounds CpCr(nacnacAr,Ar ) with nacnacAr,Ar = Ar-N—C(Me)—CHC(Me)—N-Ar0
(Ar = Ar0 = C6H2Me3-2,4,6 or mes, 2; C6H3Et2-2,6 or dep, 3; Ar = C6H3Me2-2,6 or xyl and Ar0 =
C6H3iPr2-2,6 or dipp, 4) have been synthesized and used in polymerization experiments in addition to
the previously known analogues with Ar = Ar0 = xyl, 1, or dipp, 5. The compounds were used as
moderators for the polymerization of vinyl acetate (VAc) initiated by V-70, according to an OMRP
mechanism. The alkylchromium(III) thermal initiator CpCr(nacnacxyl,xyl)(CH2CMe3) (8) was synthe-
sized from CpCr(nacnacxyl,xyl)(OTs) (7) and Mg(CH2CMe3)2(dioxane), while 7 was obtained from
CpCr(nacnacxyl,xyl)Cl (6) and AgOTs. The polymerizations carried out with (1-5)/V-70/VAc at
elevated temperatures yielded rapid deactivation, suggestive of irreversible radical trapping. On the
other hand, room-temperature polymerizations carried out with 6/VAc proceeded, albeit slowly, to
greater conversions. A labilizing effectof the Ar/Ar0 steric bulk issuggested by QM/MM calculationsof
CrIII-C BDE for models of the OMRP dormant species with 1, 5, and the parent system where Ar =
Ar0 = Ph. Thermal deactivation of the nacnacxyl,xyl system has been evidenced, with formation of the
acetate complex CpCr(nacnacxyl,xyl)(OAc), 9, as confirmed by an UV-vis study and by independent
synthesis. This product is proposed to form by β-acetate transfer from the growing radical chain,
triggered by a head-head monomer insertion. Compounds 2, 3, 6, 7, 8, and 9 have been structurally
characterized by X-ray diffraction methods.
Introduction
Organometallic radical polymerization (OMRP) provides
a new route to poly(vinylacetate). In the OMRP mechanism,
unwanted bimolecular radical coupling reactions are effec-
tively prevented by reversible formation and homolytic
cleavage of metal-alkyl bonds.6 Cobalt-mediated radical
polymerization is the most well explored system for OMRP
of vinyl acetate. The mechanistic details of these polymeriza-
tions are rather complex, with a range of termination and
chain transfer steps playing a role dictated by the degree of
solvent coordination and the spin state of the Co(II) and
Co(III) intermediate species.7
The controlled polymerization of functionalized monomers
such as vinyl acetate continues to pose a formidable challenge
to transition metal chemists. Early metal catalysts that effec-
tively polymerize nonpolar monomers by an insertion
mechanism are incompatible with vinyl ester substrates.1
Morefunctional group tolerant latemetal catalysts experience
problems in copolymerization of vinyl acetate and ethylene
due to low π-bonding affinity, chelate formation, and
β-acetate elimination.2-4 Vinyl acetate is also a difficult
monomer for controlled radical polymerization methods
due to the relatively high reactivity of the propagating radical
species compared to more tractable substrates such as styrene
or methyl acrylate.5
In principle, well-defined high-spin CrII and CrIII com-
plexes could serve as a simpler system for preparing poly-
(vinylacetate) via OMRP. Chromium(II) species such as
aqueous [Cr(H2O)6]2þ or CrCl2 in coordinating aprotic
solvents have long been known to effectively trap organic
radicals at close to diffusion-controlled rates to form CrIII
-
*Corresponding authors. E-mail: kevin.m.smith@ubs.ca.
(1) Chen, E. Y.-X. Chem. Rev. 2009, 109, 5157–5214.
(2) Williams, B. S.; Leatherman, M. D.; White, P. S.; Brookhart, M.
J. Am. Chem. Soc. 2005, 127, 5132–5146.
alkyl intermediates.8 This fundamental reactivity mode is the
key step in the chromium-mediated coupling of organic
€
(3) Berkefeld, A.; Drexler, M.; Moller, H. M.; Mecking, S. J. Am.
Chem. Soc. 2009, 131, 12613–12622.
(4) (a) Ito, S.; Munakata, K.; Nakamura, A.; Nozaki, K. J. Am.
Chem. Soc. 2009, 131, 14606–14607. (b) Nakamura, A.; Ito, S.; Nozaki, K.
Chem. Rev. 2009, 109, 5215–5244.
(6) Poli, R. Angew. Chem., Int. Ed. 2006, 45, 5058–5070.
ꢀ ^
ꢀ
(7) Debuigne, A.; Poli, R.; Jerome, C.; Jerome, R.; Detrembleur, C.
Prog. Polym. Sci. 2009, 34, 211–239.
(5) Satoh, K.; Kamigaito, M. Chem. Rev. 2009, 109, 5120–5156.
(8) Smith, K. M. Coord. Chem. Rev. 2006, 250, 1023–1031.
r
2009 American Chemical Society
Published on Web 12/10/2009
pubs.acs.org/Organometallics

168 Organometallics, Vol. 29, No. 1, 2010
Champouret et al.
halides and aldehydes.9,10 The homolytic bond dissociation
energy (BDE) of a CrIII-alkyl bond is typically greater than
that of a CoIII-alkyl. Homolytic transfer of alkyl groups
from CoIII-R to CrII is used in the Takai-Utimoto reaction,
where catalytic amounts of B12 or cobalt phthalocyanine are
used to activate alkyl halides and transfer the organic radical
from Co to Cr prior to coupling with aldehydes.11 While the
trapping of organic radicals with CrII to form octahedral
CrIII organometallic species is usually considered to be rapid
and irreversible, the reaction can be reversible for secondary
or tertiary radicals due to the lowering of the Cr-R BDE
through adverse steric interactions.12,13
CpCr(nacnac) complexes are high-spin, well-characterized
species that do not bind coordinating solvents but have been
shown to trap CH3 radicals.14,15 CpCr(nacnac)(CH3) com-
plexes can be independently synthesized by the reaction of
Grignard reagents with the corresponding CrIII triflate or
halide compounds. Modifying the size of the ortho substi-
tuents of the nacnac N-aryl groups should exert a dramatic
influence on the CrIII-R BDE for CpCr(nacnac)R com-
plexes, as indicated by preliminary DFT calculations.16 This
desired tunability of the M-R BDE is critical for both
exploring structure-activity relationships for these reagents
and developing a class of well-defined organometallic com-
plexes capable of mediating OMRP reactions for a range of
activated olefin substrates.6
In 2008, we communicated our preliminary studies toward
OMRP of vinyl acetate using CpCr(nacnac) complexes and
V-70.16 In this paper, we further examine the structure-
activity relationships of vinyl acetate polymerization using
isolated CpCr(nacnac) compounds with V-70. Although the
paramagnetic nature of the CrII and CrIII complexes makes
interpretation of their NMR spectra difficult, the UV-vis
spectra of these intensely colored species proved to be quite
informative. Employing elevated temperatures to accelerate
initiation in the CpCr(nacnac)/V-70/vinyl acetate system
unexpectedly led to a decrease in the observed rate of
polymerization due to formation of an inactive CrIII acetate
thermal decomposition product. As an alternative to ther-
molysis of CpCr(nacnac) with V-70, a well-defined CrIII
neopentyl complex was developed as a single-component
OMRP reagent. The choice of the neopentyl group was
based on the high steric demand of this specific alkyl ligand,
which has long been used to provoke unique organometallic
reactivity, both of nonradical17-20 and of radical type.21-23
Experimental Section
Materials. All reactions, unless otherwise stated, were carried
out under dry, oxygen-free argon or nitrogen, using standard
Schlenk and glovebox techniques. Solvents were dried by using
the method of Grubbs,24 or they were distilled under argon from
appropriate drying agents and degassed by three freeze-
vacuum-thaw cycles prior to use.25 Celite (Aldrich) was dried
overnight at 110 ꢀC before being evacuated and then stored
under argon or nitrogen. Vinyl acetate (VAc, 99%, Alfa Aesar
or 99þ%, Aldrich) was passed through a neutral alumina
column to remove the stabilizer, dried over calcium hydride,
distilled at 90 ꢀC, degassed by three freeze-vacuum-thaw
cycles, and stored under argon or nitrogen at -20 ꢀC. 2,20-
Azobis(4-methoxy-2,4-dimethylvaleronitrile) (V-70, 96%, Wako)
was used as received. nBuLi (1.6 M in hexanes), p-toluenesulfonic
acid monohydrate, acetylacetonate, 2,6-dimethylaniline, 2,6-di-
isopropylaniline, 2,6-diethylaniline, 2,4,6-trimethylaniline, CrCl3
(anhydrous), 1,4-dioxane, silver p-toluenesulfonate, and silver
acetate were purchased from Aldrich and used as received. The
symmetric β-diketiminato ligands (2,6-Me2C6H3)NHC(Me)-
CHC(Me)N(2,6-Me2C6H3),26 (2,6-Et2C6H3)NHC(Me)CHC(Me)-
N(2,6-Et2C6H3),27 and (2,4,6-Me3C6H2)NHC(Me)CHC(Me)-
N(2,4,6-Me3C6H2)27 and the mixed N-aryl β-diketiminato ligand
(2,6-iPr2C6H3)NHC(Me)CHC(Me)N(2,6-Me2C6H3)28 were pre-
pared according to the literature procedures. NaCp was prepared
according to the literature procedure29 or was purchased from
Aldrich as a 2.0 M solution in THF and used as received. Com-
pounds Mg(CH2CMe3)2 x(1,4-dioxane),30 CrCl2(tmeda),31 1,16
3
and 516 were prepared according to literature procedures.
Characterizations. 1H NMR spectra were recorded on a Bruker
ARX 250 or a Bruker DPX 300 spectrometer. A Varian Cary
100 Bio UV-visible spectrophotometer was used to conduct
measurements using a specially constructed cell for air-sensitive
samples: a Kontes Hi-Vac Valve with PTFE plug was attached by
a professional glassblower to a Hellma 10 mm path length quartz
absorption cell with a quartz-to-glass graded seal. Size exclusion
chromatography (SEC) of poly(vinyl acetate) was carried out in
filtered THF (flow rate: 1 mL/min) at 35 ꢀC on a 300 ꢀ 7.5 mm PL
gel 5 μm mixed-D column (Polymer Laboratories), equipped with
multiangle light scattering (miniDawn Tristar, Wyatt Technology
Corporation) and refractive index (RI2000, Sopares) detectors,
witha Waters columnpack(300ꢀ 7.5 mm, Ultrastyragel 104, 103,
˚
100 A), equipped with multiangle light scattering (miniDawn
Tristar, Wyatt Technology Corp.) and refractive index (waters
410) detectors or at 30 ꢀC on a Polymer Laboratories PL-GPC
50 Plus (two PLgel mixed-C columns in series) with a PL-AS RT
autosampler and PL-RI detector. The isolated polymer samples
were dissolved in THF, and the polymer solutions were filtered
(9) Furstner, A. Chem. Rev. 1999, 99, 991–1046.
(10) Wessjohann, L. A.; Scheid, G. Synthesis 1999, 1–36.
(11) Takai, K.; Nitta, K.; Fujimura, O.; Utimoto, K. J. Org. Chem.
1989, 54, 4732–4734.
(12) Takai, K.; Matsukawa, N.; Takahashi, A.; Fujii, T. Angew.
Chem., Int. Ed. 1998, 37, 152–155.
(23) Fernandez, I.; Trovitch, R. J.; Lobkovsky, E.; Chirik, P. J.
Organometallics 2008, 27, 109–118.
(13) Wessjohann, L. A.; Schmidt, G.; Schrekker, H. S. Tetrahedron
2008, 64, 2134–2142.
(14) Doherty, J. C.; Ballem, K. H. D.; Patrick, B. O.; Smith, K. M.
(24) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Organometallics 1996, 15, 1518–1520.
(25) Armarego, W. L. F.; Perrin, D. D. Purification of Laboratory
Chemicals, 4th ed.; Butterworth Heinemann, 1996.
(26) Budzelaar, P. H. M.; de Gelder, R.; Gal, A. W. Organometallics
1998, 17, 4121–4123.
(27) Cheng, M.; Moore, D. R.; Reczek, J. J.; Chamberlain, B. M.;
Lobkovsky, E. B.; Coates, G. W. J. Am. Chem. Soc. 2001, 123, 8738–
8749.
(28) Gong, S. G.; Ma, H. Y. Dalton Trans. 2008, 3345–3357.
(29) Panda, T. K.; Gamer, M. T.; Roesky, P. W. Organometallics
2003, 22, 877–878.
(30) (a) Andersen, R. A.; Wilkinson, G. J. Chem. Soc., Dalton Trans.
1977, 809–811. (b) Dryden, N. H.; Legzdins, P.; Trotter, J.; Yee, V. C.
Organometallics 1991, 10, 2857–2870.
Organometallics 2004, 23, 1487–1489.
(15) MacLeod, K. C.; Conway, J. L.; Tang, L.; Smith, J. J.; Corcoran,
L. D.; Ballem, K. H. D.; Patrick, B. O.; Smith, K. M. Organometallics
2009, 28, 6798-6806.
(16) Champouret, Y.; Baisch, U.; Poli, R.; Tang, L.; Conway, J. L.;
Smith, K. M. Angew. Chem., Int. Ed. 2008, 47, 6069–6072.
(17) Schrock, R. R. Acc. Chem. Res. 1979, 12, 98–104.
(18) Foley, P.; DiCosimo, R.; Whitesides, G. M. J. Am. Chem. Soc.
1980, 102, 6713–6725.
(19) Pamplin, C. B.; Legzdins, P. Acc. Chem. Res. 2003, 36, 223–233.
(20) Mindiola, D. J. Acc. Chem. Res. 2006, 39, 813–821.
(21) Tsou, T.-T.; Loots, M.; Halpern, J. J. Am. Chem. Soc. 1982, 104,
623–624.
(22) Wayland, B. B.; Poszmik, G.; Mukerjee, S.; Fryd, M. J. Am.
(31) Hao, S. K.; Song, J. I.; Berno, P.; Gambarotta, S. Organo-
Chem. Soc. 1994, 116, 7943–7944.
metallics 1994, 13, 1326–1335.

Article
Organometallics, Vol. 29, No. 1, 2010 169
(pore size = 0.45 μm) before chromatographic analysis. The
columns were calibrated against linear polystyrene standards
(Polymer Laboratories). Elemental analyses were performed by
Guelph Chemical Laboratories, Guelph, ON, Canada. Solution
magnetic susceptibilites were determined by the Evans method.32
Synthesis of CpCr(nacnacmes,mes) (2). Following a procedure
similar to that previously reported for compound 1,16 compound
(2,4,6-Me3C6H2)NHC(Me)CHC(Me)N(2,4,6-Me3C6H2) (744mg,
2.23 mmol) was dissolved in THF (12 mL). n-BuLi (1.60 mL, 2.56
mmol, 1.15 equiv) was added dropwise, and the resulting yellow
solution was stirred for 30 min at room temperature. In a separate
Schlenk flask, CrCl2(tmeda) (531 mg, 2.22 mmol, 1 equiv) was
suspended in THF (35 mL) followed by the addition of NaCp
(1.25 mL, 2.50 mmol, 1.13 equiv). The resulting mixture was stirred
for 30 min at room temperature. To this solution was added
dropwise the lithium salt prepared above, and the mixture was
stirred at room temperature overnight. The solvent was evaporated
in vacuo, and the residue was extracted with hexanes, followed by
filtration through Celite. The solvent was again removed in vacuo,
and the complex was dissolved in hexanes (20 mL), filtered, and
cooled to -35 ꢀC for several days. Black crystals (730 mg) were
isolated in two crops. Yield: 73%. μeff (Evans, C6D6): 4.8(1) μB.
Anal. Calcd for C28H34CrN2: C, 74.64; H, 7.61; N, 6.22. Found: C,
74.83; H, 7.96; N, 6.42. UV/vis (hexanes; λmax, nm (ε, M-1 cm-1)):
308 (11 700), 427 (7210), 573 (376).
The green to incident and orange to transmitted light filtrate
was concentrated and cooled to -20 ꢀC to yield 1.66 g of
dark green crystals over several days in three crops. Yield: 78%.
Anal. Calcd for C26H30CrN2Cl: C, 68.19; H, 6.60; N, 6.12.
Found: C, 67.88; H, 6.50; N, 5.73. UV/vis (hexanes; λmax, nm
(ε, M-1 cm-1)): 418 (7220), 581 (504).
Synthesis of CpCr(nacnacxyl,xyl)OTs (7). Compound 6 (1.28 g,
2.79 mmol) and AgOTs (781 mg, 2.80 mmol, 1.00 equiv) were
placed in a Schlenk flask followed by the addition of THF
(60 mL). The mixture was stirred overnight at room temperature
and filtered through Celite, and the solvent was evaporated in
vacuo. The residue was extracted with 32 mL of a hexanes/
dichloromethane mixture (4:1), filtered through Celite, and
rinsed with hexanes (2 ꢀ 5 mL). The green to incident and
orange to transmitted light filtrate was cooled to -20 ꢀC to yield
1.40 g of black crystals over several days in four crops. Yield:
84%. Anal. Calcd for C33H37CrO3N2S: C, 66.76; H, 6.28; N,
4.72. Found: C, 66.50; H, 6.20; N, 4.34. UV/vis (diethyl ether;
λ
max, nm (ε, M-1 cm-1)): 411 (8140), 571 (548).
Synthesis of CpCr(nacnacxyl,xyl)CH2CMe3 (8). Compound 7
(600 mg, 1.01 mmol) was added to a Schlenk flask followed by
the addition of diethyl ether (30 mL). Mg(CH2CMe3)2 1.05(1,4-
3
dioxane) (143 mg, 0.554 mmol, 0.549 equiv) in diethyl ether
(5 mL) was added dropwise to the Schlenk. The mixture was
stirred for 1.5 h at room temperature, the solvent was evacuated
in vacuo, and the residue was extracted with hexanes (30 mL),
filtered through Celite, and rinsed with hexanes (3 ꢀ 5 mL). The
solvent was again evacuated in vacuo, and the residue was
extracted with hexanes (15 mL), filtered, and cooled to -35 ꢀC
to yield 335 mg of black crystals over several days in four crops.
Yield: 67%. Anal. Calcd for C31H41CrN2: C, 75.42; H, 8.37; N,
Synthesis of CpCr(nacnacdep,dep) (3). Following a procedure
similar to that previously reported for compound 1,16 compound
(2,6-Et2C6H3)NHC(Me)CHC(Me)N(2,6-Et2C6H3) (1.00 g, 2.76
mmol) was dissolved in THF (10 mL) and cooled to -40 ꢀC in an
acetonitrile/liquid nitrogen bath. n-BuLi (1.75 mL, 2.80 mmol,
1.01 equiv) was added dropwise, and the resulting yellow solution
was stirred for one hour at -40 ꢀC. In a separate Schlenk flask,
CrCl2(tmeda) (660 mg, 2.76 mmol, 1 equiv) and NaCp (243 mg,
2.76 mmol, 1 equiv) were suspended in THF (15 mL) and the
contents stirred for 20 min at room temperature. To this solution
was added dropwise via a cannula the lithium salt prepared
above, and the mixture was stirred at room temperature over-
night. The solvent was evaporated in vacuo, and the residue was
extracted with pentane, followed by filtration through Celite. The
solvent was again removed in vacuo, and the complex was
dissolvedin the minimumamount ofpentane, filtered, and cooled
to -80 ꢀC overnight to yield 462 mg of black crystals. Yield: 35%.
μeff (Evans, C6D6): 4.6(1) μB. Anal. Calcd for C30H38CrN2: C,
75.28; H, 8.00; N, 5.85. Found: C, 74.97; H, 8.30; N, 5.96. UV/vis
(hexanes; λmax, nm (ε, M-1 cm-1)): 308 (12200), 428 (7910),
576 (411).
5.67. Found: C, 75.27; H, 8.69; N, 5.66. UV/vis (hexanes; λmax
,
nm (ε, M-1 cm-1)): 404 (5170), 567 (1050).
Synthesis of CpCr(nacnacxyl,xyl)OC(O)Me (9). Compound 6
(259 mg, 0.566 mmol) and AgOAc (95.2 mg, 0,570 mmol, 1.01
equiv) were placed in a Schlenk flask followed by the addition of
THF (20 mL). The mixture was stirred overnight at room
temperature in the absence of light, the solvent was evacuated
in vacuo, and the residue was extracted with 12 mL of a hexanes/
dichloromethanemixture(3:1), filteredthroughCelite, andrinsed
with hexanes (3 ꢀ 3 mL). The green filtrate was concentrated
slightly and cooled to -20 ꢀC to yield 187 mg of black crystals
over several days in three crops. Yield: 69%. UV/vis (hexanes;
λ
max, nm (ε, M-1 cm-1)): 411 (9160), 508 (446), 588 (574).
General Procedures for the Radical Polymerization of Vinyl
Acetate. a. OMRP Procedure: CrII þ V-70. All polymeriza-
tions were conducted following the same experimental protocol.
A typical experiment is described here as a representative
example with complex 4 (Cr:V-70:VAc = 1:0.8:500). All opera-
tions were carried out under a protective argon atmosphere.
Complex 4 (41 mg, 0.086 mmol, 1 equiv) and V-70 (18.7 mg
0.061 mmol, 0.8 equiv) were introduced in a Schlenk tube,
followed by the addition of degassed vinyl acetate (4 mL,
43 mmol, 500 equiv). The Schlenk tube was degassed by three
freeze-vacuum-thaw cycles and then immersed in an oil bath
preheated at 50 ꢀC. At the desired time, the Schlenk flask was
rapidly cooled to room temperature by immersion into iced
water before sample withdrawal. The monomer conversion was
determined gravimetrically after removal of the unconverted
monomer under reduced pressure, and the resulting residue was
used for SEC characterization.
Synthesis of CpCr(nacnacxyl,dipp) (4). Using a procedure iden-
tical to that described above for compound 2, compound CpCr-
(nacnacxyl,dipp) (4, 223 mg) was obtained as black crystals from
(2,6-iPr2C6H3)NHC(Me)CHC(Me)N(2,6-Me2C6H3) (316 mg,
0.872 mmol). Yield: 54%. μeff (Evans, C6D6): 4.4(1) μB. Anal.
Calcd for C30H38CrN2: C, 75.28; H, 8.00; N, 5.85. Found: C,
75.00; H, 8.38; N, 5.62.
Synthesis of CpCr(nacnacxyl,xyl)Cl (6). Compound (2,6-
Me2C6H3)NHC(Me)CHC(Me)N(2,6-Me2C6H3) (1.43 g, 4.67
mmol) was added to a Schlenk flask, dissolved in THF
(30 mL), and cooled to 0 ꢀC in an ice-water bath. n-BuLi
(3.20 mL, 5.12 mmol, 1.10 equiv) was added dropwise, and the
resulting yellow solution was allowed to warm to room tem-
perature while stirring for 1 h. The lithium salt was then
cannulated into a suspension of CrCl3 (744 mg, 4.70 mmol,
1.00 equiv) in THF (20 mL) and stirred at room temperature
overnight. NaCp (2.60 mL, 5.20 mmol, 1.11 equiv) was added to
the solution, which was again stirred at room temperature
overnight. The solvent was evaporated in vacuo, and the residue
was extracted with 40 mL of a hexanes/dichloromethane mix-
ture (3:1),filteredthroughCelite,andrinsedwithhexanes(3ꢀ 5mL).
b. OMRP Procedure: CrIII-Np. The experimental protocol
is similar to that described above for the OMRP Procedure: CrII
þ V-70. As an example, complex 8 (17.5 mg, 0.035 mmol,
1 equiv) and VAc (4 mL, 43 mmol, 1200 equiv) were introduced
into a Schlenk flask and stirred at room temperature. At the
desired time, a sample was removed from the Schlenk flask and
analyzed as described above.
X-ray Crystallography. A single crystal of each compound
was mounted on a glass fiber and centered on the optical path of
(32) (a) Baker, M. V.; Field, L. D.; Hambley, T. W. Inorg. Chem.
1988, 27, 2872–2876. (b) Schubert, E. M. J. Chem. Educ. 1992, 69, 62.

170 Organometallics, Vol. 29, No. 1, 2010
Champouret et al.
a Bruker X8 APEX II diffractometer with graphite-monochro-
mated Mo KR radiation. The data were collected at a tempera-
ture of -100.0 ( 0.1 ꢀC in a series of φ and ω scans in 0.50ꢀ
oscillations. Data were collected and integrated using the Bruker
SAINT software package33 and were corrected for absorption
effects using the multiscan technique (SADABS)34 and for
Lorentz and polarization effects. All structures were solved by
direct methods with SIR97.35 For structures 2, 3, and 6 the Cp
ring was found disordered among different orientations (four
for 2 and two for 3 and 6). Two of the four orientations in 2 are
symmetry-related to the other two since the molecule sits on a
2-fold axis with a half-molecule in the asymmetric unit. Refine-
ment of the population of each of these fragments resulted in
near equivalent values of 0.25. For compounds 3 and 6, the two
Cp orientations had equal populations. In addition, one dis-
ordered half-molecule of hexane is present in the asymmetric
unit of 6. This disorder was modeled in two orientations, with
restraints employed to maintain similar geometries. Compound
8 crystallizes with two independent molecules in the asymmetric
unit. Compound 9 crystallizes as a two-component split crystal
with the two components related by a 51ꢀ rotation about the
(0 0 1) real axis. Data were integrated for both twin components,
including both overlapping and non-overlapping reflections.
The structure was solved using non-overlapped data from the
major twin component. Subsequent refinements of 9 were
carried out using HKLF 5 format data set containing complete
from component 2 and all overlapped reflections from compo-
nent 1. The batch scale refinement showed a roughly 53:47 ratio
between the major and minor twincomponents. All non-hydrogen
atoms (except those of the Cp ring for compound 2 and the
disorder solvent atoms for 6) were refined anisotropically. All
hydrogen atoms were placed in calculated positions but were not
refined. All refinements were performed using the SHELXTL
crystallographic software package of Bruker-AXS.36 The
molecular drawings were generated by the use of ORTEP-337
and POV-Ray. Crystal data and structure refinement para-
meters are collected in the Supporting Information.
Computational Details. QM/MM calculations were carried
out by use of the Gaussian03 suite of programs38 with use of the
B3LYPfunctional39 within the DFT methodology for the QM
part and the UFF40 for the MM part. The basis set chosen for
the QM calculations comprised the 6-31G* set for the C, N, and
O atoms, the 6-31G** set for the H atoms, and the SDD set,
which includes a pseudopotential, augmented by an f polariza-
tion function with the optimized41 1.941 coefficient for the Cr
atom. The cutoff between the QM and MM parts was placed at
the level of the Ar-N bonds, with the NdC(Me)-CH-
C(Me)dN diketiminato moiety being treated quantomechani-
cally together with the Cr atom, the Cp ring, and the CH-
(OAc)CH3 ligand, while the entire aryl substituents were
handled at the MM level. The input geometries were obtained
or adapted from the crystallographically characterized com-
pounds. Spin contamination was negligible, all calculations
Scheme 1
Scheme 2
converging with ÆS2æ close to the expected values (6 for the
quintet state of CrII, 6.022 for all three compounds; 3.75 for the
quartet state of CrIII, 3.824 for the Dipp derivative). The values
reported are the electronic energies without ZPVE correction.
Results and Discussion
a. Syntheses and Characterization of CpCrII(nacnac). The
CpCrII(nacnac) compounds used in this study are shown in
Scheme 1 (xyl = 2,6-dimethylphenyl; mes = 2,4,6-trimethyl-
phenyl; dep = 2,6-diethylphenyl; dipp = 2,6-diisopropyl-
phenyl). Of these, only compounds 116 and 514,16 have
previously been described in the literature.
The new CpCrII(nacnac) complexes 2-4 were prepared
following the literature procedure reported for 1,14,16 which
consists of the one-pot reaction of CrCl2(tmeda) with 1 equiv
of NaCp, followed by 1 equiv of the appropriate nacnacLi
salt (Scheme 2). Compounds 1-5 are highly air sensitive.
The 1H NMR spectra of compounds 2-4 in C6D6 all
displayed multiple broad, overlapping, unassignable signals
between 0 and 13 ppm. The magnetic susceptibilities of
complexes 2-4 were determined using the Evans method32
and were consistent with the high-spin CrII configuration
previously determined for 1.14 As previously reported for
compound 5,14 solutions of the CrII complexes 1-4 are green
to incident light with a distinctive magenta color to trans-
mitted light. They exhibit two very strong bands at 307-308
and 427-430 nm, comparable to those observed for three-
coordinate CrII nacnac alkyl complexes.42 Compounds 1-5
also have a less intense single band at 558-576 nm.
Compounds 2 and 3 have also been characterized by single-
crystal X-ray diffraction. The geometry of the two CrII com-
pounds (Figure 1) can be described as a “two-legged piano
stool”, with the Cp ring centroid lying close to the plane defined
by the Cr1, N1, and N2 or N1* atoms. Relevant bond distances
and angles are collected in Table 1. The bonding parameters for
compounds 2and 3are very similar to those found in previously
reported structures for CpCr(nacnac) CrII complexes.14,16
(33) SAINT, version 7.03A; Bruker Analytical X-ray System: Madison,
WI, 1997-2003.
(34) SADABS, Bruker Nonius area detector scaling and absorption
correction, V2.10; Bruker AXS Inc.: Madison, Wisconsin, 2003.
(35) Altomare, A.;Burla, M.;Camalli, M.;Cascarano, G.;Giacovazzo,
C.; Guagliardi, A.; Moliterni, A.; Polidori, G.; Spagna, R. J. Appl.
Crystallogr. 1999, 32, 115–119.
(36) Sheldrick, G. M. SHELXTL, version 5.1; Bruker AXS, Inc.:
Madison WI, 1997.
(37) Farrugia, L. J. J. Appl. Crystallogr. 1997, 32, 565.
(38) Frisch, M. J.; et al. et al. Gaussian 03, Revision D.01; Gaussian,
Inc.: Wallingford, CT, 2004.
(39) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
(40) Rappe, A. K.; Casewit, C. J.; Colwell, K. S.; Goddard, W. A.;
Skiff, W. M. J. Am. Chem. Soc. 1992, 114, 10024–10035.
(41) Ehlers, A. W.; Boehme, M.; Dapprich, S.; Gobbi, A.; Hoellwarth,
A.; Jonas, V.; Koehler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G.
Chem. Phys. Lett. 1993, 208, 111–114.
(42) Fan, H. J.; Adhikari, D.; Saleh, A. A.; Clark, R. L.; Zuno-Cruz,
F. J.; Cabrera, G. S.; Huffman, J. C.; Pink, M.; Mindiola, D. J.; Baik,
M. H. J. Am. Chem. Soc. 2008, 130, 17351–17361.

Article
Organometallics, Vol. 29, No. 1, 2010 171
Figure 2. Conversion as a function of time for the VAc polym-
erization initiated by V-70 in the presence of compounds 1-5.
Conditions: VAc/V-70/CrII = 500:0.8:1. T = 50 ꢀC for 4 h, then
90 ꢀC.
and Ph (19.7 and 28.4 kcal mol-1 at the B3LYP/6-31G* level,
respectively) indicated a strong effect of the steric congestion
created by the aryl substituents on the ability of the organo-
metallic dormant chain to release free radicals in solution.
We now report subsequent studies with compounds 1 and
5, as well as with compounds 2-4, which revealed a more
complex and interesting state of affairs. Compounds 2-4
were synthesized with the idea of fine controlling the
CrIII-PVAc BDE, thus allowing the development of a
system capable of yielding a suitable polymerization rate
and degree of control. However, a steric bulk increase on
going from 1 to 2 and then to 3 and 4 did not produce any
significant labilization of the CrIII-PVAc bond since the
polymerization did not proceed in any of these cases beyond
12% conversion for experiments carried out at 90 ꢀC; see
Figure 2. Indeed, the polymerization essentially stops in each
case after an initial burst of monomer consumption, within
the first 5-10 h.
On the other hand, new polymerization experiments with
compound 5, run at higher temperatures, gave a lower
apparent polymerization rate constant, only 34% conver-
sion after 84 h at 90 ꢀC. The Mn value increased linearly with
conversion and the Mw/Mn decreased to reach a value of only
1.21 for the final sample (see Figure 3), although the Mn was
much greater than expected (54 800 vs 14 600). A polymer-
ization process that takes place rapidly under mild condi-
tions (30 ꢀC)16 cannot become slower at a higher temperature
while maintaining the same mechanism. It is more logical to
think of a deactivation process (vide infra).
c. Design of a Single-Component Chromium Reagent as
OMRP Initiator. Well-defined CrIII-R complexes could
serve as single-component OMRP reagents if they possessed
a Cr-alkyl bond that was sufficiently weakened by steric
interactions to readily undergo homolysis.12,13 To date, the
success of synthetic routes to CpCr(nacnac)X complexes has
depended on the degree of steric bulk in the target molecule.
For the 2,6-iPr2-substituted system, CpCr(nacnacdipp,dipp)Cl
was difficult to prepare by salt metathesis and did not react
cleanly with Grignard reagents.14 In contrast, reducing the
size of the ortho substituents in the nacnac ligands was found
to greatly improve both the synthesis and salt metathesis
reactivity.15,43 For example, CpCr(nacnacxyl,xyl)Me was
readily prepared from CpCr(nacnacxyl,xyl)Cl with MeMgI.15
Synthesis of the corresponding CpCr(nacnacdipp,dipp)Me
had necessitated the use of the corresponding CrIII triflate
complex.14
Figure 1. Views of compounds 2 (a) and 3 (b), with thermal
ellipsoids drawn at the 50% probability level. All H atoms are
omitted and only one Cp orientation is shown for clarity.
˚
Table 1. Selected Bond Distances (A) and Angles (deg) for
Compounds 2 and 3
2
3
Distances
CNT-Cr
Cr-N
Cr-N
2.00(1)
2.007(2)
2.013(12)
2.0276(12)
2.0260(12)
Angles
CNT-Cr-N
CNT-Cr-N
CNT-(CrN2)a
N-Cr-N
134.0(3)
134.8(3)
134.5(3)
173.8(3)
90.37(5)
177.7(3)
91.2(1)
a Angle between the CNT-Cr vector and the CrN2 plane
b. VAc Polymerizations under OMRP: V-70 Initiator.
Before presenting the new results, it is necessary to review
those already described in a recent communication.16 Com-
pound 1 was shown to trap radicals produced by V-70 in the
presence of VAc, since a polymerization test with VAc/V-70/
1 = 500:0.8:1 at 50 ꢀC for 4 h and then at 90 ꢀC gave only an
11% monomer conversion after 8 h (Mn,SEC = 11 500 vs the
expected value of 4730, Mw/Mn = 1.81), which no longer
increased upon warming at 90 ꢀC for 66 h. Conversely,
compound 5 gave a much higher monomer conversion, since
a polymerization with VAc/V-70/5 = 500:0.8:1 gave a
linearly growing Mn up to a conversion of 70% (Mn,SEC
=
67 300 vs the expected value of 30 100, Mw/Mn = 1.80) in
46 h under much milder conditions (T = 30 ꢀC). Although
the controlling ability was poor (low initiator efficiency, high
polydispersity), the sustained polymerization and the Mn
growth with conversion demonstrated the reversibility of
radical trapping. The hypothesis that both complexes oper-
ate with formation of an organometallic dormant chain,
Cp(nacnacAr,Ar)CrIII-PVAc, with a much weaker CrIII
-
PVAc bond when Ar = dipp (compound 5) relative to
Ar = xyl (compound 1) seemed fully consistent with DFT
calculations of the bond strengths. The larger system with
Ar = dipp was not calculated, but a comparison of the bond
dissociation energy for the two related systems with Ar = xyl
(43) Huang, Y. B.; Jin, G. X. Dalton Trans. 2009, 767–769.

172 Organometallics, Vol. 29, No. 1, 2010
Champouret et al.
Figure 4. Thermal ellipsoid diagram (50%) of compound 6. All
H atoms are omitted and only one Cp orientation is shown
˚
for clarity. Selected bond lengths (A): Cr(1)-N(1), 2.018(2);
Cr(1)-N(2), 2.020(2); Cr(1)-CNT, 1.92(2); Cr(1)-Cl(1),
2.3082(5). Selected bond angles (deg): N(1)-Cr(1)-N(2),
90.47(7); N(1)-Cr(1)-Cl(1), 94.28(5); N(2)-Cr(1)-Cl(1),
93.14(5); CNT-Cr(1)-N(1), 124.8(4); CNT-Cr(1)-N(2),
125.0(4); CNT-Cr(1)-Cl(1), 120.34(6); CNT-(CrN2)a,
160.5(4). aAngle between the CNT-Cr vector and the CrN2
plane.
Figure 3. Number average molecular weight and polydispersity
index for the PVAc obtained in the presence of compound 5.
Conditions are as in Figure 2.
Isolated CpCr(nacnacxyl,xyl)Me does not serve as an effec-
tive single-component OMRP reagent. When 16.1 mg of this
CrIII methyl complex was dissolved in 4 mL of neat vinyl
acetate, only a 9% mass conversion was observed after 48 h
at room temperature. The resulting polymer had both a high
Mn (83 900 compared to 9140 expected Mn) and a high PDI
of 3.4. The lack of steric pressure on the small methyl ligand
The synthesis of compound CpCr(nacnacxyl,xyl)Cl (6) was
achieved by reacting CrCl3 with 1 equiv of Li(nacnacxyl,xyl),
followed by 1 equiv of NaCp to yield an air-stable crystalline
solid. The geometry of 6 (Figure 4) can be described as a
“three-legged piano stool” and is ubiquitous of half-sand-
wich CrIII complexes. The Cr-Cl bond length of 2.3082(5) A
˚
is similar to that found in CpCr(nacnacdipp,dipp)Cl,14 as well
as other cylcopentadienyl CrIII complexes with terminal Cl
groups, such as CpCr(acac)Cl and Cp*Cr(acac)Cl (Cr-Cl of
and the relative instability of the CH3 radical should make
3
the CrIII-CH3 BDE high and the homolytic dissociation
unfavorable. Initiation to form CH3 and the CrII radical
3
2.299(1) and 2.307(1) A, respectively).52 It is also similar to
˚
trap will thus be inefficient at room temperature. However,
the observation of a small amount of uncontrolled polym-
erization leading to high Mn is consistent with the slow
the Cr-Cl distance in six-coordinate CrIII nacnac complexes
with terminal Cl ligands (2.346(1) A for Cr(nacnacPh,Ph)Cl2-
˚
253
(THF) and 2.2947(12) A for Cr(nacnacdipp,dipp)(O2CMe)-
release of CH3 radicals, which then react rapidly in neat
3
˚
vinyl acetate.44
Cl(THF)54), while five-coordinate CrIII nacnac complexes
For an efficient single-component OMRP reagent, a CrIII
alkyl complex is required that has a low Cr-R BDE and that
display shorter Cr-Cl bond lengths (2.233(2) and 2.2294(14)
A for the terminal chloride ligands in [Cr(nacnacdipp,dipp)Cl-
˚
2
generates a R radical capable of reacting rapidly with vinyl
3
(μ-Cl)] and Cr(nacnacdipp,dipp)[(OCPh)2CH]Cl, respectively).54
acetate. The steric pressure exerted by the neopentyl ligand
has often been used to encourage not only intramolecular
deprotonation reactions17-20 but also metal-alkyl bond
homolysis.21-23 While several classes of even-electron mono-
meric chromium neopentyl complexes are known for CrII,31,45
CrIV,46,47 and CrVI,48-50 well-defined CrIII neopentyl com-
plexes are relatively unexplored.51
Unlike the geometries of 2 and 3, the Cr atom deviates
significantly from the plane of the nacnac ligand, toward the
˚
Cl atom. The Cr in 6 lies 0.615(3) A out of the imaginary plane
defined by the N1-C2-C4-N2 atoms, which is smaller than
˚
the 0.72 A out-of-plane distortion previously observed for the
bulkier CpCr(nacnacdipp,dipp)Cl.14 Relative to the CrII struc-
tures, the Cr-CNT distance is slightly shorter, whereas the
Cr-N distances are not significantly different.
While 6 is a useful precursor to CpCr(nacnacxyl,xyl)Me,15
attempts to install more sterically demanding alkyl ligands
once again required the use of a better leaving group. The
reaction of compound 6 with 1 equiv of AgOTs provides
compound 7 in high yields as an air-stable crystalline solid
(Figure 5). Preliminary reactions between 6 and AgOTf or
AgOTs had indicated that the CrIII tosylate was less air
sensitive and more crystalline than the CrIII triflate complex.
(44) Solution phase measurements give the bimolecular rate constant
,
for the reaction of methyl radical with vinyl acetate as 1.4 ꢀ 104 M-1 s-1
while the same reaction with the more stabilized benzyl radical has k =
14 M-1 s-1: (a) Zytowski, T.; Fischer, H. J. Am. Chem. Soc. 1996, 118,
437–439. (b) Fischer, H.; Radom, L. Angew. Chem., Int. Ed. 2001, 40,
1340–1371.
(45) Hermes, A. R.; Morris, R. J.; Girolami, G. S. Organometallics
1988, 7, 2372–2379.
(46) Mowat, W.; Shortland, A. J.; Hill, N. J.; Wilkinson, G. J. Chem.
Soc., Dalton Trans. 1973, 770–778.
(47) Schulzke, C.; Enright, D.; Sugiyama, H.; LeBlanc, G.;
Gambarotta, S.; Yap, G. P. A.; Thompson, L. K.; Wilson, D. R.;
Duchateau, R. Organometallics 2002, 21, 3810–3816.
˚
The Cr-O bond length in 7, 1.9839(15) A, is slightly shorter
than the 2.030(1) observed for the CpCr(nacnacdipp,dipp)-
(OTf),14 which may be attributable to the lower electron-
withdrawing power of the para-tolyl substituent compared
(48) Meijboom, N.; Schaverien, C. J.; Orpen, A. G. Organometallics
1990, 9, 774–782.
(49) Danopoulos, A. A.; Wilkinson, G.; Sweet, T. K. N.; Hursthouse,
M. B. J. Chem. Soc., Dalton Trans. 1995, 2111–2123.
(50) Coles, M. P.; Gibson, V. C.; Clegg, W.; Elsegood, M. R. J.;
Porrelli, P. A. J. Chem. Soc., Chem. Commun. 1996, 1963–1964.
(51) The metallacylobutane Cp*Cr(pyridine)(CH2)2CMe2 has been
prepared via the route developed for the analogous CH2SiMe3 reactions:
Heintz, R. A.; Leelasubcharoen, S.; Liable-Sands, L. M.; Rheingold,
A. L.; Theopold, K. H. Organometallics 1998, 17, 5477–5485.
€
(52) Heinemann, O.; Jolly, P. W.; Kruger, C.; Verhovnik, G. P. J.
J. Organomet. Chem. 1998, 553, 477–479.
(53) Kim, W.-K.; Fevola, M. J.; Liable-Sands, L. M.; Rheingold,
A. L.; Theopold, K. H. Organometallics 1998, 17, 4541–4543.
(54) Gibson, V. C.; Newton, C.; Redshaw, C.; Solan, G. A.; White,
A. J. P.; Williams, D. J. Eur. J. Inorg. Chem. 2001, 1895–1903.

Article
Organometallics, Vol. 29, No. 1, 2010 173
Figure 5. Thermal ellipsoid diagram (50%) of compound 7. All
˚
H atoms are omitted for clarity. Selected bond lengths (A):
Cr(1)-N(1), 2.0038(16); Cr(1)-N(2), 2.0019(17); Cr(1)-CNT,
1.896; Cr(1)-O(1), 1.9839(15). Selected bond angles (deg):
N(1)-Cr(1)-N(2), 90.17(7); N(1)-Cr(1)-O(1), 91.45(7);
N(2)-Cr(1)-O(1), 92.48(7); CNT-Cr(1)-N(1), 125.74;
CNT-Cr(1)-N(2), 123.82; CNT-Cr(1)-O(1), 123.32;
Cr(1)-O(1)-S(1), 135.77(9).
Figure 7. (a) ln([VAc]0/[VAc]) as a function of time for the VAc
polymerization initiated by compound 8. Conditions: VAc/8 =
1200:1. T = room temp. (b) Variation of the Mn,SEC (diamonds)
and Mw/Mn (triangles) as a function of conversion; the solid line
represents Mn,th
.
monitor the kinetics of iodomethane activation with CpCr-
(nacnac) complexes 1-3.15 Although 8 was stable in solution
to allow for recrystallization from hexanes, decomposition
took place upon dilution (necessary for UV/vis charac-
terization) to yield CpCr(nacnacxyl,xyl), 1. The decrease in
stability of 8 at high dilution is consistent with the facile
homolytic Cr-C bond cleavage, paralleling the reactivity of
other neopentyl complexes,21-23 with formation of com-
pound 1 and an extremely reactive neopentyl radical, and
indicated the potential for complex 8 to be used as a single-
component reagent to initiate and control the radical
polymerization of vinyl acetate.
d. OMRP of VAc with Complex 8. Polymerizations of
VAc in the presence of complex 8 (VAc/8 = 1200:1) at room
temperature gave a linearly growing Mn up to a monomer
conversion of 14% after 400 h (Mn,SEC = 16 200 vs the
expected value of 15 100, Mw/Mn = 1.46); see Figure 7.
These results contrast with the uncontrolled polymerization
observed with CpCr(nacnacxyl,xyl)Me under the same condi-
tions as discussed above. There is no visible induction period,
and the good agreement between the observed and calculated
Mn demonstrates the increased initiator efficiency compared
with the OMRP reactions with the V-70 initiator. However,
the progressive decrease of the polymerization rate constant
(decreasing slope in Figure 7a) suggests that partial deactiva-
tion of the growing chains occurs, which also causes the
broad molecular weight distribution.
Spectroscopic analysis of the polymerization reaction of
vinyl acetate with 8 revealed absorbance peaks at 422 and
556 nm, at a reaction time of 10 min, indicating that 8 had
been completely consumed, presumably being transformed
into the CpCr(nacnacxyl,xyl)(PVAc) dormant species. This is
again in contrast to the previously discussed reaction of
CpCr(nacnacxyl,xyl)Me in vinyl acetate, where after 48 h the
absorbance peak of the CrIII methyl starting material was
still evident at 546 nm. Throughout the progress of the
polymerization reaction initiated by 8, the UV/vis spectrum
further evolved to yield a shift of the major band toward
Figure 6. Thermal ellipsoid diagram (50%) of compound 8.
Compound 8 has two independent molecules in the crystal
lattice; only one is shown and all H atoms are omitted for
˚
clarity. Selected bond lengths (A): Cr(1)-N(1), 2.046(2);
Cr(1)-N(2), 2.051(2); Cr(1)-CNT, 1.949; Cr(1)-C(27),
2.136(3). Selected bond angles (deg): N(1)-Cr(1)-N(2),
89.71(9); N(1)-Cr(1)-C(27), 93.50(10); N(2)-Cr(1)-C(27),
93.63(10); CNT-Cr(1)-N(1), 123.03; CNT-Cr(1)-N(2),
122.29; CNT-Cr(1)-C(27), 125.39; Cr(1)-C(27)-C(28),
135.1(2).
to CF3, the decreased steric demand of the 2,6-Me2C6H3-
substituted nacnac, or both.
The reaction of compound 7 with 0.5 equiv of Mg-
(CH2CMe3)2 1.05(1,4-dioxane) gave CpCr(nacnacxyl,xyl)-
3
Np (8), where Np = neopentyl, CH2CMe3. The use of the
halide-free dialkyl Mg reagent was required to avoid un-
wanted substitution of the tosylate ligand in 7 with a halide
prior to alkylation to form 8. Complex 8 was found to be
highly soluble in nonpolar solvents, similar to the previously
reported CpCr(nacnacxyl,xyl)Me.14,15 The structural charac-
terization of 8 has been achieved by single-crystal X-ray
diffraction (Figure 6). The Cr-C(27) bond is significantly
˚
˚
elongated in 8 (2.136(3) A) compared to the 2.076(2) A
observed previously for the corresponding CrIII methyl
complex.15 The Cr-N bond lengths in 8 are also slightly
longer, and the Cr-C(27)-C(28) angle is 135.1(2)ꢀ, indica-
tive of the strain imposed by the tBu substituent of the
neopentyl ligand. Solutions of 8 are purple, exhibiting strong
bands in the UV-visible spectrum at 404 and 567 nm. The
strong absorbance around 550 nm seems characteristic of
CpCr(nacnac)(alkyl) complexes: the increase in absorbance
at 530 nm due to formation of CpCr(nacnac)Me was used to

174 Organometallics, Vol. 29, No. 1, 2010
Champouret et al.
Figure 9. Optimized geometries and QM/MM relative energies
(in kcal/mol) of geometry-optimized CpCrII(nacnacAr,Ar) and
CpCrIII(nacnacAr,Ar)(CHMeOOCMe) for Ar = Ph, Xyl, and
Dipp.
e. DFT Study of the OMRP Trapping Processes. The
previous communication reported geometry optimizations
at the full QM level using density functional theory (DFT)
for CpCrII(nacnacAr,Ar) and CpCrIII(nacnacAr,Ar)(CHMe-
OOCMe) (a model of the OMRP dormant chain), leading
to the calculation of the CrIII-C BDE values of 28.4 and
19.7 kcal/mol for Ar = Ph and Xyl, respectively.16 This shows
a tremendous steric effect of the nacnac aryl substituents on
the CrIII-C bond fragility, which is also reflected in the
Figure 8. Evolution of the UV/vis properties during the VAc
polymerization controlled by compound 8 (conditions as shown
in Figure 7). (a) Initial spectrum of compound 8; (b) after 10 min
of polymerization; (c) after 400 h of polymerization; (d) after
warming to 70 ꢀC (spectrum of the decomposition product).
III
˚
optimizedCr -C distances (2.109 and 2.124A, respectively).
Calculations of the Ar = dipp system were not carried out
because they are too time-consuming at the full QM level. We
now report QM/MM results for the same systems as well as
for the bulkier dipp system in terms of both energetics and
radical-trapping barriers. The energetic results and views of
the optimized structures are shown in Figure 9, whereas the
essential bonding parameters are given in Table 2. All opti-
mized geometries are available in the Supporting Information
in the form of Cartesian coordinates. The calculations were
carried out with imposition of the experimentally determined
spin state (S = 3/2 for the alkylchromium(III) complexes and
S = 2 for the chromium(II)-trapping species), leading to
calculated structures in excellent agreement with those deter-
mined crystallographically for complexes 1 and 5. The calcu-
lated Cr-C bond lengths for the cyclopentadienyl ligand are
longer than those observed experimentally:55 the CNT-Cr
distances for the CrII Xyl and Dipp nacnac complexes were
higher frequency (412 nm), while the 556 nm band decreased
in intensity; see Figure 8. This spectral evolution parallels the
observed decrease in polymerization rate (Figure 7). After
400 h of polymerization process, the sample was heated at
70 ꢀC for 3.5 h to produce a color change from purple to
green (incident light) and orange (transmitted light) with
the higher energy absorption band increasing in intensity
and the 556 nm band disappearing and being replaced by a
less intense band at 575 nm. The results suggested that,
even though the rate of polymerization had slowed sig-
nificantly (see Figure 7), there was still some CpCrIII
-
(nacnacxyl,xyl)(PVAc) compound present after 400 h of
polymerization, which underwent further reactivity once
heated.
˚
2.022 and 2.016 A for 1 and 5, respectively.
The energetic results in terms of BDE(CrIII-C) for the Ph
and Xyl systems are quite close to those previously obtained
at the full QM level, although the steric labilization exerted
by the four Me groups is predicted as less severe by the QM/
MM calculations relative to the full QM level. The present
calculation uses a polarized SDD basis set for the Cr atom,
which is considered as more balanced than the LANL2DZ
basis set without polarization functions previously used in
the full QM calculations.16 The presence of a steric effect is
confirmed on going from the Xyl to the Dipp substituents,
where the four iPr groups induce a further CrIII-C bond
labilization by 2.0 kcal/mol relative to four Me groups. This
labilization is accompanied by a significant lengthening of
It is striking to compare the behavior of the polymeriza-
tion initiated by compound 8 on one side (14% in 400 h at
room temperature, Figure 7) with that initiated by the 1/V-70
mixture (5% in 50 h at 90 ꢀC, Figure 2), both leading in
principle to the same OMRP equilibrium. This clearly proves
that the CpCrIII(nacnacxyl,xyl)(PVAc) dormant chain can be
reversibly reactivated under mild conditions to sustain the
OMRP of vinyl acetate, but it also suffers irreversible
thermal deactivation, even at room temperature at a slow
rate and much faster at more elevated temperatures.
It is interesting to compare the relatively slow polymer
growth in the presence of the CpCr(nacnacxyl,xyl) system and
the previously reported16 faster polymer growth in the pre-
sence of the CpCr(nacnacdipp,dipp) system (70% conversion
in 46 h at 30 ꢀC). This difference confirms the previously
proposed steric effect on the homolytic bond dissociation
energy, which is further investigated at the theoretical level in
the next section.
the CrIII-C bond, in the order Ph (2.091 A) < Xyl (2.115 A)
˚
˚
˚
< Dipp (2.139 A).
(55) Gallant, A. J.; Smith, K. M.; Patrick, B. O. Chem. Commun.
2002, 2914–2915.

Article
Organometallics, Vol. 29, No. 1, 2010 175
Table 2. Selected Bond Distances (A) and Angles (deg) for the B3LYP//UFF Optimized Geometries
˚
complex
CpCrII(nacnacAr,Ar
)
TS
CpCrIII(nacnacAr,Ar)(R)a
Xyl Dipp
Ar
Ph
Xyl
Dipp
Ph
Xyl
Dipp
Ph
Distances
CNT-Cr
Cr-N
2.081
2.014
2.014
2.081
2.009
2.009
2.070
2.028
2.030
2.072
2.021
2.020
4.290
2.080
2.024
2.021
3.440
2.079
2.051
2.040
3.396
2.070
2.018
2.024
2.091
2.081
2.033
2.034
2.115
2.096
2.062
2.088
2.139
Cr-C
Angles
CNT-Cr-N
133.3
133.2
133.6
133.6
132.3
133.9
134.2
133.3
108.7
178.6
92.4
131.7
131.0
106.0
161.4
91.6
131.9
132.1
101.4
167.5
93.4
121.6
124.4
117.6
141.8
92.3
123.5
124.5
114.9
143.3
90.1
123.8
124.5
114.1
143.7
91.7
CNT-Cr-C
CNT-(CrN2)b
N-Cr-N
175.8
93.3
179.5
92.7
178.5
93.8
N-Cr-C
77.5
75.1
92.2
91.4
92.3
89.1
97.5
97.2
100.2
97.8
102.0
94.8
a R = CHMeOOCMe. b Angle between the CNT-Cr vector and the CrN2 plane.
We have also optimized the transition state leading from
the CpCrII(nacnacAr,Ar) complex plus free radical to the
OMRP dormant species. The coordination geometry is quite
close to that of the CrII complex with a rather long CrII-C
distance; that is, the transition state is early for the radical-
trapping process, consistent with the very low calculated
energy barrier. The barrier is lowest, as expected, for the less
encumbered Ph derivative, whereas the most encumbered
Dipp system yields a lower barrier than the Xyl system with
intermediate steric demand. Although the barrier for trap-
ping the radical is very low for all three systems, the
unexpectedly lower barrier for the Dipp derivative may be
attributable to the structure of the CrII complex. The calcu-
lated ground-state structure of CpCr(nacnacdipp,dipp) faith-
fully reproduces the slight bowing of the N-aryl substituents
out of the plane defined by the nacnac ligand that was
observed in the X-ray structure of 1.14 As suggested by a
helpful reviewer, this sterically induced initial deformation of
the CrII Dipp complex may permit a slightly lower energy
approach of the alkyl radical compared to CpCr(nacnacxyl,
xyl), which does not display this distortion.
Figure 10. Thermal ellipsoid diagram (50%) of 9. All H atoms
˚
are omitted for clarity. Selected bond lengths (A): Cr(1)-N(1),
2.010(5); Cr(1)-N(2), 2.004(6); Cr(1)-CNT, 1.914; Cr(1)-O(1),
1.952(5); C(27)-O(1), 1.278(9); C(27)-O(2), 1.230(9). Selected
bond angles (deg): N(1)-Cr(1)-N(2), 90.7(2); N(1)-Cr(1)-O-
(1), 92.7(2); N(2)-Cr(1)-O(1), 88.6(2); CNT-Cr(1)-N(1),
124.94; CNT-Cr(1)-N(2), 123.34; CNT-Cr(1)-O(1), 125.92;
Cr(1)-O(1)-C(27), 133.6(5).
f. Isolation of the Deactivated Complex. As discussed
above, UV-vis analysis of the CrIII neopentyl compound
in vinyl acetate suggested that while 8 was consumed within
minutes at room temperature, a subsequent thermal decom-
position reaction occurred and that this is accelerated by
heating. To determine the ultimate fate of the organochro-
mium complex, the thermolyzed reaction mixture of 35.8 mg
of 8 in 4 mL of vinyl acetate was evacuated and the unknown
Cr species was extracted from the PVAc with diethyl ether.
After filtration, concentration, and storage at -35 ꢀC for two
weeks, 7.9 mgof X-rayqualitycrystals of CpCr(nacnacxyl,xyl)-
OC(O)Me (9) (Figure 10) was obtained, corresponding to a
23% isolated yield based on 8.
The independent synthesis of 9 was achieved by the
reaction of 6 with 1 equiv of AgOAc, with UV-visible
spectroscopic analysis confirming that 9 was the thermal
decomposition product of the polymerization reaction.
Upon reexamination of the V-70-initiated OMRP experi-
ments, UV-vis spectroscopy confirmed that the CrIII acetate
complex was a common termination pathway for all of the
Cr-mediated vinyl acetate polymerization reactions con-
ducted at elevated temperatures. Suspicions arose about
the possibility of CrIII acetate formation by oxidative addi-
tion of a dCH-OC(O)CH3 moiety from the polymer chain.
However, a control experiment showed no signs of reactivity
for compound 1 toward commercially obtained PVAc upon
heating to 50 ꢀC for 24 h. Similarly, prolonged heating of 1 in
vinyl acetate does not result in any change in the UV/vis
spectrum of 1. These results suggested a more complicated
mechanism for the formation of CpCr(nacnac)OAc. The pro-
posed mechanism for the formation of 9 is that of
β-acetate elimination arising from a 2,1-insertion (head-to-
head) of the monomer, as shown in Scheme 3. Head-to-head
insertion is known to be particularly problematic for the radical
polymerization of vinyl acetate, due to the relatively poor
regioselectivity of radical addition.1,5 Note that the 2,1-inser-
tion step may lead to the generation of a dormant species with a
presumably stronger CrIII-C bond, which could also contri-
bute to a gradual slow down of the polymerization process.
Elimination of β-acetate groups has been well documented
as a decomposition mode in attempts to copolymerize ethy-
lene and vinyl acetate with Ni and Pd catalysts.2,3 We point
out, however, that the β-acetate elimination process from the
newly formed dormant species is not likely to follow a classical
β-elimination mechanism, because this requires an open co-
ordination site on Cr capable of accepting the two additional
electrons furnished by the incoming acetate group. While
the metal is electronically unsaturated (15 electrons) and
metal-based orbitals are indeed available, the latter are however

176 Organometallics, Vol. 29, No. 1, 2010
Champouret et al.
Scheme 3
Scheme 4
Conclusions
The present study has revealed a more complex situation
than previously appreciated for the OMRP of vinyl acetate
mediated by half-sandwich CrII complexes of type CpCr-
(nacnac). The new data reported here confirm the sterical
labilization of the CrIII-PVAc bond, with the more encum-
bering nacnacdipp,dipp ligand resulting in a faster apparent
rate constant for polymer growth than the nacnacxyl,xyl
ligand. However, the relatively stronger bond of Cp-
(nacnacxyl,xyl)Cr-PVAc is still sufficiently labile to sustain
the OMRP of vinyl acetate even at room temperature. On the
other hand, an irreversible deactivation process comes into
play, slowly at room temperature and faster upon warming,
to yield a new material now firmly identified as the acetate
complex CpCr(nacnacxyl,xyl)OAc.
half-occupied because of the ubiquitous spin quartet config-
uration of half-sandwich CrIII. Expansion to a 17-electron
configuration would require an energetically costly (for CrIII)
electron-pairing process.56-58 A likely alternative is acetate
group transfer from the β-C atom of the radical chain to the
Cr center, as shown in Scheme 4. Unfortunately, we were
unable to observe vinyl chain end resonances in the 13C
NMR spectrum of the polymer isolated from thermolysis of
8 in vinyl acetate to provide additional support for the mechan-
ism proposed in Scheme 4. Homolytic bond rupture followed
by atom (or group) transfer has been observed for R-Co-
(porphyrin) complexes where the coordination geometry has
no cis vacant site available to accommodate the migrating atom
or group.59 It is quite possible that, following a 2,1-insertion,
the resulting CH2-terminated radical undergoes competitive
β-acetate transfer, as shown in Scheme 4, or CrIII-C bond
formation, as shown in Scheme 3, thus yielding a mixture of
acetate complex and a less active Cp(nacnacxyl,xyl)Cr-CH2CH-
(OAc)-PVAc dormant species. Only upon warming can the
latter be reactivated and eventually be completely transformed
into the final acetate product.
Acknowledgment. R.P. thanks the Agence Nationale
de la Recherche (contract ANR No. NT05-2_42140) and
the Institut Universitaire de France for financial support
and the Centre Interuniversitaire de Calcul de Toulouse
(CICT, project CALMIP) for granting free CPU time.
K.M.S. thanks the Natural Sciences and Engineering
Research Council of Canada (NSERC) for financial
support and thanks Jeffrey A. Therrien and Julia L.
Conway for obtaining single crystals of 3 and 7, respec-
tively, suitable for X-ray diffraction.
Supporting Information Available: A listing of the Cartesian
coordinates and final energies of all calculated complexes,
UV-visible spectra of complexes 1, 2, 3, 6, 7, 8, and 9, and
complete crystallographic data for complexes 2, 3, 6, 7, 8, and 9.
This material is available free of charge via the Internet at http://
pubs.acs.org.
(56) Poli, R. Chem. Rev. 1996, 96, 2135–2204.
(57) Fettinger, J. C.; Mattamana, S. P.; Poli, R.; Rogers, R. D.
Organometallics 1996, 15, 4211–4222.
(58) Mattamana, S. P.; Poli, R. Organometallics 1997, 16, 2427–2433.
(59) de Bruin, B.; Dzik, W. I.; Li, S.; Wayland, B. B. Chem.;Eur. J.
2009, 15, 4312–4320.