Initiators for the stereoselective ring-opening polymerization of meso-lactide

Abstract

Poly(lactides) (PLAs), or poly(lactic acid)s, are among the first commercial biodegradable polymers that have the potential to become commodity plastics. meso-Lactide, a by-product of L-lactide production, will become more easily available. Highly stereoregular poly(meso-lactides) should be crystalline and thus interesting as polymeric material. In this short review, initiators capable of inducing the stereoselective ring-opening polymerization of meso-lactide to ideally give syndiotactic or heterotactic PLAs will be discussed. Mechanistic discussions with regard to understanding the reactivity differences between the various lactide monomers are included.

---

Introduction

Poly(lactic acid)s (PLAs) are biodegradable and biocompatible polymers and have been intensely studied in recent years.¹ Applications include their use as biomedical and pharmaceutical materials.^1c,d Currently their use as commodity polymers is under scrutiny, since lactide monomer, the cyclic diester of lactic acid, is derived from biomass and can be regarded as “carbon-neutral”. As polycondensation of lactic acid is not practical, ring-opening polymerization (ROP) of lactide monomer evolved as the most efficient method for production. The two stereogenic centers in one lactide (LA) molecule result in three distinct configurational isomers (S,S)-LA (L-LA), (R,R)-LA (D-LA) and (R,S)-LA (meso-LA). A 1 : 1 mixture of (S,S)-LA and (R,R)-LA is referred to as rac-LA (Scheme 1).²

Scheme 1 Lactide monomers.

The physical and mechanical properties of any polymeric material critically depend on the microstructure (tacticity) of the main chain. Polymers that have stereocenters in the repeat unit can exhibit two structures of maximum order; isotactic (from L-, D- or rac-LA) and syndiotactic (from meso-lactide). Adjacent stereocenters of isotactic polymers are of the same relative stereochemistry, whereas those of syndiotactic polymers are of opposite relative configuration. Due to their stereoregularity, isotactic (T_m > 180 °C) and syndiotactic polymers (T_m = 152 °C) are crystalline, an important feature for many applications. Heterotactic PLAs (obtained from rac- or meso-LA) have been reported to be amorphous and to show no melting temperature so far. Mixtures of poly(L-lactide), PLLA, and poly(D-lactide), PDLA, forming stereocomplexes (with T_m values up to 220 °C) are currently at the center of focus.³

Stereoregular PLA materials can be prepared from meso-LA by using metal complexes as single-site initiators. Two different mechanisms are conceivable. One is chain-end-control, where the insertion of the incoming meso-LA monomer is determined by the stereogenic center in the last repeating unit in the propagating chain. If the stereogenic center in the last unit is repeated, syndiotactic PLA will be obtained. If the stereogenic center of the last unit favors alternating enchainment through the opposite stereogenic center, heterotactic PLA will result. The other mechanism is enantiomorphic site-control where the insertion of the incoming meso-LA monomer is controlled by the ligand sphere. In this case, syndiotactic PLA will be obtained from meso-LA (Scheme 2).

Scheme 2 Structures of PLAs synthesized from meso-lactide.

In this review, we will summarize the current status of the synthesis of poly(meso-lactide) by ring-opening polymerization using single-site metal initiators.

Early studies

Kricheldorf et al. were the first to use soluble metal complexes for the ring-opening polymerization of meso-lactide.⁴Aluminium tri(isopropoxide) Al(OⁱPr)₃, AlEt₃/neopentyl alcohol, AlEt₃/(+)-menthol and methylalumoxane were shown to polymerize meso-lactide in xylene solutions at 120 °C. meso-Lactide was polymerized at slower rates compared to rac-lactide.^4a This was rationalized by suggesting higher stability of meso-lactide with respect to rac-lactide due to the trans disposition of the two methyl groups. When the time of the polymerization of meso-lactide was increased, the polymer stereochemistry was affected. Due to transesterifications (“back-biting”) the iis/siitetrad fraction increased after 48 h.

Di(n-butyl)magnesium polymerized meso-lactide in toluene at room temperature with full conversion reached after 8 days resulting in nearly atactic polylactides.^4b

Zinc, lead, antimony, or bismuth compounds were also used as initiators in the ring-opening polymerization of meso-lactide at 120 °C in xylene.^4c The polymerizations using lead oxide and bismuth(2-ethylhexanoate) showed high conversion but resulted in low molecular weight polymers. When zinc stearate and antimony (2-ethylhexanoate) were used, meso-lactide was polymerized faster than rac-lactide to afford PLAs in similar yields and inherent viscosities.

Aluminium initiators

Inspired by the pioneering work of Spassky et al. on the isospecific polymerization of rac-lactide,⁵ a breakthrough in the stereoselective polymerization of meso-lactide was reported by Ovitt and Coates in 1999 with the use of an aluminium isopropoxide complex ((R)-1, Scheme 3).⁶ The enantiopure complex, (R)-1, achieved 94% conversion at 70 °C after 40 h (M_n,exp = 12 030 g mol⁻¹ and M_w/M_n = 1.05). Due to its high degree of stereoregularity, the PLA with the highest level of syndiotacticity (96%) exhibited crystallinity with a melting temperature, T_m = 152 °C, following annealing at 95 °C for 60 min (T_g = 38.1 °C). Interestingly, rac-1 showed 98% conversion under similar conditions, but produced an amorphous polymer (M_n,exp = 13 600 g mol⁻¹ and M_w/M_n = 1.07) with glass transition temperature T_g = 43.2 °C.⁷ The most notable feature of the reaction using the racemic catalyst is that the polymer was heterotactic (where 80% of the linkages formed during the polymerization occurred between lactate units of the identical stereochemistry). A polymer exchange mechanism was proposed to account for the formation of this heterotactic microstructure, whereby each individual polymer chain effectively switches between enantiomeric aluminium centers before insertion of a subsequent monomer unit.

Scheme 3 Aluminium initiators for the polymerization of meso-lactide.

Feijen et al. showed that meso-lactide polymerization using chiral aluminium salan complexes in the presence of isopropanol (2a–c, Scheme 3) occurred in a controlled fashion (M_n,exp = 6000–7000 g mol⁻¹ and M_w/M_n = 1.09–1.12), resulting in syndiotactically biased PLAs (P_s = 0.64, 0.70 and 0.69, respectively) (Table 1).⁸

Table 1 Polymerization of meso-lactide using aluminium complexes^a

Initiator	[Initiator]0/mol L^−1	[mon.]0/[init.]0	Time/h	% Conv.	P s	Ref.
a Polymerization conditions: toluene, 70 °C.
1	0.002	100	40	94	0.96	7
2a	0.534	50	23	97	0.64	8
2b	0.534	50	27	94	0.69	8
2c	0.534	50	27	83	0.70	8
3a	0.474	96	1.5	93	0.56	9
3b	0.474	96	2.5	47	0.57	9
3c	0.474	96	4	94	0.53	9

---

They also demonstrated that achiral aluminium salen complexes (3a–c, Scheme 3) gave atactic polymers (P_s = 0.56, 0.57 and 0.53, respectively) in the presence of isopropanol.⁹ The minor difference in the tacticity of these polymers suggests that substituents on the ligand have little effect on the chain-end control during meso-lactide polymerization. In all cases, the polymerization rates for meso-lactide were lower than those for rac-lactide.

Group 3 metal initiators

Carpentier et al. reported 97% conversion of meso-lactide after 0.5 h at 20 °C using the amine bis(phenolate) yttrium complex 4 (Scheme 4).¹⁰ Analysis of the polymer microstructure revealed that initiator 4 formed syndiotactically enriched PLA (P_s = 0.75).

Scheme 4 Group 3 metal initiators for the polymerization of meso-lactide.

Initiators based on group 3 bis(phenolate) complexes 5a–f (Scheme 4), active in heteroselective ROP of rac-LA, were found to polymerize meso-lactide, producing highly syndiotactic PLAs.¹¹ These complexes contain a “temporary” chiral reaction site that apparently allows for selective ring-opening of one of the diastereotopic ester functions.¹¹ The C₃-linked complex 5e gave the highest syndiotacticity (P_s = 0.90) when the tert-butylortho-substituents complexes 5a, 5c, and 5f are compared with respect to their ROP activity. Fast conversion and high syndiotacticity were achieved using 5b and 5d, C₂-bridged scandium derivative with ortho-cumyl (CMe₂Ph) substituents in the phenoxy group (P_s > 0.92). Apparently, complexes with bulkier orthocumyl substituents showed better control over syndiotacticity than those with tert-butyl substituents.

Polymerizations of meso-lactide in melt (60–130 °C) using scandium complex 5f were rapid (50% monomer conversion in less than 10 min) and controlled. Tacticity control (P_s > 0.83) was as high as that observed in solution. The scandium complex 5d showed complete conversion within 30 min (P_s = 0.92), whereas the homologous yttrium complex led to 79% conversion (P_s = 0.71) (Table 2).¹¹

Table 2 Polymerization of meso-lactide using group 3 complexes

Initiator	[Initiator]0/mol L^−1	[mon.]0/[init.]0	% Conv.	P s	Ref.
a Polymerization conditions: 30 min, THF, 20 °C. b Polymerization conditions: 30 min, toluene, 25 °C.
4	0.004^a	100	97	0.75	10
5a	0.002^b	100	99	0.88	11
5b	0.002^b	100	88	0.93	11
5c	0.002^b	100	99	0.89	11
5d	0.002^b	100	99	0.92	11
5e	0.002^b	100	99	0.90	11
5f	0.002^b	100	99	0.89	11
6	0.002^b	100	99	0.72	13
7	0.002^b	100	99	0.73	13

---

The use of scandium amide complexes containing cyclen-derived (NNNN)-type macrocycle ligands 6 and 7 for meso-lactide polymerization was reported (Scheme 4). Complexes 6 and 7 were fast (full conversion at room temperature in less than 30 min), giving syndiotactic PLA (P_s = 0.73) with molecular weight efficiency values¹² of 3.81 for 6 (M_n,exp = 46 000–55 000 g mol⁻¹) and 0.85 for 7 (M_n,exp = 12 250–15 000 g mol⁻¹).¹³

Indium initiators

Tolman et al. reported the use of a mixture of InCl₃ in the presence of benzyl alcohol and triethylamine for the polymerization of meso-lactide.¹⁴ The polymerizations were fast; full conversion reached after 30 h at 0 °C and 5 h at 25 °C. The polylactides obtained at 0 °C were mainly atactic (P_s < 0.62). Using well-characterized indium initiators with a zwitterionic structure [{InCl₃(3-diethylamino-1-propanol)(H₂O)}₂] (8, Scheme 5), they have shown a lack of stereoselectivity in the polymerization of meso-lactide.¹⁵

Scheme 5 Indium initiators for the polymerization of meso-lactide.

Indium bis(phenolate) complex bearing tert-butylortho-substituent (9, Scheme 5) gave highly syndiotactic PLA with low polydispersity (P_s = 0.93 and M_w/M_n = 1.05) at room temperature, with full conversion reached after 16 h (Table 3).¹¹

Table 3 Polymerization of meso-lactide using indium compounds

Initiator	[Initiator]0/mol L^−1	[mon.]0/[init.]0	Time/h	% Conv.	P s	Ref.
a Polymerization conditions: [InCl3]0/[PhCH2OH]0 = 1, dichloromethane and triethylamine, 0 °C. b Polymerization conditions: [InCl3]0/[PhCH2OH]0 = 1, dichloromethane and triethylamine, 25 °C. c Polymerization conditions: [InCl3]0/[PhCH2OH]0 = 1, melt and triethylamine. d Polymerization conditions: toluene, 25 °C.
InCl3	0.276^a	100	30	>99	0.62	14
InCl3	0.276^b	100	5	>99	0.56	14
InCl3	0.276^c	50	5	>99	0.44	14
9	0.520^d	100	16	>99	0.93	11

---

Group 4 metal initiators

Heteroselective group 4 metal initiators containing a 1,ω-dithiaalkanediyl-bridged bis(phenolato) ligand with bulky groups at the ortho-substituent R were obtained from [M(OR′)₄] (M = Ti or Zr, and R′ = ⁱPr or ^tBu) (10a–e, Scheme 6).¹⁶ When complexes 10a, 10d, and 10e with tert-butylortho-substituents are compared with respect to their ROP activity, the C₃-linked complex 10e gave the highest conversion, under the same polymerization conditions (38% for 10a, 8% for 10d and 71% for 10e, Table 4), indicating the importance of a flexible backbone. It is noteworthy that the rigid trans-1,2-cyclohexanediyl backbone (complex 10d) gave only 8% conversion after 24 h at 100 °C.

Scheme 6 Group 4 metal initiators for the polymerization of meso-lactide.

Table 4 Polymerization of meso-lactide using group 4 metal complexes

Initiator	[Initiator]0/mol L^−1	[mon.]0/[init.]0	Time/h	% Conv.	P s	Ref.
a Polymerization conditions: toluene, 100 °C. b Polymerization conditions: C6D6, 100 °C. c Polymerization conditions: toluene, 50 °C.
10a	0.520^a	100	24	38	0.63	16
10b	0.520^a	100	24	73	0.73	16
10c	0.520^a	100	24	>99	0.71	16
10d	0.520^a	100	24	8	—	16
10e	0.520^a	100	24	71	0.70	16
11a	0.520^b	50	24	71	0.62	16
11b	0.520^b	50	24	94	0.71	16
12a	0.520^c	100	48	79	0.18	17
12b	0.520^c	100	48	85	0.25	17
12c	0.520^c	100	48	93	0.62	17

---

The syndiotacticity is higher when the ortho-position is a cumyl (2-phenylpropyl) substituent in complex 10b (73% conversion, P_s = 0.73) when compared with complex 10a with tert-butyl substituent in the phenoxy moiety (38% conversion, P_s = 0.63). A change in the metal has a significant effect on the polymerization of meso-lactide. Zirconium complex 10c polymerized meso-lactide faster (>99% conversion) than the homolog titanium complex 10b (73%) with a monomer : initiator ratio of 100. Asymmetric group 4 metal initiators were obtained from [Ti(OⁱPr)_n−xCl_x] (11a,b, Scheme 6).¹⁶

When complex 11a was used as an initiator 71% conversion was attained (M_w/M_n = 1.06 and P_s = 0.62) at 100 °C for 24 h, in C₆D₆ (monomer : initiator ratio of 50). However, 94% conversion was obtained with 11b (M_w/M_n = 1.07, P_s = 0.71).¹⁶ The efficiency values for these initiators were approximately 0.5.

C ₂-Symmetric complexes of tetravalent metals containing two 1,ω-dithia-alkanediyl-bridged bis(phenolato) (OSSO)-type ligands polymerized meso-lactide to give heterotactic or syndiotactic polylactides depending on the metal center.¹⁷Polymers at 50 °C in toluene obtained by using zirconium complex 12a showed the lowest syndiotacticity (P_s = 0.18) compared with that of polymers obtained by the hafnium complex 12b (P_s = 0.25) and cerium complex 12c (P_s = 0.62), with the latest showing the highest monomer conversion. This was associated with the space available between the two (OSSO)-type ligands in the cerium complex, as indicated the results of single crystal analysis of 12c.¹⁷

Zinc initiators

Polymerization of meso-lactide using BDI zinc isopropoxide complex (13a,b, Scheme 7) afforded PLAs with different microstructures.¹⁸ Complex 13a provides access to syndiotactic PLAs without the use of chiral ligands. PLA was produced with a molecular weight of 22 400 g mol⁻¹ (M_w/M_n = 1.07, P_s = 0.76) after 5 h at 0 °C in dichloromethane (monomer : initiator ratio of 200). However, complex 13b produced moderately heterotactic PLA and suggested chain-end stereocontrol to be operative during polymerization. Complex 13b polymerized meso-lactide at 0 °C with 97% of conversion after 4 h, with a monomer : initiator ratio of 99 (M_n,exp = 13 800 g mol⁻¹, M_w/M_n = 1.02 and P_s = 0.37).

Scheme 7 Zinc complexes for polymerization of meso-lactide.

Group 2 metal initiators

Alkaline earth metal amide complexes containing a cyclen-derived (NNNN)-type macrocyclic ligand polymerized meso-lactide (14a,b, Scheme 8) to afford syndiotactically enriched PLAs.¹⁹Magnesium complex 14a and calcium complex 14b are fast initiators; full conversions were reached at room temperature, in toluene or in tetrahydrofuran, in less than 30 min. Syndiotacticity P_s was 0.64 for 14a (0.70 in THF) and 0.63 for 14b in toluene. The initiators showed molecular weight efficiency values of 0.60 for 14a (M_w/M_n = 1.50) and 1.18 for 14b (M_w/M_n = 1.41) for meso-lactide polymerization.

Scheme 8 Alkaline earth metal complexes for polymerization of meso-lactide.

Organocatalysts

Waymouth, Hedrick et al. investigated the polymerization of meso-lactide using N-heterocyclic carbenes (15a–c, Scheme 9).²⁰ Compounds 15a and 15b afforded heterotactic PLA (P_s = 0.33 and 0.17, respectively) at −40 °C in dichloromethane (monomer : initiator ratio of 100). Complex 13c produced practically atactic PLAs (P_s = 0.58).

Scheme 9 NHC used for polymerization of meso-lactide.

Mechanistic studies

In the early 1990s, Kricheldorf showed that meso-lactide was polymerized more slowly than rac-lactide when using aluminium and magnesium initiators.⁴ In 2000, Chisholm et al. analyzed the molecular structure of meso-lactide and its polymerization by single-site initiators.²¹ They showed preferential polymerization of meso-lactide over L-lactide and rac-lactide. At −40 °C, zinc and magnesium tris(pyrazolyl)hydroborate complexes exclusively polymerized meso-lactide in a 50 : 50 mixture of meso- and rac-lactide, leaving rac-lactide unreacted. Furthermore, according to a computational study the ground state conformation of meso-lactide is between a twist-boat and a planar structure, and thus the six-membered ring is close to planar. Calculations suggested that the planar structure is favored because the lone pairs of the two oxygen atoms in each ester group repel each other (Fig. 1). This supports the view that meso-lactide is less stable than rac-lactide.²²

Fig. 1 Molecular structures of meso-lactide according to ref. 21: (a) side-on view and (b) side view.

Polymerization of meso-lactide by the chiral aluminium initiator (R)-1 (Scheme 3) proceeded with a slow rate (k_obs = 2.0 × 10⁻³ min⁻¹). Interestingly, this is approximately half the rate of the formation of the heterotactic PLA, prepared from meso-lactide using rac-1 (k_obs = 4.4 × 10⁻³ min⁻¹). This result can be rationalized on the basis that (R)-1 can only attack meso-lactide at one of the two diastereotopic acyl-oxygen bonds (Scheme 10).⁷

Under the same conditions, aluminium complexes 3a–c (Scheme 3) polymerized meso-lactide with various rates (k_obs of 29.4, 4.6 and 13.5 × 10⁻³ min⁻¹, respectively).⁹ When the complexes 3a and 3c (R = R′) were used, rac-lactide was polymerized ca. 1.3 times faster than meso-lactide. However, using complex 3b (R ≠ R′), the polymerization of L-lactide is ca. 2.1 fold faster than rac-lactide and 5 times faster than meso-lactide; demonstrating a direct effect of the substituent on the mechanism of the polymerization. This was attributed to the chain-end selection caused by the presence of both stereogenic centers.⁹

When indium complex 8 (Scheme 5) was used, the polymerization rates for rac-lactide and meso-lactide (k_obs = 13.8 × 10⁻³ min⁻¹) were similar and approximately 10 times higher than that for L-lactide. These findings support a mechanism where both the rate of polymerization and selectivity of monomer enchainment are influenced by multiple stereocenters in the monomer and/or the polymer chain end.¹⁵

The polymerization of meso-lactide (k_obs = 0.8 × 10⁻³ min⁻¹) using titanium complex 10e demonstrated a rate comparable to that of rac-lactide which was ca. 2.2 times slower than L-lactide polymerization. This was explained by a reaction site with a preference for L-lactide.¹⁶ The asymmetric chlorinated analog titanium complex 11b indicated a polymerization of meso-lactide 4 times slower (k_obs = 0.2 × 10⁻³ min⁻¹) than when complex 10e was used.¹⁶

When the tetravalent complexes containing two (OSSO)-type ligand (12a–c) were compared, the cerium complex 12c showed the fastest polymerization rate for meso-lactide (k_obs = 3.2 × 10⁻³ min⁻¹) and ca. 2.1 times faster than rac- and L-lactide. Similar tendency was found using hafnium 12b, meso-lactide (k_obs = 2.8 × 10⁻⁴ min⁻¹) was polymerized twice as fast compared to rac- and L-lactide. However, zirconium complex 12a polymerized rac-lactide and L-lactide approximately 20 times faster than meso-lactide (k_obs = 1.1 × 10⁻⁴ min⁻¹).¹⁷

At room temperature, zinc complex 13a polymerized rac-lactideca. 1.6 times faster than meso-lactide (k_obs = 37.0 × 10⁻³ min⁻¹) (Table 5).¹⁸

Table 5 Rate constants of the polymerization of meso-lactide

Initiator	[Initiator]0/mol L^−1	[mon.]0/[init.]0	Solvent	k obs/10^−3 min^−1	T/°C	Ref.
R-1	0.200	100	Toluene	2.0	70	7
rac-1	0.200	100	Toluene	4.4	70	7
3a	0.474	96	Toluene	29.4	70	9
3b	0.474	96	Toluene	4.6	70	9
3c	0.474	96	Toluene	13.5	70	9
8	0.840	100	CD2Cl2	13.8	21	15
10e	0.520	100	C6D6	0.8	100	16
11b	0.520	100	C6D6	0.2	100	16
12a	0.520	100	C6D6	0.1	100	17
12b	0.520	100	C6D6	0.3	100	17
12c	0.520	100	C6D6	3	100	17
13a	1.023	494	CH2Cl2	37.0	25	18

---

Conclusions and future perspectives

Although meso-lactide has been known for 20 years,²³ the number of stereoselective single-site initiators is still small compared to the plethora of well-defined initiators for L- and rac-lactide. Peculiarly, only two examples of melting temperature for poly(meso-lactide) have been reported so far due to the lack of highly syndiotactic PLAs.^6,11

In contrast, when rac-lactide is polymerized, chirality or rigidity of the backbone of the ligand was not necessary to obtain high tacticity.^11,18 In most cases, metal complexes which polymerized rac-lactide to give highly heterotactic PLAs, also yielded highly syndiotactic PLAs from meso-lactide.^10,11,18 Trivalent metal complexes appeared to afford higher syndiotacticity than the divalent or tetravalent metal complexes. Furthermore, metals (aluminium^6,7 and indium¹¹) having slower rates of polymerization than group 3 metals^10,11 resulted in better polydispersities of the poly(meso-lactide).

Scheme 10 Formation of syndiotactic PLA according to ref. 6 and 7.

Since meso-lactide is a by-product of L-lactide synthesis, it will become more easily available with increasing capacity of L-lactide production.²³ Therefore, efficient and selective polymerization will be of considerable interest, provided crystalline poly(meso-lactides) can be obtained. This will give large benefits as it will be possible to use L-lactide to form isotactic PLAs and its by-product (meso-lactide) to form syndiotactic PLAs reducing the cost of production and increasing the interest towards biorenewable poly(lactic acid).

Acknowledgements

We thank the Deutsche Forschungsgemeinschaft, the Fonds der Chemischen Industrie for financial support, and Uhde Inventa-Fischer for a generous gift of meso-lactide.

Notes and references

1. (a) R. E. Drumright, P. R. Gruber and D. E. Henton, Adv. Mater., 2000, 12, 1841; (b) S. Inkinen, M. Hakkarainen, A.-C. Albertsson and A. Södergård, Biomacromolecules, 2011, 12, 523; (c) O. Dechy-Cabaret, B. Martin-Vaca and D. Bourissou, Chem. Rev., 2004, 104, 6147; (d) J. K. Oh, Soft Matter, 2011, 7, 5096.
2. (a) P. J. Dijkstra, H. Du and J. Feijen, Polym. Chem., 2011, 2, 520; (b) C. M. Thomas, Chem. Soc. Rev., 2010, 39, 165; (c) M. J. Stanford and A. P. Dove, Chem. Soc. Rev., 2010, 39, 486; (d) C. A. Wheaton, P. G. Hayes and B. J. Ireland, Dalton Trans., 2009, 4832; (e) R. H. Platel, L. M. Hodgson and C. K. Williams, Polym. Rev., 2008, 48, 11; (f) A. Amgoune, C. M. Thomas and J.-F. Carpentier, Pure Appl. Chem., 2007, 79, 2013; (g) J. Wu, T.-L. Yu, C.-T. Chen and C.-C. Lin, Coord. Chem. Rev., 2006, 250, 602; (h) B. J. O'Keefe, M. A. Hillmyer and W. B. Tolman, J. Chem. Soc., Dalton Trans., 2001, 2215.
3. K. Fukushima and Y. Kimura, Polym. Int., 2006, 55, 626.
4. (a) H. R. Kricheldorf and C. Boettcher, Makromol. Chem., 1993, 194, 1653; (b) H. R. Kricheldorf and S.-R. Lee, Polymer, 1995, 36, 2995; (c) H. R. Kricheldorf and C. Boettcher, J. Macromol. Sci., Part A: Pure Appl. Chem., 1993, 30, 441.
5. N. Spassky, M. Wisniewski, C. Pluta and A. Le Borgne, Macromol. Chem. Phys., 1996, 197, 2627.
6. T. M. Ovitt and G. W. Coates, J. Am. Chem. Soc., 1999, 121, 4072.
7. T. M. Ovitt and G. W. Coates, J. Am. Chem. Soc., 2002, 124, 1316.
8. H. Du, A. H. Velders, P. J. Dijkstra, J. Sun, Z. Zhong, X. Chen and J. Feijen, Chem.–Eur. J., 2009, 15, 9836.
9. H. Du, A. H. Velders, P. J. Dijkstra, Z. Zhong, X. Chen and J. Feijen, Macromolecules, 2009, 42, 1058.
10. A. Amgoune, C. M. Thomas, T. Roisnel and J.-F. Carpentier, Chem.–Eur. J., 2006, 12, 169.
11. J.-C. Buffet, A. Kapelski and J. Okuda, Macromolecules, 2010, 43, 10201.
12. Molecular weight efficiency values, f, were calculated according to M_n,exp/M_n,theo, with M_n,theo = [mon]₀/[initiator]₀ × 144.13 × conv.
13. J.-C. Buffet and J. Okuda, Dalton Trans., 2011, 40, 7748.
14. A. Pietrangelo, M. A. Hillmyer and W. B. Tolman, Chem. Commun., 2009, 2736.
15. A. Pietrangelo, S. C. Knight, A. K. Gupta, L. J. Yao, M. A. Hillmyer and W. B. Tolman, J. Am. Chem. Soc., 2010, 132, 11649.
16. J.-C. Buffet and J. Okuda, Chem. Commun., 2011, 47, 4796.
17. A. Sauer, J.-C. Buffet, T. P. Spaniol, H. Nagae, K. Mashima, and J. Okuda, unpublished results.
18. B. M. Chamberlain, M. Cheng, D. R. Moore, T. M. Ovitt, E. B. Lobkovsky and G. W. Coates, J. Am. Chem. Soc., 2001, 123, 3229.
19. J.-C. Buffet, J. P. Davin, T. P. Spaniol and J. Okuda, New J. Chem., 2011 10.1039/c1nj20259f.
20. A. P. Dove, H. Li, R. C. Pratt, B. G. G. Lohmeijer, D. A. Culkin, R. M. Waymouth and J. L. Hedrick, Chem. Commun., 2006, 2881.
21. M. H. Chisholm, N. W. Eilerts, J. C. Huffman, S. S. Iyer, M. Pacold and K. Phomphrai, J. Am. Chem. Soc., 2000, 122, 11845.
22. I. A. Shuklov, H. Jiao, J. Schulze, W. Tietz, K. Kühlein and A. Börner, Tetrahedron Lett., 2011, 52, 1027.
23. G. Entenmann, and D. Bendix, Boehringer Ingelheim, Ger. Offen. DE 3820299, 1988; G. Entenmann and D. Bendix, Chem. Abstr., 1989, 110, 213592w.

---