Lihua
Guo
a,
Wenjing
Liu
a and
Changle
Chen
*b
aSchool of Chemistry and Chemical Engineering, Qufu Normal University, Qufu, 273165, China
bCAS Key Laboratory of Soft Matter Chemistry, Department of Polymer Science and Engineering, University of Science and Technology of China, Hefei, 230026, China. E-mail: changle@ustc.edu.cn
First published on 29th August 2017
In this review, recent developments on late transition metal catalyzed α-olefin polymerization and copolymerization with polar comonomers are described. First, the polymerization mechanisms of early and late transition metal catalyzed α-olefin polymerization are compared. Second, cationic catalysts bearing α-diimine and related ligands as well as neutral catalysts bearing anionic ligands for α-olefin homopolymerization are discussed in detail. Third, late transition metal catalyzed α-olefin copolymerization with polar functionalized comonomers is summarized. Special attention is paid to the regio- and stereo-selectivity induced by late transition metal catalysts in these polymerization and copolymerization reactions.
The non-polar nature of polyolefins represents one of their biggest limitations. The introduction of some polar functional groups could greatly improve their many properties and further expand their applications.17 As the most direct and economic strategy, transition metal catalyzed olefin copolymerization with polar monomers is highly fascinating but also highly challenging. The industrially relevant early transition metal catalysts are incompetent for this task because they are easily poisoned by polar functional groups. Because of their low oxophilicity, late transition metal catalysts have achieved great success in the area of ethylene copolymerization with polar monomers. However, most of these catalysts are not able to realize regio- or stereo-controlled α-olefin polymerization. Moreover, very few efforts have been made in developing efficient catalytic systems for α-olefin/polar monomer copolymerization. As a result, the synthetic approach for functional polypropylenes and poly(α-olefins) is generally limited to multi-step techniques such as post-polymerization functionalization or the masking/protection of polar monomers. The direct synthesis of polar functionalized polypropylenes from the transition metal catalyzed copolymerization of propylene and polar monomers remains a big challenge in this field.
In the 1990s, Brookhart et al. reported that α-diimine Pd(II) complexes could copolymerize olefin with polar monomers.18–21 A key insight from these early studies is that increasing steric bulk on the metal axial position could efficiently retard chain transfer and lead to high molecular weight polymers. Since then, tremendous progress has been made and a huge amount of high performance late transition metal catalysts have been designed and synthesized.22–24 The use of late transition metal catalysts for the coordination–insertion polymerization of ethylene, cycloolefins, dienes, trienes, and norbonene type monomers and copolymerization with polar monomers has been extensively studied3–15 and is beyond the scope of this review. In contrast, only a small fraction of these late transition metal catalysts have been investigated in α-olefin polymerization or copolymerization reactions. This review will focus on recent investigations on late transition metal catalyzed propylene and higher α-olefin homopolymerization and copolymerization with polar monomers. These catalysts can be categorized into two groups based on the nature of the active species during polymerization: cationic catalysts and neutral catalysts.
In this review, we will first discuss the mechanism of transition metal catalyzed ethylene and α-olefin polymerization. Subsequently, the cationic catalysts based on diimine ligands and other ligands, as well as neutral catalysts bearing phosphine–sulfonate and derivative ligands for α-olefin homopolymerization and copolymerization with polar monomers will be discussed. Specifically, the control of polymer microstructures including branching structure, regioselectivity and stereoselectivity will be focused on.
α-Olefin polymerization catalyzed by early transition metal catalysts usually exhibits highly selective 1,2-insertion of the monomer, in which the terminal or “1” carbon of the α-olefin is directly bound to the metal center and the neighboring “2” carbon is attached to the incipient polymer chain, resulting in poly(α-olefins) with n-alkyl side chains (Scheme 2a).25–27 The occasional 2,1-insertion can lead to a 2,1-enchainment or 1,ω-enchainment, which has been observed as isolated regioerrors.28
In contrast, polymerization of α-olefins using cationic α-diimine Ni(II) and Pd(II) catalysts occurs through a chain straightening mechanism to give polymers that contain fewer branches per 1000 carbons than predicted from simple 1,2-monomer enchainment (Scheme 2b).21,29,30 The monomer undergoes 1,2- as well as 2,1-insertion into the M–C bond and the metal usually walks to the end of the branch before the next insertion happens. When 1,2-insertion takes place, the subsequent β-hydride elimination followed by metal migration to the primary carbon atom leads to 2,ω-enchainment to form a polymer chain with a methyl branch. On the other hand, 2,1-insertion followed by chain walking results in 1,ω-enchainment to give repeating units with a linear structure.
Catalyst | T (°C) | Time (h) | Olefin (conc.)a | TOFb (h−1) | M n (×10−3) | PDIc | B /1000C | Thermal analysise (°C) | Ref. |
---|---|---|---|---|---|---|---|---|---|
a P = propylene, H = 1-hexene, D = 1-decene. b Turnover frequency, which is calculated as the moles of monomer consumed per mole of catalyst per hour. c Determined by Gel Permeation Chromatography (GPC). d Branching density, branches per 1000 carbon atoms, determined by 1H NMR. nd: not determined. e Determined by Differential Scanning Calorimetry (DSC). f Broad and weak endotherms, amorphous. | |||||||||
C2 | 25 | 16 | P (1 atm) | 579 | 15 | 4.30 | 213 | −43 (Tg) | 18 |
C4 | 0 | 2 | P (1 atm) | 3005 | 150 | 1.60 | 300 | −20, −78 (Tg) | 18 |
C4 | 0 | 1.5 | H (0.8 M) | 907 | 74 | 1.15 | 83 | nd | 29 |
C5 | 0 | 2 | H (0.8 M) | 528 | 74 | 1.19 | 98 | nd | 29 |
C6 | −78 | 96 | P (15 g) | 1 | 5.7 | 1.37 | nd | −0.5 (Tg), 137.3 (Tm) | 37 |
C6 | 0 | 2 | P (5 g) | 385 | 48.2 | 1.08 | nd | −54.4 (Tg) | 37 |
C7 | 20 | 1 | P (1.2 atm) | 380 | 53.8 | 1.13 | 141 | 30.1 (Tm) | 33 |
C8 | 22 | 24 | D (0.1 M) | 0.9 | 10 | 1.30 | nd | 109 (Tm) | 43 |
C9 | 75 | 3 | H (2.66 M) | 1822 | 529 | 1.17 | 52 | −47 (Tg), 58 (Tm) | 44 |
C9 | 50 | 1 | P (1 atm) | 2032 | 133 | 1.13 | 105 | −55 (Tg) | 44 |
C11 | −20 | 2 | P (1 atm) | 12900 | 6.5 | 2.1 | nd | nd | 51 |
C12 | −60 | 8 | P (40 mL) | 13 | 6.4 | 1.30 | nd | 134 (Tm) | 53 |
C12 | −20 | 3 | P (20 mL) | 170 | 42 | 1.10 | nd | 76 (Tm) | 53 |
C13 | 25 | 8 | H (1.06 M) | 26.4 | 31.8 | 1.23 | 132.2 | —f | 54 |
C14 | 0 | 8 | H (1.06 M) | 5.2 | 5.2 | 1.40 | 17.1 | 93.4, 107.2 (Tm) | 54 |
C15 | 20 | 3 | D (0.1 M) | 4.5 | 13 | 1.49 | 26 | 105.5 (Tm) | 55 |
C19 (Dip-H) | 100 | 3 | P (10 g) | 808 | 16 | 2.20 | nd | nd | 59 |
C20 (Men-Ph) | 50 | 12 | P (6 g) | 140 | 21 | 1.90 | nd | −9.6 (Tg), 46.6, 63.9 (Tm) | 60 |
The activities of α-diimine Ni(II) and Pd(II) catalysts in propylene polymerization are lower than in ethylene polymerization. For the α-diimine Ni(II) catalyzed propylene polymerization, the rate-limiting step is the trapping of propylene, making the chain growth process overall first order in olefin concentration.31 In contrast, the chain propagation is limited by migratory insertion rates in the α-diimine Pd(II) system.18,20 Furthermore, these catalysts can polymerize α-olefin in a living fashion and afford block copolymers with α-olefins.20,32,33
The structure of the polymers depends on the structures of the α-diimine Ni(II) and Pd(II) catalysts and the polymerization conditions. Considerable research efforts have been made in this field. Pellecchia et al. found that the regioselectivity of the α-diimine Ni(II) catalysts is temperature dependent. At below −45 °C, C3 and C4 showed preferred 1,2-propylene insertion and afforded regioregular syndiotactic polypropylene with rr values of above 74%.34,35 However, regioirregular polypropylene was obtained at 0 °C.36
Brookhart et al. reported the effects of ligand symmetry on α-diimine Ni(II)-catalyzed 1-hexene polymerization in an effort to probe the possibility of realizing stereoselective α-olefin polymerization.29 The Ni(II) catalyst C4/MAO exhibited a high propensity for 1,2-insertion and the preference decreased with temperature increasing from 0 °C to 50 °C. In contrast, 2,1-insertion was preferred when the “di-t-butyl” catalyst C5/MAO was used (Scheme 3). This is mainly due to steric effects exerted from different ligand structures.
Coates et al. reported that C2-symmetric α-diimine Ni(II) catalyst C6 bearing a chiral sec-phenethyl moiety polymerized propylene and higher α-olefins in a living fashion and showed an interesting temperature-dependent regioselectivity (Scheme 4).37–41 The regioregular and highly isotactic polypropylene was obtained at low temperature (e.g. −78 °C) while regioirregular polypropylene with 1,2- and 3,1-enchainments was generated at higher temperature (e.g. 0 °C). An isotactic–atactic regioregular–regioirregular block copolymer can be synthesized by changing the reaction temperature during propylene polymerization. When this catalyst was used for higher α-olefin polymerization, poor regio-control was observed and an ethylene–propylene copolymer structure was obtained through 2,ω-enchainment.
As outlined in the mechanism shown in Scheme 2b, 1,ω-enchainment can lead to a polymer product with a highly linear microstructure. However, it is highly challenging to achieve this via precise α-olefin 2,1-insertion and a chain straightening mechanism using α-diimine Ni(II) and Pd(II) catalysts. Wu et al. reported the synthesis of polypropylene and poly(1-hexene) with an obvious Tm using camphyl based α-diimine Ni(II) catalyst C7.33,42 Coates et al. recently reported high selectivity of 1,ω-enchainment using C8 and some derivative catalysts bearing the “sandwich” type α-diimines.43 These catalysts polymerized higher α-olefins such as 1-decene to semicrystalline polyethylene-type polymers with linear units of up to 76% and melting points (Tm) of above 100 °C. Most interestingly, a similar polymer microstructure could be obtained using a mixture of α-olefins (Scheme 5). Guan et al. reported that cyclophane-based α-diimine Ni(II) catalyst C9 promoted living polymerizations of propylene and 1-hexene at elevated temperatures of up to 75 °C.44 The significantly lower level of the branching density in polypropylene and poly(1-hexene), which is only half of the value compared with traditional α-diimine Ni(II) catalysts, indicated higher selectivity for 1,ω-enchainment through 2,1-insertion of α-olefin and chain walking. Our group recently reported the tuning of polypropylene and poly(1-hexene) microstructures using a series of α-diimine Pd(II) catalysts C10 bearing both the dibenzhydryl moiety and systematically varied ligand sterics.45 These polymers are amorphous with no melting point, suggesting poor regio-control for the polymerization process.
In addition to linear α-olefin polymerization, the polymerization of branched α-olefins such as 4-methyl-1-pentene has also been investigated using conventional α-diimine Ni(II) and Pd(II) catalysts.46–48 The obtained polymers are amorphous elastomers or oils with various types of branches generated through the chain walking process.
Bazan et al. reported that α-keto-β-diimine Ni(II) catalyst C12 can realize stereo-controlled propylene polymerization at low temperature (Scheme 6).53 For example at −60 °C, the chain walking process can be significantly suppressed and isotactic polypropylenes can be obtained. At high temperatures, atactic and regioirregular polypropylenes with 1,3-enchainment errors were obtained.
The ligand structure in Ni(II) catalysts C13 and C14 can dictate the insertion fashion of 1-hexene (Scheme 7).54 Catalyst C13 with a tert-butyl substituent showed 80% selectivity of 1,2-insertion of 1-hexene, producing amorphous polymers with predominant methyl branches. In contrast, catalyst C14 with two methyl substituents showed 90% selectivity of 2,1-insertion of 1-hexene, producing semicrystalline “polyethylene” with melting points of up to 107 °C.
Our group recently reported that iminopyridyl Ni(II) catalyst C15 bearing both dibenzhydryl and naphthyl moieties can polymerize α-olefins with high selectivity of 1,ω-enchainment to afford polymers with melting points of up to 105 °C.55
Nozaki et al. recently reported that C19 and related catalysts bearing imidazo-[1,5-a]quinolin-9-olate-1-ylidene ligands can suppress chain walking and exhibit high 1,2-selectivity for propylene insertion, generating polypropylenes with regio-defects of less than 1.5 mol%.59 These catalysts were also effective for 1-butene polymerization and afforded poly(1-butene)s with regio-defects of less than 0.1 mol%. However, stereo-control was not achieved in this system.
Using C20 and related phosphine-sulfonate palladium catalysts, Nozaki et al. accomplished regio- and stereo-controlled homopolymerization of propylene at 50 °C.60 A moderate triad ratio (mm = 0.59) could be achieved in this system. It is worth noting that most previously reported catalysts based on group 10 metals such as α-diimine Ni(II) complexes can only achieve stereo-control at very low temperatures (e.g. −60 °C).
Catalyst | T (°C) | Time (h) | Propylene | Comonomer | Activity (kg mol−1 h−1) | M n (×10−3) | PDIa | Incorpb (mol%) | Thermal analysisc (°C) | Ref. |
---|---|---|---|---|---|---|---|---|---|---|
a Determined by GPC. b Determined by 1H or 13C NMR. c Determined by DSC. nd: not determined. | ||||||||||
C2 | 35 | 18.5 | 6 atm |
![]() |
2.7 | 37 | 1.8 | 0.6 | nd | 20 |
C11 | 30 | 16 | 5.7 g |
![]() |
7.2 | 0.97 | 2.1 | 1.4 | nd | 61 |
C11 | 30 | 16 | 7.0 g |
![]() |
4.5 | 1.3 | 7.8 | 2.1 | nd | 61 |
C11 | 30 | 16 | 7.3 g |
![]() |
2.1 | 2.0 | 1.9 | 1.7 | nd | 61 |
C11 | 0 | 16 | 11 g |
![]() |
1.3 | 3.2 | 2.4 | 0.3 | nd | 61 |
C19 | 100 | 6 | 10 g |
![]() |
2.8 | 11 | 2.3 | 0.9 | nd | 59 |
C19 | 100 | 6 | 10 g |
![]() |
5.4 | 13 | 2.5 | 0.5 | nd | 59 |
C19 | 100 | 6 | 10 g |
![]() |
0.74 | 3.0 | 3.0 | 1.1 | nd | 59 |
C20 | 50 | 12 | 6.0 g |
![]() |
0.38 | 5.8 | 2.6 | 2.9 | nd | 60 |
C20 | 50 | 12 | 6.0 g |
![]() |
0.79 | 7.5 | 1.9 | 1.5 | nd | 60 |
C20 | 50 | 12 | 6.0 g |
![]() |
0.58 | 11 | 2.2 | 1.7 | −10.0 (Tg), 42.9, 61.7 (Tm) | 60 |
C20 | 50 | 12 | 6.0 g |
![]() |
0.83 | 4.6 | 2.2 | 1.6 | −14.3 (Tg), 43.3, 61.8 (Tm) | 60 |
The typical cationic α-diimine Pd(II) catalyst C2 is effective for copolymerization of α-olefins with methyl acrylate.20 The majority of ester groups are located at the end of the branches in the resulting copolymers. The branching density of the copolymers is lower than the theoretical value from an exclusive 1,2-insertion due to the chain straightening mechanism, which is similar to the α-olefin homopolymerization scenario. The acrylate incorporation ratio is higher than that of the ethylene copolymers. For example, a 1-hexene to methyl acrylate ratio of 7:
1 at equal molar concentration of the two monomers was observed in 1-hexene/methyl acrylate copolymerization, whereas the ratio was ca. 150
:
1 for the ethylene case. In addition, the catalytic activity and the copolymer molecular weight are markedly suppressed in the presence of polar comonomers.
Our group recently reported that sterically bulky α-diimine palladium catalysts C10 could copolymerize 1-octene with biorenewable comonomer acrylic acid, with high comonomer incorporation and high copolymer molecular weights.45a This provides an alternative strategy to access branched carboxylic acid functionalized polyolefins.
Nozaki et al. reported C11 catalyzed copolymerization of propylene with allylbenzene and various allyl sily ethers.61 The allyl comonomers were incorporated mainly into the terminal chain-end position of the copolymers.
Despite the successful synthesis of various functional copolymers, the above-mentioned late transition metal catalysts suffer from poor regio- and stereo-control in olefin (co)polymerization mainly because of the chain straightening process. If this process can be suppressed using novel catalysts, polar functionalized isotactic copolymers may be synthesized.
Recently, Nozaki et al. reported regio-selective copolymerization of propylene with a series of polar monomers using Pd(II) catalysts C19.59 However, the obtained functional polypropylenes exhibited an atactic stereoregularity. Using the phosphine–sulfonate palladium catalysts C20, the same group achieved moderate isotacticity in the copolymerization of propylene with various polar monomers (see Scheme 8).60 Under optimized conditions, semicrystalline isotactic polar polypropylene was obtained bearing various functional groups such as acetoxy, cyano, chloro and alkoxycarbonyl moieties. This is a very nice demonstration of the great potentials of late transition metal catalysts in achieving regio- and stereo-controlled copolymerization of α-olefin with various polar monomers.
It is easy to achieve regio- and stereo-specific α-olefin polymerization using early transition metal catalysts, which is currently practiced on a large scale in industry. There has been a continuous interest in preparing polar functionalized iPP (isotactic polypropylene) through direct copolymerization of propylene with polar monomers. However, these early transition metal catalysts are easily poisoned by polar monomers. In this sense, late transition-metal catalysts are promising candidates because of their low oxophilicity. Unfortunately, most late transition-metal catalysts are prone to generating amorphous and atactic polymers as a result of poor regioselectivity, poor stereoselectivity and a fast chain-walking process. It is highly challenging to solve this dilemma. Meanwhile, this also represents a major driving force for future studies.
This journal is © the Partner Organisations 2017 |