Ethylene polymerization catalyzed by Ni and Pd catalysts is an area of great interest both in academic research and for industrial production. The structures of Ni and Pd catalysts can dramatically influence polymerization performance. In this review, the influence of newly reported Ni and Pd catalysts on ethylene polymerization, from ligand effects to metal–metal cooperativity, as well as the polymerization conditions, are discussed to provide fresh guidelines for rational design of high-performances catalysts, highly efficient study of polymerization mechanisms, and successful synthesis of PE variants with unique microstructures and properties.
After Ziegler-type and metallocene-type catalysts, late-transition-metal catalysts became highlights for olefin polymerization. Ni and Pd catalysts, as post-transition-metal catalysts, have been widely studied since researchers such as Brookhart[1] and Drent[2] carried out pioneering work on finding that these catalysts, bearing ligands of several classes, possess some unique properties such as high tolerance to polar monomers and production of polymers with topology from linear to branched and even hyperbranched. Obviously, the structures of these catalysts create their intriguing performances.
Recent progress has revealed special applications of various new Ni and Pd catalysts. Unusual polymers such as ultra-high-molecular-weight PE (UHMWPE),[3] elastomeric polymers,[4] and branched high-molecular-weight PE[4b], [5] (BHMWPE) can be obtained through ethylene polymerization catalyzed by these catalysts. PE containing linear short branches,[6] and hyperbranched–branched diblock polyolefins[7] have been synthesized by a “tandem catalysis” one-pot approach. More intriguingly, multiphase polyolefin particles of high-density PE (HDPE)/branched PE core-shell structure have been prepared successfully by tandem addition of two kinds of water-soluble catalysts[8] To illustrate the formation of HDPE nanocrystals in this system, Godin et al. studied the molecular mechanisms of catalyst activation and deactivation.[9]
Copolymerization of ethylene with polar monomers to synthesize functionalized PE is an attractive field for Ni and Pd catalysts. Pd catalysts can catalyze dimerization, trimerization, and (co)polymerization of polar olefins such as vinyl ethers.[10] Pd catalysts could also be used for copolymerization of higher α-olefins with polar monomers, affording functionalized copolymers with high incorporation of polar comonomers as well as high molecular weights.[11] It should be noted that insertion of polar groups into the nonpolar PE chain could endow the PE with diversified functions and make it more “colorful”.
In addition, many catalytic systems with special applications based on new Ni and Pd catalysts have been designed. Smith et al. reported a tantalizing catalytic system with in situ catalyst modification through metal capture[12] and provided a new model for catalyst design. Gladysz et al. designed a process termed “phase-transfer catalyst activation” to eliminate the negative effects of pre-coordinating bases and to avoid the use of cocatalyst.[13] Shu et al. used ligand-asymmetric homobimetallic catalyst to imitate “tandem catalysis” and heterobimetallic catalysis.[5]
With regard to characterization methods, electron paramagnetic resonance (EPR) spectroscopy is popular as a useful method for studying the mechanisms of Ni-catalyzed ethylene polymerization and the nature of active species.[14] Density functional theory (DFT) methods, as theoretical computational methods, have been widely used in combination with experimental results to develop polymerization mechanisms at the molecular level.[15] For data analysis, the response surface method (RSM) has been successfully applied to the modelling of the relationships between multiple operational parameters and results.[16]
For a catalytic system, the polymerization mechanism is a key factor for making newly designed systems understood in depth and widely acknowledged. However, not all catalytic systems can be assigned reasonable mechanisms, due to internal complex interaction. Li et al., for example, explored the types of special catalytic performance that could be achieved in the presence of reactive hydrogen atoms in the ligands of a series of Ni catalysts, but failed to uncover the exact functions of the hydrogen atoms, thus leaving many puzzles unsolved.[17] Therefore, simple catalytic systems are more useful and should be recommended for mechanistic studies, and investigations into Ni- and Pd-catalyzed systems are still indispensable and meaningful.
Although there have been several reviews about Ni and Pd catalysts in recent years,[18] there are only a limited number focusing on the relationships between properties of Ni- and Pd-catalyzed systems (structures of catalysts and conditions of systems) and the characters of the resulting polymers (microstructures etc.) for synthesizing PE with unique properties or applications. Furthermore, the synthesis of catalysts is a time-consuming and costly procedure accompanied by a high risk of failure. In this context, we aim to provide some guidelines for rational design of high-performance catalytic systems and successful synthesis of unique PE variants for future research. Specifically, the structural effects of catalysts, including steric hindrance by ligands, electronic effects of ligands, secondary coordination sphere effects, and metal–metal cooperativity effects are discussed broadly. To summarize the highlights of this field and to study the laws of the catalytic processes more efficiently, this review focuses on ethylene polymerizations (excluding ethylene oligomerization) and on a series of representative references reported recently. Nevertheless, some representative copolymerizations are also included because of their high research value.
Commonly, neutral Ni catalysts are highly attractive because their tolerance towards polar functional groups is higher than that of their cationic Ni counterparts. It has already been explained that the low electrophilicity weakens the chelation between polar groups and the metal center, thus facilitating coordination and insertion of olefinic monomers.[3], [19] However, the catalytic activities of Ni catalysts are usually consistent with the trend in net charges observed by X-ray photoelectron spectroscopy (XPS), which shows that a high net charge on the Ni leads to high catalytic activity.[20]
Obviously, there are some differences between Ni and Pd catalysts. Firstly, fewer ligands can be used to ligate Pd metal to form active Pd complexes for polymerization of ethylene than is the case with Ni complexes, because of the more electron-rich metal center.[21] Secondly, it is widely accepted that Pd catalysts offer more potential for chain walking in ethylene polymerization.[22] Thus, Pd catalysts usually produce PE with short branches and high branching density, corresponding to fast chain walking along the growing polymer chain, whereas Ni catalysts tend to produce PE with longer side chains. Recently, researchers have verified that β-H elimination in Pd-catalyzed polymerization is faster than that in Ni-catalyzed polymerization by Eyring analysis, which shows Pd agostic complexes having a lower enthalpy of activation for β-H elimination than their Ni counterparts.[23] Intriguingly, Zhang et al. proposed that γ-agostic, δ-agostic, or greater interaction between the metal center and the growing chain might exist in Ni-catalyzed processes, because of the discovery of PE with long chain branches and internal vinylene groups.[24] In other words, the differences between Ni and Pd analogues can be utilized to synthesize PE variants with different microstructures and functions.
For ethylene polymerization catalysts, it can sometimes be observed that a desirable catalytic property can be achieved only when an appropriate ligand is combined with a particular metal, whereas many other ligands may fail.[25] Because this is still a puzzle remaining to be solved, further research into revealing the “action rules” of ligands is necessary for the development of new, high-performance catalysts.
Ligands do have critical influences on the ethylene polymerization process, thus greatly influencing the microstructures of the resulting polymers.[26] Ligands such as α-diimines,[4f], [11], [15b], [16], [17], [27] phosphine-sulfonates,[6], [15a], [28] salicylaldimines,[3], [12], [15d], [19a], [29] and iminopyridines,[30] forming five- or six-membered chelate rings with the metal center are still common for catalyst design because of their higher stability, and modification of ligands in substituents and backbone is a common – but not easy – way to design new, high-performance catalysts. In fact, there are still some examples of failed attempts to design catalysts with special performances, as can be seen from recent reports.[31]
Electronic and steric effects of ligands are two main factors influencing the behavior of catalysts, as can be observed in previous reports. Electronic effects mainly originate from the electron-donating or electron-withdrawing natures of substituents whereas steric effects are related to interatomic distances and angles, the sizes of substituents, the ionic radii of the transition metals, and so on.[15c] Usually, bulky substituents near the metal can block the axial site of the metal center, thus suppressing chain transfer[19b], [26], [27l] and affording polymers with higher molecular weights.[16] Electronic perturbation of ligands can modify the net charge at the metal center, to yield polymers with a variety of microstructures (including changing the insertion locations of polar comonomers for copolymers) and can markedly influence the activities of ethylene polymerization.
From the point of view of coordinating groups (atoms), most studies have found that bromide pre-catalysts show higher activities than their chloride analogues, which has been explained in terms of the bromide analogues having higher solubility.[4a], [17], [32] The evidence, however, is still not so sufficient to make such a conclusion.
In the following sections, the influence of Ni and Pd catalysts on ethylene polymerization is discussed in terms of electronic and steric effects of ligands, secondary coordination sphere effects, and metal–metal cooperativity. However, the performances of catalysts cannot usually be explained in terms of just one of these effects, but are more likely the results of combinations of several.[27l], [27t]
In the case of Ni catalysts, electronic perturbation of many ligands usually has very limited effects on the performance of ethylene polymerization. In some cases the electronic effects are moderate, unless the electron-donating or -withdrawing substituents are quite near the Ni center.[19a], [27a] However, there are still some reports that show that the catalytic activities of Ni catalysts can be enhanced in terms of increasing electronegativity of ligands.
The catalytic performances of Pd catalysts, though, can be dramatically influenced by electronic effects. Notably, Pd catalysts with α-diimine ligands usually afford highly branched copolymers with the comonomers located at the ends of the branches in catalysis of copolymerization of α-olefins and polar comonomers,[1c], [1e] whereas those bearing electronically asymmetrical ligands such as phosphine-sulfonate can produce linear functionalized copolymers.[2], [33]
Redox behavior is also a form of electronic perturbation. Intriguing results – such as changes in the microstructures of the resulting polymer chains[27o], [27v] or in the activities of catalysts[28d] – can be observed when redox-active ligands are introduced into metal catalysts. Studies of the effects of the redox behavior of bis(N-arylimino)acenaphthene (BIAN) ligands are still meaningful to illustrate the mechanism of ethylene polymerization catalyzed by catalysts with redox-active BIAN ligands.[14e]
Besides experimental methods, the use of DFT methods to study the electronic structures of catalysts has become commonplace and useful. For example, relating calculated net charges on the metal center (allowing for modification by the electron-withdrawing substituents) with the polymerization mechanism can make it easier to understand the electronic effects of ligands.[15d]
On the other hand, with steric tuning of ligands through the design of substituents with different bulkiness, asymmetric/symmetric ligand structures,[34] or different geometric structures of chelate rings,[21], [26] β-H elimination and chain transfer[15b], [27e], [27j] can be tuned to some degree, so the microstructures of the resulting polymers, such as their branching distribution and branching density,[27m] can be controlled over a wide range, as well as the polymer molecular weight. Ethylene polymerization activities, however, are not usually consistent with trends of ligand bulk.[27j]
In this section, recently reported mononuclear catalysts with strong electronic or steric effects are discussed according to ligand classification.
To the best of our knowledge, α-diimine and iminopyridine ligands are the two main types of ligands containing N–N chelating moieties used for catalysts. Catalysts with α-diimine ligands usually show higher activity for ethylene polymerization and yield PE with higher molecular weight (Table 1). In addition, the β-diketiminate ligand (Figure 1) is not common in studies of the electronic and steric effects of ligands but the two reported catalysts do show some different influences on the microstructures of the resulting polymers.[35]
| Catalyst | Cocatalyst/Ni (or Pd) | T[°C] | P[bar] | Act.a | Mwb (× 10–4) | Mw/Mnb | Bc | Tmd[°C] | Ref. |
|---|---|---|---|---|---|---|---|---|---|
|
|||||||||
| 2a | MAO,e 3000 | 30 | 10.13 | 14.88 | 73.3 | 2.3 | – | 51.3 | [27r] |
| 2b | MAO, 3000 | 30 | 10.13 | 7.19 | 150 | 2.1 | 202 | 44.6 | [27r] |
| 2c | EASC,f 500 | 50 | 10.13 | 16.48 | 4.74 | 2.1 | – | 88.9 | [27q] |
| 2d | EASC, 600 | 30 | 10.13 | 10.50 | 66.3 | 2.3 | – | 69.8 | [4a] |
| 2e | EASC, 600 | 30 | 10.13 | 21.95 | 58.6 | 2.4 | 85 | 49.4 | [4a] |
| 2f | MAO, 2500 | 30 | 10.13 | 13.6 | 20 | 2.1 | – | 101.8 | [27s] |
| 2g | MAO, 2500 | 30 | 10.13 | 5.03 | 100 | 2.1 | – | 116.8 | [27s] |
| 2h | MAO, 2000 | 30 | 10.13 | 12.14 | 28.6 | 2.56 | – | 54.3 | [27t] |
| 2i | Et2AlCl,g 500 | 20 | 10.13 | 4.01 | 153 | 2.3 | – | 110.9 | [27u] |
| 2j | EASC, 700 | 30 | 10.13 | 17.45 | 3.68 | 2.38 | – | 50.9 | [4c] |
| 2k | Et2AlCl, 600 | 30 | 10.13 | 10.62 | 5.0 | 3.1 | – | 52.5 | [4b] |
| 2l | Me2AlCl,h 700 | 30 | 10.13 | 10.70 | 4.44 | 2.1 | – | 59.6 | [4f] |
| 3a | Et2AlCl, 600 | 100 | 9.12 | 6.18 | 123.0 | 2.14 | 62 | 41.5 | [27e] |
| 3c | Et2AlCl, 600 | 100 | 9.12 | 3.76 | 238.6 | 1.52 | 66 | 45.3 | [27e] |
| 4d | NaBARF,i 1.2 | 40 | 9.12 | 2.86 | 61.33 | 1.14 | 25 | 98 | [27i] |
| 5a | Et2AlCl, 600 | 30 | 0.2 | 0.788 | 20.1 | 1.81 | 119 | – | [27g] |
| 6a | MAO, 500 | 20 | 9.12 | 16.2 | 17.9 | 2.18 | 91 | 49 | [27a] |
| 6b | MAO, 500 | 20 | 9.12 | 3.6 | 84.0 | 2.01 | 56 | 75 | [27a] |
| 7a | MMAO,j 600 | 20 | 0.2 | 3.13 | 30.10 | 1.71 | 113 | – | [27l] |
| 7b | MMAO, 600 | 20 | 0.2 | 3.12 | 34.37 | 1.92 | 125 | – | [27l] |
| 7c | EASC, 400 | 20 | 10.13 | 9.82 | 22.1 | 2.90 | – | 125.5 | [27n] |
| 8b | NaBARF, 1.2 | 60 | 8.11 | 4.096 | 42.05 | 1.23 | 20 | 119.0 | [27f] |
| 9a | Et2AlCl, 600 | 30 | 10.13 | 9.24 | 18.8 | 2.69 | – | 124.3 | [27p] |
| 10a | Et2AlCl, 1000 | 20 | 8.11 | 4.3 | 14.77 | 1.78 | – | 67.8 | [30b] |
Structures of catalysts 1a and 1b containing β-diketiminate ligands.[35]
Recent reports showed that electron-withdrawing groups could enhance the activities of catalysts in particular cases (Table 1). Sun et al. synthesized catalysts 2a–i (Figure 2), a series of α-diimine Ni catalysts with some electron-withdrawing groups (F, Cl, NO2) introduced onto the imine-linked aryl ligands. These exhibit high activities up to 107 g of PE (mol of Ni)–1 h–1.[4a], [27q], [27r], [27s], [27t] Chen et al. reported that Ni catalyst 3a (Figure 2), containing the electron-withdrawing CF3 group, shows the highest activity in the series of catalysts 3a–d.[27e] Antonov et al. reported a series of Ni catalysts with 2-(imino)pyridine ligands containing electron-withdrawing groups and showing activity up to 6600 kg product (mol of Ni)–1 h–1 bar–1, which has been the highest activity for reported 2-(imino)pyridine Ni catalysts.[30a] However, another series of Pd catalysts reported by Chen's group show catalyst 4d (Figure 2), with the electron-donating OMe group at the para-position of the N-aryl system, with the highest activity.[27i] Certainly the metal center makes a big difference in these two series of catalysts.[27e], [27i]
Structures of catalysts 2a–l, 3a–d, 4a–d, 5a–d, 6a, 6b, 7a–c, 8a, 8b, 9a–e, 10a–c, and 11a–d.[4f], [27a], [27i], [27l], [27n], [27p], [27q], [27r], [27s], [27t], [27u]
Inspired by iminopyridine ligands, ligands containing cyclic imine components (e.g., Figure 3) have recently been widely studied in terms of backbone geometry and ring tension but just in the case of Ni catalysts.[20], [32], [36] This series of catalysts exhibited moderate activities below 107 g of PE (mol of Ni)–1 h–1.
Structures of ligands containing cyclic imine units.[20], [32], [36]
Polymerizations catalyzed by catalysts containing distorted planar chelate ring systems such as α-diamine[26] (e.g., catalyst 12, as shown in Figure 4) and amine-imine[21] (e.g., catalyst 13, Figure 4) ligands or a bulky backbone[27d] and bulky substituents at the ortho-aryl position[27f], [27j] (e.g., catalysts 14 and 15, Figure 4) show “living” character at room temperature, thanks to their special structure near the metal center, whereas almost all α-diimine catalysts show similar characteristics at low temperature.[37] To the best of our knowledge, ligand modification to increase axial shielding and then suppress the chain-transfer process is the key factor in these cases. Therefore, these phenomena may give rise to new strategies for catalysts design for well-controlled polymer chains.
Structures of catalysts 12–15 with “living” character at room temperature.[21], [26], [27d], [27j]
Other nitrogen heterocyclic (such as pyrazole, pyrrole, pyrimidine, indazole) catalysts containing N–N ligands[38] need further structural modifications to achieve high performances.
However, the relationship between polymerization activities and steric effects of ligands is still unclear. In some cases, substituents with less steric hindrance at the ortho-aryl position[20], [27p], [27q], [27r], [27s] or the backbone (substituents at the imine carbon atom, e.g., catalysts 11a–d, Figure 2)[30a], [30c] have positive influence on activities, whereas other cases show that bulky ligands[27a], [27f], [27g], [27i], [27m], [34] can enhance the activities of catalysts. Interestingly, there are cases in which particular sizes of the substituents on the N-aryl ligand are needed in order to obtain high activities.[26], [36c] It is possible that the unique structures of these catalysts facilitate chain propagation to some extent.
Notably, several catalysts containing α-diimine ligands show high performance for copolymerization of ethylene with polar comonomers (Table 2). Obviously, the specific electronic perturbation[27a] and/or steric hindrance[27d], [27f] around the metal centers delivers some positive effects on their catalytic behavior.
| Catalyst | T[°C] | PE[bar] | Comonomer [m] | Act.a | Incorp.b[%] | Mwc(× 10–3) | Mw/Mnc | Bd | Tme[°C] | Ref. |
|---|---|---|---|---|---|---|---|---|---|---|
|
||||||||||
| 6a | 30 | 1.013 | MA (1) | 0.56 | 7.9 | 20.9 | 1.70 | 112 | – | [27a] |
| 8a | – | 8.104 | MA (2.4) | 12.4 | 0.61 | 742.6 | 1.68 | 18 | 120.0 | [27f] |
| 14 | 20 | 1.722 | methyl undec-10-enoate (0.088) | 13 | 1.0 | 98.0 | 1.58 | – | 97 | [27d] |
Electronically nonsymmetrical ligands such as phosphine-sulfonate and bisphosphine monoxide (BPMO) are two main types of P–O ligands widely studied recently. Catalysts with such electronically asymmetrical ligands can generate linear polymer chains. Although ethylene/polar monomers copolymerization to yield functionalized linear polymers[1d], [2], [39] is the main goal of designing catalysts of these series, their performances in ethylene polymerization are still the fundamental evaluation criteria for catalytic properties.
Similarly to the catalysts containing N–N ligands, the phosphine-sulfonate ligand with different bulky substituents shielding the axial positions of the metal, as well as substituents with different electronic properties, can influence the behavior of catalysts remarkably. In addition, labile coordinating bases[28c] can prompt the activation process and decrease competition between coordinating bases and ethylene in the coordination process, thus enhancing ethylene polymerization activities. For reference, some high-performance catalysts containing P–O chelating ligands are listed in Table 3 and Table 4, for ethylene homopolymerization and copolymerization with polar comonomers, respectively.
| Catalyst | T [°C] | P [bar] | Act.b | Mwc (× 10–4) | Mw/Mnc | B d | Tme [°C] | Ref. |
|---|---|---|---|---|---|---|---|---|
|
||||||||
| 16c | 80 | 8.11 | 4.5 | 1.27 | 2.70 | – | 126.1 | [28c] |
| 18b | 110 | 9.12 | 7.5 | 0.85 | 2.48 | 7 | 124 | [28a] |
| 19b | 100 | 9.12 | 13.4 | 6.08 | 3.80 | – | 123.7 | [40] |
| Catalyst | T[°C] | PE[bar] | Comonomer [m] | Act.a | Incorp.b[%] | Mwc(× 10–3) | Mw/Mnc | Tmd[°C] | Ref. |
|---|---|---|---|---|---|---|---|---|---|
|
|||||||||
| 17a | 80 | 8.104 | NB–NP (0.8), NB–OAc (0.8) | 343 | NB–NP (1.1), NB–OAc (8.0) | 162.6 | 2.04 | 102 | [28e] |
| 17b | 80 | 8.104 | MA (2) | 7.9 | 12.6 | 67.54 | 1.68 | 77.3 | [28f] |
| 17c | 80 | 8.104 | MA (1) | 2.75 | 6 | 25.06 | 1.74 | 95.9 | [28g] |
| 18b | 85 | 5 | MA (2.5) | 6 | 40 | 3.6 | 1.89 | – | [28a] |
| 19b | 100 | 9.117 | MA (1) | 6.5 | 33 | 16.44 | 1.37 | – | [40] |
| 23b | 70 | 30 | MA (0.042) | 86 | 4.5 | 136 | 2.0 | 104 | [41] |
For electronic effects of ligands, Chen et al. designed catalyst 16c (Figure 6, below) with a strongly electron-withdrawing C6F5 substituent at the position ortho to the sulfonic group. It exhibits the highest ethylene polymerization activity in the series of phosphine-sulfonate catalysts 16a–e.[28c] Intriguingly, Pd catalyst 17a (Figure 6, below), bearing a ligand similar to that in 16a and 16b, can exhibit high activity in terpolymerization of ethylene with functionalized norbornenes such as NB-OAc and NB-NP (Figure 5).[28e] Chen's group found that catalysts 17b–c with suitable side groups, similarly to catalyst 17a, show high performance in copolymerization (Table 4).[28f], [28g] Chen's group also reported a phosphine-sulfonate ligand with a naphthalene backbone which could potentially have influence through both electronic and steric perturbation on the performances of catalysts 18a–c(Figure 6).[28a] The catalysts with the naphthalene backbone have much higher stability and activities than their analogies with the benzene backbone.[42] In addition, out of the catalysts with the naphthalene backbone, catalyst 18b, with the o-OMe-C6H4 substituent on the P atom, is one of the most active phosphine-sulfonate Pd catalysts for ethylene polymerization (Table 3),[28a] which might be due to the electronic-donating effect of the o-OMe-C6H4 substituent. Intriguingly, Chen's group designed ferrocene-containing phosphine-sulfonate ligands (e.g., complex 20 and 20ox, as shown in Figure 6) to tune polymerization processes as an example of redox-controlled polymerization by changing the valence state of iron, thus tuning the electron density around the active center.[28d]
Structures of the functionalized norbornenes NB-OAc and NB-NP.[28e]
Structures of catalysts 16a–e, 17a–c, 18a–c, 19a–d, 20, 20ox, 21, 22, and 23a–i.[28a], [40], [41], [43]
Interestingly, the phosphine oxide ligand in a BPMO catalyst can offer an additional site for ligand modifications to tune the behavior of the catalyst, relative to catalysts containing phosphine-sulfonate ligands.[40], [43] It is worth mentioning that some interesting phenomena could be observed in a study by Nozaki's group of a series of BPMO-Pd catalysts 21 and 22 (Figure 6).[43] When two less bulky iPr groups are replaced by two phenyl groups on the P atom in the BPMO ligand, the activity of the catalyst decreases dramatically. However, with ortho-substituted aryl groups on the P atom, the activity of the catalyst is improved markedly, although it is still lower than that of the catalyst with iPr on the P atom. Obviously, the effects of ligands in this system are too complex to be understood easily. In addition, both Chen's group[40] and Nozaki's group studied the steric perturbation of phosphine oxide substituents on the behavior of catalysts. Chen's group found that decreasing the size of the substituents on the P atom can enhance the activities (including copolymerization of ethylene with polar monomer), whereas Nozaki's group replaced one substituent on the P atom with an amine group, finding that putting less bulky substituents on N atom leads to very low catalytic activity. Obviously, there are many puzzles still remaining to be solved in the area of the steric effects of ligands in BPMO-Pd catalysts.
Recently, Ni catalysts 23a–i (Figure 6), featuring o-bis(aryl)-phosphinophenolate ligands showing different degrees of electronic and steric perturbation, have been reported for high-performance copolymerization of ethylene and alkyl acrylates without any activators.[41] It can be seen clearly that outstandingly high performances are achieved for this series of catalysts (Table 4). The researchers also found that an increase in the activity appears with stepwise replacement of phenoxy group(s) by MeO group(s) at the aryl groups around the metal center, which may be the result of the effects of MeO–Ni interaction.[41]
Recent reports show that FI catalysts (phenoxy-imine ligand catalysts) have been widely studied in the context of their high tolerance for functional additives among catalysts with N–O chelating ligands.
Intriguingly, the influence of fluoro substituents is a highlight for salicylaldimine catalysts.[3], [15d], [19a] From the point of view of electronic effects, Chen's group designed catalysts 24a–d (Figure 7), with electron-withdrawing or -donating substituents in the para-position to the N-aryl group, which are covalently closer to the metal center thus they may have remarkable inductive effects to eliminate or weaken the influence of (ligand)C–F···H–C(polymer) interactions.[30a] Ethylene polymerization was run with bis(cycloocta-1,5-diene)nickel [Ni(COD)2], and the activity was enhanced gradually with increasing electron-withdrawing effects (Table 5).
Structures of catalysts 24a–d, 25a–c, 26a–c, and 27a–c.[19b], [25], [30a], [44]
| Catalyst | Actb | Mwc (× 10–4) | Mw/Mnc | B d | Tme [°C] |
|---|---|---|---|---|---|
|
|||||
| 24a | 1.6 | 11.8 | 2.03 | 25 | 112 |
| 24b | 2.1 | 15.2 | 2.27 | 26 | 111 |
| 24c | 2.9 | 13.6 | 3.03 | 39 | 109 |
| 24d | 3.0 | 13.3 | 3.13 | 47 | 107 |
To study the steric effects of ligands, Damavandi et al. designed a series of phenoxyiminato catalysts and found that the joint presence of two bulky groups at the ortho positions of the N-aryl group in catalyst 25c (Figure 7) leads to higher activity for ethylene polymerization.[25]
The N-heterocyclic carbene (NHC) structure is not a common ligand for high-performance catalysts. However, Nazaki's group reported 26a–c, a series of neutral Ni complexes (Figure 7) containing NHC ligands.[19b] In the presence of Ni(COD)2, high activity of up to 106 g of PE (mol of Ni)–1 h–1 was achieved with catalyst featuring bulkier substituents near the metal center. In addition, this series of catalysts can catalyze ethylene polymerization at 50–100 °C with moderate activity in the absence of cocatalyst.
Zhang et al. reported 27a–c, a class of cyclopentadienyl phenyl phosphine complexes (Figure 7) with differing degrees of bulkiness around the metal center.[44] Although the cyclopentadienyl group is convenient in terms of coordination and modification, the activities of Ni catalysts of this class are not high [9.6–59.0 kg PE (mol of Ni)–1 h–1]. However, the influence of steric perturbation of ligand allowed the conclusion that the catalytic activities generally decrease as the bulkiness of substituents increases.
Obviously, different steric effects of ligands on catalytic activities can be found from the two examples above. Therefore, well-designed experiments followed by valid characterization methods are necessary to illustrate the mechanisms.
Universidad Autónoma de Madrid © 2008 · Ciudad Universitaria de Cantoblanco · 28049 Madrid · Información y Conserjería: 91 497 43 31 E-mail: informacion.ciencias@uam.es Gestión de estudiantes de Grado y Posgrado: 91 497 8264 / 4329 / 4353 / 4349 / 6879 / 8362 E-mail: administracion.ciencias@uam.es
