Homolytic bond strengths and formation rates in half-sandwich chromium alkyl complexes: Relevance for controlled radical polymerization

Article abstract of DOI:10.1002/anie.200801498

(Chemical Equation Presented) Radicals in check: The steric properties of the aryl substituents in chromium β-ketiminato complexes can be tuned to achieve reversible radical trapping of a growing poly(vinyl acetate) radical chain (see scheme; V-70 = radical initiator, VOAc = vinyl acetate).

Full text of DOI:10.1002/anie.200801498

Angewandte
Chemie
DOI: 10.1002/anie.200801498
Chromium Complexes
Homolytic Bond Strengths and Formation Rates in Half-Sandwich
Chromium Alkyl Complexes: Relevance for Controlled Radical
Polymerization**
Yohan Champouret, Ulrich Baisch, Rinaldo Poli,* Liming Tang, Julia L. Conway, and
Kevin M. Smith*
Dedicated to Jan Reedijk on the occasion of his retirement
In the past decade, controlled/living radical polymerization
(CRP) processes have seen a considerable surge of interest
owing, in part, to their relevance to the accessibility of a
variety of well-defined polymer structures (e.g. predeter-
mined molecular mass, narrow molecular weight distribu-
tion).[1] We have been interested in the one-electron reactivity
of transition-metal complexes and its relevance in CRP.[2] One
way in which transition-metal complexes can be used to
control radical polymerization is through a reversible deac-
tivation.The growing radical chain is trapped by formation of
a metal–carbon bond to yield a metal-capped polymer chain,
which is a dormant organometallic species (Figure 1).We
refer to this particular control mechanism as “organometallic
radical polymerization” (OMRP).[2] One of the outstanding
challenges in this area is the possibility to control the
polymerization of less reactive monomers (e.g. vinyl chloride,
vinylidene dichloride, vinyl acetate), for which activation is
made difficult by the relatively strong bonds established with
common radical traps.
Reasonable control for the radical propagation of poly-
(vinyl acetate) (PVAc) has been achieved on the basis of
another control mechanism (degenerative transfer, DT, based
on the use of xanthates or dithiocarbamates).[3,4] Results
obtained by atom transfer radical polymerization (ATRP)
have not been nearly as good,[5–7] while good control was
recently achieved in the presence of [Co(acac)2] (Mw/Mn as
low as 1.1; acac = acetylacetonate).[8,9] Recent studies have
shown that this process occurs either by DT or by OMRP,
depending on the presence of additional ligands such as
pyridine or water.[10,11]
Figure 1. Free-energy profile of controlled radical polymerizations by
OMRP. Px =growing polymer chain, Mt=metal, L=ancillary ligand(s),
M=monomer.
[*] Dr. Y. Champouret, Dr. U. Baisch, Prof. R. Poli
Laboratoire de Chimie de Coordination
UPR CNRS 8241 liØe par convention
à l’UniversitØ Paul Sabatier et
à l’Institut National Polytechnique de Toulouse
205 Route de Narbonne, 31077 Toulouse Cedex (France)
Fax: (+33)5-6155-3003
E-mail: rinaldo.poli@lcc-toulouse.fr
Homepage: http://www.lcc-toulouse.fr/equipe_g/pages_
personnelles/poli/index.html
Scheme 1. Compounds used in this study.
Dr. L. Tang, J. L. Conway, Prof. K. M. Smith[+]
Department of Chemistry, University of Prince Edward Island
550 University Avenue, Charlottetown, PE, V6T 1Z1 (Canada)
Homepage: http://web.ubc.ca/okanagan/chees/faculty/
kmsmith.html
In search of other transition-metal complexes capable of
controlling the polymerization of less reactive monomers, we
have considered the use of half-sandwich b-diketiminato
systems of CrII (Scheme 1).Previous studies have shown that
stable CrIII complexes with methyl ligands can be prepared by
oxidation of 1 with silver triflate and subsequent alkylation
with methyl lithium,[12] similar to the preparation of related
chromium b-diketiminato organometallic complexes.[13–15]
However, attempts to synthesize half-sandwich CrIII com-
plexes with larger alkyl ligands led to unexpected products,
[+] Current address: Department of Chemistry, UBC Okanagan
3333 University Way, Kelowna, BC, V1V 1V7 (Canada)
Fax: (+1)250-807-8005
E-mail: kevin.m.smith@ubc.ca
[**] R.P. thanks the ANR (contract No. NT05-2_42140) and IUF for
financial support and the CICT (Project CALMIP) for granting free
computational time, and K.M.S. thanks NSERC of Canada for
financial support and Dr. Brian O. Patrick (UBC) for performing the
single crystal X-ray structural determinations and refinements.
presumably resulting from homolysis of the CrIII alkyl bond
À
and subsequent hydrogen-atom abstraction from the sol-
vent.[16] Thus, this metal system shows promise for applica-
tions in OMRP.Herein, we show by a combination of
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200801498.
Angew. Chem. Int. Ed. 2008, 47, 6069 –6072
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6069

Communications
experimental and computational methods that the CrIII
With the idea in mind that the CrIII PVAc bond dissoci-
À
À
PVAc bond strength is dramatically influenced by the steric
effects of the b-diketiminato aryl substituents and can be
tuned to a suitable range for the CRP of the vinyl acetate
monomer (VAc).Early reports on the use of Cr II acetate and
benzoyl peroxide for controlled radical polymerization were
discussed on the basis of what we now call an OMRP
mechanism, but the polymerization activity was low for vinyl
acetate.[17–19]
ation enthalpy (BDE) could be lowered by greater ligand
steric bulk, we turned our attention to compound 1.Indeed,
when an experiment was carried out with complex 1 at 508C
in the presence of V-70 (VAc/V-70/1 = 500/0.8/1), the con-
version reached 70% after only 20 h, but control of the
polymerization was poor (Mn = 5.67 ” 104 gmolÀ1, Mw/Mn =
2.5, Mn(th) = 3.01 ” 104 gmolÀ1), suggesting that the reversible
trapping equilibrium of Figure 1 is not sufficiently displaced
toward the dormant state under these conditions.When the
reaction was repeated at lower temperature (VAc/V-70/1 =
500/0.8/1 at 308C), the polymerization was slower (Table 1,
In addition to the known compound [CpCrII{Ar1NC-
(CH3)CHC(CH3)NAr2}]
with
Ar1 = Ar2 = 2,6-iPr2C6H3
(Dipp), Cp = C5H5, 1[12] we have now prepared the analogue
with Ar1 = Ar2 = 2,6-Me2C6H3 (Xyl) 2, as well as asymmetric
analogues with Ar1 = Dipp and Ar2 = Ph (3), p-C6H4OMe (4),
and p-C6H4CF3, (5).The structures of complexes 2, 4, and 5
have been confirmed using single crystal X-ray diffraction.
Synthetic and structural details are provided in the Support-
ing Information.
k
app = 3.3 ” 10À6 sÀ1).Although the level of control is not yet
ideal (at 70% conversion Mn = 6.73 ” 104 gmolÀ1, Mw/Mn =
1.80, Mn(th) = 3.01 ” 104 gmolÀ1), the continuous growth of
Mn demonstrates the occurrence of an OMRP process.When
the same experiment was performed in the presence of a
larger excess of V-70 (VAc/V-70/1 = 500:4:1), the polymeri-
zation was faster and uncontrolled (PDI > 2.4). This result
clearly shows that, contrary to the [Co(porphyrin)][22] and
[Co(acac)2][10] systems, the present CrII complex is not capable
of mediating an associative DT process, consistent with the
absence of vacant sites on the Cr atom in the CrIII-capped
dormant chain to promote an associative radical exchange.
The effect of the ligand steric encumbrance (Xyl vs.Dipp)
À
Initial studies showed that compound 2 traps the growing
polystyrene (PS) radical chain inefficiently.Indeed, the V-70-
initiated polymerization in the presence of 2 (bulk, styrene/V-
70/2 = 250:0.8:1, V-70 = 2,2’-azobis(4-methoxy-2,4-dimethyl-
valeronitrile)) gave a polydispersity index (PDI = Mw/Mn) of
2.5–3.7 and a final polymer with Mn = 7.9 ” 104 gmolÀ1 (the-
oretical number-average molecular weight
Mn(th) = 1.7 ”
104 gmolÀ1) at 65% conversion (additional data are presented
in the Supporting Information).Since the growing PVAc
radical is expected to establish a stronger bond than the
growing PS radical,[20,21] we anticipated more efficient radical
trapping for the growing PVAc radical chain.
and the nature of the alkyl chain (PS vs.PVAc) on the Cr III
C
BDE was probed by DFT calculations on [CpCrIII{ArNC-
(CH3)CHC(CH3)NAr}(R)], where Ar= Ph or Xyl and R =
CH2Ph, CHMePh (a model of the PS growing chain) and
CHMeOOCMe (a model of the growing PVAc chain).All
calculations were carried out by full quantum mechanics, thus
we avoided the larger Dipp-substituted b-diketiminato
system; the steric effects were conveniently probed by the
comparison of Ph and Xyl systems.Amongst different
rotamers, the lowest energy ones are those in which the Ph
or OOCMe group is oriented endo relative to the Cp ring, as
shown in Figure 2.The optimized geometries of the separate
[CpCrII{ArNC(CH3)CHC(CH3)NAr}] complexes and R, as
well as full details on the DFT calculations, are given in the
Supporting Information.
In agreement with the above expectation, a preliminary
screening of VAc polymerization initiated by V-70 in the
presence of 2 at 508C (VAc/V-70/2 = 500/0.8/1) yielded
essentially no polymer in 4 h (ca.six half-lives of V-70).
Subsequent warming to 908C yielded an initial polymeri-
zation (11% conversion after an additional 4 h) and then
essentially no further conversion in the following 60 h (Mn =
1.12 ” 104 gmolÀ1, Mw/Mn = 1.68 at 12% conversion; Mn(th)
=
5.16 ” 103; Table 1), indicating a weak propensity for the
Cp{XylNC(CH3)CHC(CH3)NXyl}CrIII PVAc bond to break
À
homolytically.
Table 1: Radical polymerization of vinyl acetate initiated by V-70 with
complexes 2 and 1.
[a]
[b]
[a]
Complex
t [h]
Conv [%]
Mn SEC
Mn(th)
Mw/Mn
2[c]
3.5
8
2
–
–
–
11
12
11
12
12
2
10470
10200
10740
11270
11270
–
23800
29800
38000
67300
4730
5160
4730
5160
5160
–
2500
9900
14600
30100
1.81
1.86
1.74
1.67
1.68
–
1.36
1.75
1.74
1.80
18
22.5
48
66
3.5
8
24
32
46
1[d]
6
23
34
70
[a] Determined by size exclusion chromatography (SEC). [b] Mn(th)
=
([M]0/[Cr]0)(Mmono)(conv). [c] VAc/V-70/2=500/0.8/1, 508C for 4 h,
then 908C. [d] VAc/V-70/1=500/0.8/1, 308C.
Figure 2. DFT-optimized structures of [CpCr{ArNC(CH3)CHC-
(CH3)NAr}(R)] compounds.
6070 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 6069 –6072

Angewandte
Chemie
[a]
Table 2: Calculated CrIII C bond dissociation enthalpies (BDE) and
30 h), although the Mn and Mw/Mn were low (5.6 ” 103 gmolÀ1
À
bond lengths[b] for compounds [CpCrIII{ArNC(CH3)CHC(CH3)NAr}(R)].
and 1.15, respectively), in agreement with a controlled
process.Compound 4 gave 12% conversion within the initial
5 h at 608C, then no further increase after 40 h at 808C.
However, further heating to 1008C yielded an increased
conversion (40% after an additional 65 h).Finally, in the
presence of complex 5 an initial burst of polymerization (41%
conversion) occurred within the first 2 h at 508C, but then
conversion reached a plateau at 50% for the next 52 h and
subsequent heating to 808C took the conversion up to 71%
after an additional 15 h.It seems that the para substituent on
the aryl group (X in Scheme 1) has an important effect on the
alkyl radical trapping ability of the CrII complex (barrier to
bond formation in Figure 1).While the complexes containing
the Xyl, Dipp, and Ph substituents (1, 2, and 3) are able to trap
the growing PVAc radical chain effectively, those containing
the p-C6H4OMe and especially the p-C6H4CF3 substituents (4
and 5) are much less efficient.The origin of this electronic
effect is currently unknown; it is the subject of continuing
investigations.
Ar=Ph
Ar=Xyl
À
À
R
BDE
Cr C
BDE
Cr C
CH2Ph
CH(Me)Ph
CH(Me)OOCMe
20.8
11.8
28.4
2.136
2.173
2.109
13.3
2.0
19.7
2.146
2.197
2.124
[a] B3LYP/6-31G**; ZPVE and PVcorrections on the basis of the ideal gas
model. BDE in kcalmolÀ1. [b ] In Š.
The calculated BDEs are reported in Table 2.They show,
in agreement with the experimental evidence, a stronger
CrIII R bond for the PVAc model than for the PS model,
À
whereas the CrIII CH2Ph bond has an intermediate strength.
À
Introducing o-Me substituents on the aryl rings of the b-
diketiminato ligand considerably weakens the CrIII R bonds
À
(by 7.5 kcalmolÀ1 for the secondary CH2Ph group and by 9.8
and 8.7 kcalmolÀ1 for the tertiary PS and PVAc models,
respectively).
The effect of steric bulk on the CrIII C BDE is further
In conclusion, we have shown for the first time that a CrII/
À
revealed by trends in the bond lengths.Addition of the o-CH3
groups to the b-diketiminato aryl substituents lengthens the
bond for all R groups.Addition of the a-CH3 group to R
(going from benzyl to 1-phenylethyl) also lengthens the bond.
Conversely, replacement of Ph with OOCMe (going from the
PS to the PVAc model) strengthens the bond.This effect is
electronic rather than steric and is related to the smaller
delocalization of the radical spin density.These results are in
perfect harmony with the inefficient PS trapping and with an
irreversible PVAc trapping by the Xyl system.Extrapolation
of the calculated BDEs to the Dipp system rationalizes the
reversible trapping observed for the radical polymerization of
VAc, which shows signs of controlled growth.On the basis of
these results, we are now developing new [CpCrII{Ar1NC-
(CH3)CHC(CH3)NAr2}] systems with intermediate bulk.
Preliminary experiments in the presence of the isosteric
complexes 3–5 indicated that the steric pressure of one Dipp
CrIII system can provide a platform for controlled radical
III
À
polymerization of less reactive monomers.The Cr
C bond
strength can be tuned by modulation of the steric bulk on the
b-diketiminato ligand.However, electronic factors affecting
the rate of radical trapping must also be considered for the
optimization of the OMRP process.
Experimental Section
Ligands: The symmetric b-diketiminato ligand XylNHC(Me)CHC-
(Me)NXyl (Xyl = 2,6-Me2C6H3) was prepared according to the
literature procedure.[23] The mixed N-aryl b-diketiminato ligands
DippNHC(Me)CHC(Me)Ar (Ar= C6H5, p-C6H4OMe, or p-
C6H4CF3) were prepare by reacting the appropriate aniline derivative
with DippNHC(Me)CHC(Me)O according to the literature proce-
dure for the corresponding Ar= o-C6H4OMe derivative.[24]
2: In
a glovebox, XylNHC(Me)CHC(Me)NXyl (2.1834 g,
7.124 mmol) in THF (20 mL) was treated with nBuLi (2.0m in
pentane, 3.6 mL, 7.2 mmol) at À308C.In a separate Schlenk flask,
NaCp (2.0m solution in THF, 3.6 mL, 7.2 mmol) was added to
[CrCl2(tmeda)][25] (1.7034 g, 7.124 mmol, tmeda = N,N,N’,N’-tetrame-
thylethylenediamine) suspended in THF (25 mL) at 258C.After
30 min, the yellow solution of the deprotonated ligand was added
dropwise to the reaction mixture in the Schlenk flask.After the
reaction mixture had stirred overnight at 258C, the solvent was
removed in vacuo, and the residue was extracted with pentane and
filtered through Celite.The solvent was again removed in vacuo, and
the residue was extracted into a minimum of pentane, filtered, and
cooled to À308C overnight to yield black crystals (1.2185 g, 41.7%
yield).
and one Ph substituent does not sufficiently labilize the CrIII
PVAc bond, but revealed an unexpected electronic effect in
the radical trapping rate (Figure 3).Use of compound
yielded almost no polymer at 908C (4% conversion after
À
3
3–5: Compounds 3, 4, and 5 were prepared using the procedure as
for compound 2 but with the DippNHC(Me)CHC(Me)NAr ligands,
where Ar= C6H5, p-C6H4OMe, or p-C6H4CF3, respectively.
Controlled polymerizations: Under strict exclusion of air and
moisture, the reaction components (e.g. metal complex and V-70 for
the OMRP runs) were placed in a Schlenk tube equipped with a
stirring bar.Freshly distilled monomer and the solvent (when this was
used) were added at 08C, and three freeze-pump-thaw cycles were
carried out.The tube was then placed in a preheated oil bath at the
desired temperature.Samples were periodically withdrawn by glass
syringe after quenching the solution by cooling with an ice bath.The
conversion was calculated by weight difference after removing all
~
*
Figure 3. Behavior of complexes 3 ( VAc/V-70/3=250:0.9:1), 4 (
&
VAc/V-70/4=500:0.9:1) and 5 ( VAc/V-70/5=250:0.9:1) as OMRP
traps for PVAc. Further details are given in the Supporting Information.
Angew. Chem. Int. Ed. 2008, 47, 6069 –6072
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6071

Communications
residual monomer by evaporation under vacuum to constant weight.
The polymer residue was directly analyzed by size exclusion
chromatography after dissolution in THF.
[11] A.Debuigne, Y.Champouret, R.Jꢀrꢁme, R.Poli, C.Detrem-
bleur, Chem. Eur. J. 2008, DOI: 10.1002/chem.200800371.
[12] J.C. Doherty, K.H.D. Ballem, B.O. Patrick, K.M. Smith,
Organometallics 2004, 23, 1487 – 1489.
[13] VC. .Gibson, C.Newton, C.Redshaw, GA. .Solan, AJ.P. .
White, D.J.Williams, Eur. J. Inorg. Chem. 2001, 1895 – 1903.
[14] L.A. Macadams, W.K. Kim, L.M. Liable-Sands, I.A. Guzei,
Received: March 30, 2008
Published online: July 4, 2008
A.L. Rheingold, K.H. Theopold,
952 – 960.
[15] L.A.Macadams, G.P.Buffone, C.D.Incarvito, A.L.Rheingold,
K.H.Theopold, J. Am. Chem. Soc. 2005, 127, 1082 – 1083.
[16] J.L. Conway, K.H.D. Ballem, R. Poli, B.O. Patrick, K.M.
Smith, unpublished results.
[17] M.Lee, Y.Minoura, J. Chem. Soc. Faraday Trans. 1 1978, 74,
1726 – 1737.
[18] M.Lee, T.Morigami, Y.Minoura, J. Chem. Soc. Faraday Trans. 1
1978, 74, 1738 – 1749.
[19] M.Lee, K.Utsumi, Y.Minoura, J. Chem. Soc. Faraday Trans. 1
1979, 75, 1821 – 1829.
[20] M.B.Gillies, K.Matyjaszewski, P-.O.Norrby, T.Pintauer, R.
Poli, P.Richard, Macromolecules 2003, 36, 8551 – 8559.
[21] K.Matyjaszewski, R.Poli, Macromolecules 2005, 38, 8093 – 8100.
[22] B.B. Wayland, C.-H. Peng, X. Fu, Z. Lu, M. Fryd,
molecules 2006, 39, 8219 – 8222.
[23] P.H.M.Budzelaar, R.De Gelder, A.W.Gal,
1998, 17, 4121 – 4123.
[24] A.P.Dove, V.C.Gibson, E.L.Marshall, A.J.P.White, D.J.
Williams, Dalton Trans. 2004, 570 – 578.
[25] S.K.Hao, J.I.Song, P.Berno, S.Gambarotta,
Organometallics 2002, 21,
Keywords: chromium · homolytic bond cleavage ·
poly(vinyl acetate) · polymerization · stereoelectronic effects
.
[1] Macromolecular Engineering: Precise Synthesis, Materials Prop-
erties, Applications (Eds.: K. Matyjaszewski, Y. Gnanou, L.
Leibler), Wiley-VCH, Weinheim, 2007.
[2] R.Poli, Angew. Chem. 2006, 118, 5180 – 5192; Angew. Chem. Int.
Ed. 2006, 45, 5058 – 5070.
[3] M.Destarac, D.Charmot, X.Franck, S.Z.Zard,
Macromol.
Rapid Commun. 2000, 21, 1035 – 1039.
[4] M.Destarac, W.Bzducha, D.Taton, I.Gauthier-Gillaizeau, S.Z.
Zard, Macromol. Rapid Commun. 2002, 23, 1049 – 1054.
[5] M.Wakioka, K.Y.Baek, T.Ando, M.Kamigaito, M.Sawamoto,
Macromolecules 2002, 35, 330 – 333.
Macro-
[6] J.Xia, H-.J.Paik, K.Matyjaszewski,
8310 – 8314.
Macromolecules 1999, 32,
Organometallics
[7] H.-J. Paik, M. Teodorescu, J. Xia, K. Matyjaszewski, Macro-
molecules 1999, 32, 7023 – 7031.
[8] A.Debuigne, J.R.Caille, R.Jꢀrꢁme,
Angew. Chem. 2005, 117,
1125 – 1128; Angew. Chem. Int. Ed. 2005, 44, 1101 – 1104.
[9] A.Debuigne, J-.R.Caille, R.Jꢀrꢁme, Macromolecules 2005, 38,
5452 – 5458.
Organometallics
1994, 13, 1326 – 1335.
[10] S.Maria, H.Kaneyoshi, K.Matyjaszewski, R.Poli, Chem. Eur. J.
2007, 13, 2480 – 2492.
6072 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 6069 –6072