Telechelic polyethylene from catalyzed chain-growth polymerization

Article abstract of DOI:10.1002/anie.201208756

A functional chain transfer agent, bis(10-undecenyl)magnesium, was employed as cocatalyst in conjunction with [(C₅Me₅) ₂NdCl₂Li(OEt₂)₂] to catalyze polyethylene (PE) chain growth on magnesium. Vinyl-PE-Mg-PE-vinyl units were then further functionalized. This is the first example of catalytic ethylene polymerization with single-step, in situ functionalization to telechelic polyethylenes.

Full text of DOI:10.1002/anie.201208756

.
Angewandte
Communications
DOI: 10.1002/anie.201208756
Telechelic Polyethylene
Telechelic Polyethylene from Catalyzed Chain-Growth Polymeri-
zation**
Ian German, Wissem Kelhifi, Sꢀbastien Norsic, Christophe Boisson,* and Franck DꢁAgosto*
Since the discovery of the Ziegler–Natta catalyst[1,2] for the
coordinative polymerization of ethylene, continual chemical
and process optimizations have led to a broad range of
commodity polyolefins with enhanced properties.[3–5] Despite
these extensive efforts and numerous breakthroughs, tele-
chelic polyethylenes (PEs), in which both chain ends feature
the same functional group (X-PE-X) or chemically distinct
groups (X-PE-Z), are yet to be accessed using catalytic
ethylene polymerization. Telechelic polymers have important
commercial applications as cross-linkers, chain linkers, or
building blocks,[6] highlighting the opportunities reliant on the
development of telechelic PE production. In this context,
catalytic polymerization of ethylene, producing many PE
chains per transition-metal center, is the best route to
overcome the cost limitation[7] presented by other strategies,
while reliably attaining the crystalline, insoluble, thermo-
plastic properties of high density PE.
Previous methods to produce telechelic PE have involved
polymerization of butadiene followed by functionalization
and hydrogenation,[8] partial hydrogenation of polybutadiene
followed by metathesis degradation of the interior olefin
groups,[9] ring-opening metathesis polymerization of a cyclic
olefin followed by functionalization and hydrogenation,[10,11]
and the living coordinative polymerization of olefins.[12,13]
These techniques have produced valuable materials for the
fundamental understanding of structure–property relation-
ships. They are, however, either multistep processes, non-
catalytic (using stoichiometric quantities of high-cost initia-
tors), or employ monomers, such as butadiene or cyclic
olefins, that are expensive compared to ethylene and con-
sequently incompatible with the prerequisites of industrial
production.
in particular b-hydrogen transfer, that deactivate the chain
end. Approaches to overcoming these limitations have
emphasized the exploitation of the reactivity of the poly-
mer–metal bond present in living systems,[14,15] in which chain
transfer reactions are absent. The development of complexes
that mediate catalyzed chain growth (CCG)[16] of PE chains
on a main-group metal has facilitated the introduction of PE
end functions under catalytic conditions. In CCG polymeri-
zation, reversible PE chain transfer, which is rapid in
comparison to propagation, occurs between a catalytic
amount of a transition metal (on which the chains propagate)
and a main-group metal used as the chain-transfer agent
(CTA).[17–19] A PE-Mg-PE intermediate can be produced by
CCG on magnesium using [(C5Me5)2NdCl2Li(OEt2)2] in
combination with a dialkyl magnesium as CTA.[20,21] The
À
nucleophilic Mg C bonds of PE-Mg-PE can then be exploited
to install functional polymer chain ends.[22] Using this strategy,
continuing investigations have revealed the potential of end-
functional, linear PE blocks (PE-Z) to form part of more
complex architectures.[23–25] Alternatively, end-group trans-
formations with small-molecule reagents have been demon-
strated,[26–32] through which the reactivity, secondary structure,
or surface properties of PE can be modified.
The production of telechelic PE presents the additional
challenge of introducing potential reactivity at both ends of
the polymer chain, prior to or during polymerization. The
established reactivity of PE-Mg-PE towards electrophiles and
the behavior of dialkyl magnesium cocatalysts (MgR2) as
CTAs in coordinative ethylene polymerization, offer a syn-
thetic pathway to 1,w-difunctional PE, provided that R
features a functional group that remains a spectator during
ethylene polymerization. Bis(pentamethylcyclopentadienyl)-
neodymium catalysts have shown no propensity to incorpo-
rate a-olefins into predominantly PE chains,[33] which raised
the possibility of employing exo-alkenyl magnesium CTAs.
We describe herein the use of a bis(exo-alkenyl) magnesium
CTA in controlled CCG polymerization of ethylene, thereby
installing a vinyl functionality prior to the polymerization. We
then demonstrate the synthesis of 1,w-heterodifunctional,
linear PE (X-PE-Z) by CCG polymerization of ethylene
followed by single-step, in situ functionalization.
Bis(3-butenyl)magnesium (B-3-BM) was produced by
adaptation of a precedent Grignard disproportionation pro-
cedure.[34] Reaction of a THF solution of 3-butenylmagnesium
bromide with 1,4-dioxane caused the rapid precipitation of
{MgBr2(O2C4H8)2}n, leaving a THF solution of B-3-BM upon
filtration. To preclude a suppressive effect of THF, both on
the polymerization activity of the Nd-based catalyst and on
the level of functionality attainable upon post-polymerization
functionalization, the THF solvent was removed and replaced
In the field of catalytic ethylene polymerization, the scope
of end-functional PE production using high-volume methods
is limited by both the range of efficient, quantitative, and
selective transformations of transition-metal-bound polymer
chain ends and by competition from chain-transfer reactions,
[*] Dr. I. German, W. Kelhifi, Dr. S. Norsic, Dr. C. Boisson,
Dr. F. D’Agosto
Universitꢀ de Lyon, Univ. Lyon 1, CPE Lyon, CNRS UMR 5265
Laboratoire de Chimie, Catalyse, Polymꢁres et Procꢀdꢀs (C2P2),
Equipe LCPP, Bat 308F
43 Bd du 11 novembre 1918, 69616 Villeurbanne (France)
E-mail: boisson@lcpp.cpe.fr
dagosto@lcpp.cpe.fr
Homepage: http://www.c2p2-cpe.com
[**] The financial support from the University Claude Bernard Lyon 1 is
acknowledged.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201208756.
3438
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 3438 –3441

Angewandte
Chemie
with di-n-butyl ether. Polymerization of ethylene at 3 bar
pressure using a combination of [(C5Me5)2NdCl2Li(OEt2)2]
and B-3-BM (Mg/Nd = 150) in toluene with 2.5% di-n-butyl
ether proceeded rapidly with an activity of 2.7 ꢀ
105 gmolÀ1 hÀ1 barÀ1.
The PE produced during the polymerization was found to
1
have Mn = 1450 gmolÀ1 and a low ꢂ (1.12). The H NMR
spectrum was observed to contain no resonances correspond-
ing to vinyl end groups, which was attributed to the intra-
molecular insertion of vinyl chain ends into the neodymium–
carbon bonds, forming organometallic species that retained
the capacity to insert ethylene. Related alkylneodymium
=
À
À À
Scheme 3. Synthesis of CH2 CH (CH2)9 PE Z, where Z=H, I, or
SC(S)N(Et2)2.
À
species have been shown to insert vinyl units across the Nd C
rapidly to produce bis(vinylpolyethylenyl)magnesium (Vin-
PE)2Mg. The (Vin-PE)2Mg intermediate was then subjected
to one of three in situ treatments (Scheme 3): 1) cooling to
ambient temperature, followed by quenching with methanol
(No. 1, Table 1); 2) cooling to 108C, followed by stirring with
an iodine solution in THF for 1 h (No. 2, Table 1); or
3) addition of a solution of N,N,N’,N’-tetraethylthiuram
disulfide (disulfuram) in toluene, followed by stirring at
808C for 2 h (No. 3, Table 1).
bond during ethylene–butadiene copolymerization, leading to
the formation of cyclohexyl moieties, after which propagation
continued.[35,36]
The 13C NMR spectra indicated the selective cyclization
and formation of cylcopentane rings at the chain end
(Supporting Information, Figure S1). The formation of cyclo-
pentyl groups occurs after the first ethylene insertion by the
mechanism shown in Scheme 1, which is consistent with
The polymer produced by quenching the intermediate
(Vin-PE)2Mg with methanol was analyzed by HT-SEC and 1H
and 13C NMR spectroscopy in 2:1 TCE/C6D6 at 908C (TCE =
tetrachloroethylene). The narrow molecular-weight distribu-
tion observed (ꢂ = 1.13) and molecular weight close to that
targeted (Mn = 1150 gmolÀ1, Mn,target = 1000 gmolÀ1) indicate
a well-controlled polymerization with efficient initiation and
chain transfer to Mg and a low rate of termination, consistent
Scheme 1. Formation of cyclopentyl end groups after ethylene inser-
tion.
1
behavior reported for the related alkyl neodymium species
during ethylene–butadiene copolymerization.
with the CCG mechanism. The H NMR spectrum featured
multiplet resonances at 4.88 ppm and 5.69 ppm of intensity
consistent with quantitative retention of the terminal olefin
moiety throughout polymerization, which led to the structural
The use of longer terminal alkenyl chains in the CTA was
proposed to limit the potential for insertion of the vinyl chain
ends by precluding the formation of thermodynamically
favored 5 or 6-membered rings. 11-bromo-1-undecene was
treated with a magnesium suspension in di-n-butyl ether to
form a solution of 10-undecenylmagnesium bromide (UMB),
which was converted into a bis(10-undecenyl)magnesium
(BUM) solution by addition of dioxane to trigger MgBr2
precipitation (Scheme 2). The 1H NMR spectrum showed
=
À
À
assignment of the expected PE product CH2 CH (CH2)9
(CH2CH2)n H, that was confirmed by 13C NMR analysis
À
(Supporting Information, Figure S2).
The heterogeneous reaction between excess solution-
phase iodine and precipitated (Vin-PE)2Mg, as previously
established for monofunctional PE-I synthesis,[27] produced
=
À
À
À
CH2 CH (CH2)9 PE I as
a pale brown powder. The
1H NMR spectra of the polymers produced contained mul-
tiplet resonances at 4.88 ppm and 5.69 ppm of intensities
consistent with quantitative retention of the vinyl end group
through polymerization and post-polymerization iodination
(Figure 1; Supporting Information, Figure S3). It was found to
be necessary to limit the temperature (108C) and reaction
Scheme 2. Synthesis of bis(10-undecenyl)magnesium.
that the carbon–magnesium bond
was Mg-alkyl, and no rearrange-
ment to an Mg-alkenyl species had
occurred. The concentration of
magnesium-bound chains capable
of chain transfer was estimated by
titration with pyrene-1-acetic acid.
Polymerization of ethylene
Table 1: Polymerizations performed at 758C using the precatalyst [(C5Me5)2NdCl2Li(OEt2)2] with bis(10-
undecenyl)magnesium as cocatalyst, and subsequent in situ functionalization to give polymers of the
form vinyl-PE-Z.[a]
No.
Quenching agent
Z
Quenching time
Mn (ꢀ)[b]
[gmolÀ1
Vinyl[c]
Z[c]
Me[c]
[h/8C]
]
1
2
3
MeOH
I2 (THF)
[Et2NC(S)S]2
H
I
0.5/20
1/10
3/80
1150 (1.13)
1500 (1.14)
1550 (1.15)
1
1
1
1
0.96
0.87
–
0.04
0.13
SC(S)NEt2
using
[(C5Me5)2NdCl2Li(OEt2)2]
[a] Polymerization conditions: P=3 bar, [Nd]=42 mmolLÀ1, Mg/Nd=150, toluene/nBu2O 40:1.
and BUM (Mg/Nd = 150) in tolu-
ene with 2.5% di-n-butyl ether,
commencing at 758C, proceeded [c] Determined by H NMR at 908C (TCE/C6D6 2:1).
[b] Determined by HT-SEC in 1,2,4-trichlorobenzene at 1508C, with reference to PE standards.
1
Angew. Chem. Int. Ed. 2013, 52, 3438 –3441
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3439

.
Angewandte
Communications
=
À
Figure 3. Expansion of the MALDI-TOF mass spectrum of CH2 CH
1
=
À
À
À
À À
Figure 1. H NMR spectrum of CH2 CH (CH2)9 PE I. ?=residual
(CH2)9 PE SC(S)NEt2 showing n=40 and n=41 peak envelopes.
proton signal of C6D5H.
=
time (1 hour) of the iodination to prevent a small proportion
of chains from undergoing iodine addition to the olefinic
double bond.
Attachment of a dithiocarbamate end group to a PE chain
presents the most high-yielding route (by a quantitative
reduction) to thiol-terminated PE, which may be further
transformed to access a large number of potential structures
using thiol–ene chemistry.[32] The solution-phase reaction
between (Vin-PE)2Mg and disulfuram at 808C was used to
characteristic of the dithiocarbamate C S stretching mode.
The Lewis basic sulfur and nitrogen sites located on the
dithiocarbamate terminus allowed the analysis of polymer
composition by MALDI-TOF mass spectrometry. The mass
spectrum (Figure 3) featured a distribution of peaks Awith m/
=
z
ratios corresponding to the structure [CH2 CH2-
(CH2)2n+9SC(S)NEt2 + H]+ ([MH]+), where n is the number
of ethylene units enchained during the polymerization step.
An additional distribution of peaks B was observed, of
generally slightly lesser intensity, which appears to shadow
the higher intensity peak series at m/z ratios of B = AÀ32.
These signals are suspected to arise from the loss of a sulfur
atom by fragmentation and recombination reactions to give
=
À
À
À
produce CH2 CH (CH2)9 PE SC(S)NEt2 in a method sim-
À
ilar to that used to prepare monofunctional PE
SC(S)NEt2.[26]
The polymer produced (molecular weight Mn =
1550 gmolÀ1) had a molecular weight distribution ꢂ = 1.15,
which is characteristic of a well-controlled polymerization.
The 1H NMR spectrum features signals corresponding to
terminal vinyl proton resonances, along with those assigned to
CH2S, NCH2CH3, and NCH2CH3 protons (Figure 2). The
integral ratios of these resonances were used to estimate
chain-end functionalities, which indicated the formation of
ions of structure [CH2 CH2(CH2)2n+9C(S)NEt2 + H]+, occur-
ring during the MALDI process prior to polymer protonation,
as previously reported for dithiocarbamate-terminated PE.[26]
Following strategies employed in our laboratories,[27]
CH2 CH (CH2)9 PE I was further quantitatively trans-
formed into CH2 CH (CH2)9 PE N3, which can be further
reduced in CH2 CH (CH2)9 PE NH2 (see the Supporting
Information, Figures S6–S9 and the general polymerization
procedure).
In summary, a route to linear, well-defined telechelic PEs
by CCG polymerization has been established. A bis(exo-
alkenyl)magnesium chain-transfer agent, featuring vinyl
functionality, has been developed to quantitatively produce
vinyl functionalized PE, free from cyclized byproducts, after
simple deactivation of the resultant (Vin-PE)2Mg intermedi-
ate with methanol. Efficient, single-step functionalization
reactions between the (Vin-PE)2Mg intermediate and elec-
trophiles of the form Z2 allowed access to 1,w-heterodifunc-
=
=
À
À
À
=
À
À
À
=
À
À
À
=
À
À
À
87% of CH2 CH (CH2)9 PE SC(S)NEt2. The presence of
the dithiocarbamate terminus was corroborated by 13C NMR
and FTIR spectroscopy (Supporting Information, Figur-
=
À
À
À
es S4,S5). The FTIR spectrum of CH2 CH (CH2)9 PE
SC(S)NEt2 revealed absorbance at n = 1414 cmÀ1, which is
=
À
À
À
tional PE, Vin-PE-Z. Telechelic PEs CH2 CH (CH2)9 PE I
=
À
À
À
and CH2 CH (CH2)9 PE SC(S)NEt2 were produced with
quantitative retention of the chain-end vinyl function and Z
functionality of 96% and 87%, respectively. After one simple
and quantitative chemical transformation, azide and amine
can further be introduced. The catalytic nature of the
polymerization and use of ethylene as the monomer enhance
the economic viability of this route to low-molecular-weight
telechelic PEs. Access to telechelic PEs in a cost-effective
process can potentially boost the development of existing
applications of functional waxes, such as imaging, inks,
1
=
À
À À
Figure 2. H NMR spectrum of CH2 CH (CH2)9 PE SC(S)N(Et2)2.
?=residual proton signal of C6D5H.
3440 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 3438 –3441

Angewandte
Chemie
[20] J.-F. Pelletier, A. Mortreux, X. Olonde, K. Bujadoux, Angew.
coatings, personal-care products, and oil and plastics addi-
tives,[37] and provide new “crystalline” building blocks to be
incorporated in greater structures.
Chem. 1996, 108, 1980 – 1982; Angew. Chem. Int. Ed. Engl. 1996,
35, 1854 – 1856.
[21] T. Chenal, X. Olonde, J. F. Pelletier, K. Bujadoux, A. Mortreux,
Polymer 2007, 48, 1844 – 1856.
[22] J. Mazzolini, E. Espinosa, F. DꢂAgosto, C. Boisson, Polym.
Chem. 2010, 1, 793 – 800.
Received: October 31, 2012
Published online: February 13, 2013
[23] R. Godoy Lopez, C. Boisson, F. DꢂAgosto, R. Spitz, F. Boisson,
D. Gigmes, D. Bertin, J. Polym. Sci. Part A 2007, 45, 2705 – 2718.
[24] E. Espinosa, M. Glassner, C. Boisson, C. Barner-Kowolik, F.
DꢂAgosto, Macromol. Rapid Commun. 2011, 32, 1447 – 1453.
[25] D. Damiron, J. Mazzolini, F. Cousin, C. Boisson, F. DꢂAgosto, E.
Drockenmuller, Polym. Chem. 2012, 3, 1838 – 1845.
[26] R. Godoy Lopez, C. Boisson, F. DꢂAgosto, R. Spitz, F. Boisson,
D. Gigmes, D. Bertin, Macromol. Rapid Commun. 2006, 27, 173 –
181.
[27] R. Briquel, J. Mazzolini, T. L. Bris, O. Boyron, F. Boisson, F.
Delolme, F. DꢂAgosto, C. Boisson, R. Spitz, Angew. Chem. 2008,
120, 9451 – 9453; Angew. Chem. Int. Ed. 2008, 47, 9311 – 9313.
[28] J. Mazzolini, I. Mokthari, R. Briquel, O. Boyron, F. Delolme, V.
Monteil, D. Bertin, D. Gigmes, F. D’Agosto, C. Boisson,
Macromolecules 2010, 43, 7495 – 7503.
Keywords: chain growth · homogeneous catalysis ·
polyethylene · polymers · telechelic polymers
.
[1] K. Ziegler, E. Holzkamp, H. Breil, H. Martin, Angew. Chem.
1955, 67, 426.
[2] G. Natta, P. Pino, P. Corradini, F. Danusso, E. Mantica, G.
Mazzanti, G. Moraglio, J. Am. Chem. Soc. 1955, 77, 1708 – 1710.
[3] R. Mulhꢁupt, Macromol. Chem. Phys. 2003, 204, 289 – 327.
[4] P. Galli, G. Vecellio, J. Polym. Sci. Part A 2004, 42, 396 – 415.
[5] P. S. Chum, K. W. Swogger, Prog. Polym. Sci. 2008, 33, 797 – 819.
[6] M. A. Tasdelen, M. U. Kahveci, Y. Yagci, Prog. Polym. Sci. 2011,
36, 455 – 567.
[7] P. D. Hustad, Science 2009, 325, 704 – 707.
[29] J. Mazzolini, O. Boyron, V. Monteil, D. Gigmes, D. Bertin, F.
DꢂAgosto, C. Boisson, Macromolecules 2011, 44, 3381 – 3387.
[30] M. M. Unterlass, E. Espinosa, F. Boisson, F. DꢂAgosto, C.
Boisson, K. Ariga, I. Khalakhan, R. Charvet, J. P. Hill, Chem.
Commun. 2011, 47, 7057 – 7059.
[31] S. Akbar, E. Beyou, P. Chaumont, J. Mazzolini, E. Espinosa, F.
DꢂAgosto, C. Boisson, J. Polym. Sci. Part A 2011, 49, 957 – 965.
[32] J. Mazzolini, O. Boyron, V. Monteil, F. DꢂAgosto, C. Boisson,
G. C. Sanders, J. P. A. Heuts, R. Duchateau, D. Bertin, D.
Gigmes, Polym. Chem. 2012, 3, 2383 – 2392.
[33] G. Jeske, H. Laude, H. Mauermann, P. N. Swepston, H.
Schumann, T. J. Marks, J. Am. Chem. Soc. 1985, 107, 8091 – 8103.
[34] J. Langer, S. Krieck, R. Fischer, H. Gçrls, D. Walther, M.
Westerhausen, Organometallics 2009, 28, 5814 – 5820.
[35] V. Monteil, R. Spitz, F. Barbotin, C. Boisson, Macromol. Chem.
Phys. 2004, 205, 737 – 742.
[8] T. Shiono, N. Naga, K. Soga, Makromol. Chem. Rapid Commun.
1993, 14, 323 – 327.
[9] T. Shiono, N. Saga, K. Soga, Makromol. Chem. Rapid Commun.
1991, 12, 387 – 392.
[10] K. Sill, T. Emrick, J. Polym. Sci. Part A 2005, 43, 5429 – 5439.
[11] L. M. Pitet, M. A. Amendt, M. A. Hillmyer, J. Am. Chem. Soc.
2010, 132, 8230 – 8231.
[12] A. C. Gottfried, M. Brookhart, Macromolecules 2003, 36, 3085 –
3100.
[13] S. Park, T. Okada, D. Takeuchi, K. Osakada, Chem. Eur. J. 2010,
16, 8662 – 8678.
[14] G. W. Coates, P. D. Hustad, S. Reinartz, Angew. Chem. 2002, 114,
2340 – 2361; Angew. Chem. Int. Ed. 2002, 41, 2236 – 2257.
[15] C. J. Domski, J. M. Rose, G. W. Coates, A.D Bolig, M. Brook-
hart, Prog. Polym. Sci. 2007, 32, 30 – 92.
[16] V. C. Gibson, Science 2006, 312, 703 – 704.
[17] G. J. P. Britovsek, S. A. Cohen, V. C. Gibson, J. P. Maddox, M.
Van Meurs, Angew. Chem. 2002, 114, 507 – 509; Angew. Chem.
Int. Ed. 2002, 41, 489 – 491.
[18] R. Kempe, Chem. Eur. J. 2007, 13, 2764 – 2773.
[19] L. R. Sita, Angew. Chem. 2009, 121, 2500 – 2508; Angew. Chem.
Int. Ed. 2009, 48, 2464 – 2472.
[36] J. Thuilliez, L. Ricard, F. Nief, F. Boisson, C. Boisson, Macro-
molecules 2009, 42, 3774 – 3779.
[37] D. W. Jennings, K. Weispfennig, Fluid Phase Equilib. 2005, 227,
27 – 35.
Angew. Chem. Int. Ed. 2013, 52, 3438 –3441
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3441