Polymerization Behavior of C1-Symmetric Metallocenes (M = Zr, Hf

Article pubs.acs.org/Organometallics

Polymerization Behavior of C1‑Symmetric Metallocenes (M = Zr, Hf): From Ultrahigh Molecular Weight Elastic Polypropylene to Useful Macromonomers Alexander Schöbel, Dominik Lanzinger, and Bernhard Rieger* Wacker Lehrstuhl für Makromolekulare Chemie, Technische Universität München, Lichtenbergstr. 4, D-85748 Garching, Germany ABSTRACT: The C1-symmetric metallocenes rac-[1-(9-η5-fluorenyl)-2(5,6-cyclopenta-2-methyl-1-η5-indenyl)ethane]zirconium dichloride (1) and rac-[1-(9-η5-fluorenyl)-2-(5,6-cyclopenta-2-methyl-1-η5-indenyl)ethane]hafnium dichloride (2) are known to produce elastic polypropylene. They were investigated concerning their temperature stability during the polymerization of propene. After activation of these complexes with triisobutylaluminum (TIBA)/[CPh3][B(C6F5)4], first polymerization experiments (80−100 °C) afforded moderate to high activities. However, at these polymerization temperatures, the molecular weights of the produced polymers are significantly decreased, resulting in a waxy appearance and, therefore, a loss of the elastic behavior. The main reason for this behavior, especially for the more significant decrease of the molecular weight in the case of hafnocene 2 compared with zirconocene 1, was revealed to occur due to a fast β-methyl transfer reaction. Hence, hafnocene 2 can form polymer chains with a high selectivity toward allylic chain ends. These macromonomers can be used in the catalytic insertion polymerization for the formation of new grafted copolymers. Initial copolymerization experiments with ethene were conducted.

■

mmmm-pentad depending on the propene concentration.25 However, molecular weights of the polymers prepared with these complexes were low. Further developments afforded metallocenes 1 and 2, which are able to form ePP with increased molecular weights and, therefore, good elastic properties (Scheme 1).26,27 Elastic polypropylenes produced with complexes 1 and 2 have a microstructure with isotactic sequences interrupted by single stereoerrors. The formation of these ePPs was reported to occur via the depicted mechanism, allowing a superior control over the amount of the isolated stereoerrors over a broad range just by variation of propene concentration and polymerization temperature. Higher propene concentrations or lower temperatures increase the amount of stereoerrors and vice versa. Increasing propene concentration favors propagation, which is usually a bimolecular reaction, over the unimolecular chain back-skip. A lower polymerization temperature results in a, relative to coordination and insertion on the less stereoselective site, more significantly decreased chain back-skip reaction rate.18 Especially, the use of hafnium instead of zirconium affords significantly higher molecular weights of the ePPs. These ultrahigh molecular weight ePPs show elastic recovery similar to that of commercially available TPEs.27 Hence, their production at industrially relevant polymerization temperatures would be desirable. For this purpose, complexes 1 and 2 were investigated concerning their high-temperature polymerization properties.

INTRODUCTION Elastic PP (ePP) was already discovered by Natta at the very first beginning of the Ziegler−Natta polymerization catalyst development in the 1950s as a heptane-soluble fraction.1 The microstructure of this polymer was assumed to be block-like, consisting of atactic and isotactic sequences. After the discovery of methylaluminoxane (MAO) as a cocatalyst and the mechanistic understanding of the complex symmetry−polymer microstructure relationship (Ewens symmetry rules), the first homogeneous metallocenes were synthesized for the formation of ePP.2−7 Under those metallocenes, rac-[ethylidene(1-η5tetramethyl-cyclopentadienyl)(l-η5-indenyl)]titanium dichloride developed by Chien and co-workers was reported to produce ePP with a block-like microstructure (atacticisotactic).8−10 Gauthier et al. used similar zirconocene and hafnium complexes and improved activities and molecular weights.11−13 Further improvements concerning the molecular weight were made by Collins and co-workers, which afforded PP with molecular weights that were still not high enough for a good elastic behavior.14 Besides this report, Collins and coworkers studied unsymmetric complexes with respect to the counteranion effect, the intrinsic stereoselectivity of the active sites, and the relative rates for monomer insertion compared to inversion, all influencing the microstructure of the formed PP.15−19 Unbridged metallocenes switching between their racand meso-forms within the time scale of chain propagation were adopted by Waymouth and co-workers.20−24 These complexes afford ePP with an isotactic-atactic blockstructure. rac-[1-(9-η5Fluorenyl)-2-(1-η5-indenyl)ethane]zirconium dichloride and similar metallocene dichlorides were reported by Rieger et al. to produce polypropylenes with a varying amount of the © XXXX American Chemical Society

Received: August 15, 2012

A

dx.doi.org/10.1021/om300781a | Organometallics XXXX, XXX, XXX−XXX

Organometallics

Article

Especially, the lower molecular weights of the PPs produced with hafnocene 2 are not expected, and thus further investigations with respect to the molecular weight dependency of metallocenes 1 and 2 as a result of their chain-transfer reactions will be separately discussed in more detail. Lower observed activities for hafnocenes are often explained with a more stable Hf−C σ bond.29 However, using the dimethylated catalyst precursors of the metallocenes 1 and 2 was shown to result in similar productivities after activation with [CPh3][B(C 6 F 5 ) 4 ]. 27 To reveal the reason for the reduced productivities in the case of hafnocene 2, activated with TIBA and [CPh3][B(C6F5)4], a more detailed investigation of the preactivation reaction with TIBA was performed. Activation. UV−vis spectroscopy can be used to follow the preactivation reaction of a metallocene dichloride with TIBA due to a shift of the ligand-to-metal charge-transfer band (LMCT band).30−32 The lowest-energy absorption band in the UV−vis spectrum of bent metallocenes arises from a charge transfer of the highest occupied (HOMO) to the lowest unoccupied molecular orbital (LUMO). A significant advantage of this analysis method is that used aluminum compounds do not have absorption bands in the region of the LMCT band. Hence, also high amounts of aluminum derivatives can be applied. In addition, compared to NMR spectroscopy, reduced concentrations that are in the range of those used for the activation and polymerization reaction are accessible. Therefore, observed reaction rates are in good accordance with those observed during the activation for the polymerization process. The obtained spectra show a hypsochromic shift of the LMCT band for both metallocenes (1, 2). This shift results from an increased HOMO−LUMO gap caused by an exchange of the σcoordinated chloro ligand by a stronger electron-donating group (Figures 1 and 2). The observed isosbestic point at 474 nm (Figure 1) is indicative of a conversion of the dichloro precursor in only one other derivative without any consecutive reaction step. In this specific case, monoalkylation of the dichloro precursor is most probable (Scheme 2). This monoalkylation reaction is also in accordance with a reaction mechanism proposed by Bochmann and co-workers for the ternary catalyst rac-Me2Si(Ind2)ZrCl2/TIBA/[CPh3][B(C6F5)4].33 However, for the hafnocene 2 at low amounts of TIBA and room temperature, the LMCT band is only slightly shifted toward lower wavelengths (Figure 2). For a preactivation reaction of hafnocene 2 with a shift of the LMCT band analogue to the one of zirconocene 1 at low amounts of TIBA, more TIBA (Al/Hf = 200/1) and higher reaction temperatures are necessary (Figure 3). NMR investigations of the reaction of Me2Si(Ind2)MCl2 (M = Zr, Hf) with TIBA also showed the necessity of increased amounts of TIBA for a quantitative monoalkylation of the hafnocene 2 compared to the zirconocene 1.34

Scheme 1. Investigated C1-Symmetric Complexes and Their Polymerization Mechanism Affording Isolated Stereoerrors Whose Amount Depends on Propene Concentration and Polymerization Temperature

■

RESULTS AND DISCUSSION Temperature Stability. To see the temperature stability of complexes 1 and 2, polymerization experiments were performed at elevated (60 °C) and higher temperatures up to 100 °C (Table 1). For the activation of the metallocene dichlorides 1 and 2, a standard procedure, including a preactivation step with TIBA (1 h at 60 °C) and the formation of the cationic species by adding [CPh3][B(C6F5)4], was applied. Even at the highest applied polymerization temperature (entry 3, Table 1), the propene uptake was constant for the complete polymerization time. Increased polymerization temperatures together with slightly raised propene pressures led to similar (entries 1 and 2) or even higher activities (entries 3 and 4). The activity of a metallocene catalyst increases with propene concentration and temperature. Hence, because of the lower solubility of propene at higher temperatures, similar or higher productivities are a result of increased propagation rates affected by higher polymerization temperatures. These observations are all indicative of a temperature-stable catalyst without any significant degradation reaction. In terms of temperature stability, both complexes behave similarly. In contrast, with respect to productivity and MW, hafnocene 2 shows lower values compared to zirconocene 1.

Table 1. Polymerization of Propene with 1 and 2 Activated with TIBA/[CPh3][B(C6F5)4] at Elevated and High Temperatures entry

complex

tpa

Tpb

pc

cd

productivitye

Mwf

PDI

1 2 3 4

1 1 2 2

12 13 11 16

80 60 100 60

7.6 7 7.6 7

1.6 1.9 1.3 1.9

69 000 70 000 34 000 22 000

17 000 60 000 4000 36 000

2.0 1.9 2.2 1.9

In minutes. bIn °C. cp = pAr + ppropene in bar, with pAr = 1.9 ± 0.2 bar. dPropene concentration in mol/L (propene concentrations were calculated using the equation reported by Busico et al.28). eIn kg (PP)/(mol(M) h). fIn g/mol.

a

B


Organometallics

Article

Figure 1. Preactivation of zirconocene 1 with TIBA (c(Zr) = 9 × 10−4 mol/L).

Figure 2. Preactivation of hafnocene 2 with TIBA (c(Hf) = 8 × 10−4 mol/L).

monoalkylation of the metallocene dichloride, an even faster consecutive reaction was observed. This indicates a crucial difference in the preactivation reaction of hafnocenes and zirconocenes with TIBA. Hence, the lower productivities of the hafnocene 2 is assumed to be a result of different precursors formed during the preactivation reaction with TIBA. These are converted into their active forms, leading to the observed different activities after the activation with [CPh3][B(C6F5)4]. Recently, lower activities for hafnocenes activated with TMA containing MAO were reported to be a result of the same resting state that is more stable in the case of hafnocenes compared to the corresponding Zr species.35 However, with respect to our investigations, an influence of the alkylation reaction, which surely depends on the whole catalytic system (metallocene dichloride, cocatalyst, Al-derivative, and their amounts and concentrations) must be considered. To obtain further information on the different precursors formed during the preactivation reaction of metallocenes 1 and 2 with TIBA, NMR spectra were recorded using different ratios

Scheme 2. Monoalkylation of Metallocene Dichlorides 1 and 2 by TIBA and the Qualitative Presentation of the Change of the HOMO−LUMO Gap

Because polymerization experiments with hafnocene 2 and zirconocene 1 were conducted using an identical preactivation procedure (200 equiv TIBA, 1 h at 60 °C), also the UV−vis spectrum of zirconocene 1 was recorded under these conditions (Figure 4). Instead of the expected fast and quantitative C


Organometallics

Article

Figure 3. Preactivation of hafnocene 2 with TIBA (c(Hf) = 8 × 10−4 mol/L).

Figure 4. Preactivation of zirconocene 1 with TIBA (c(Zr) = 9 × 10−4 mol/L).

complexes with bridged and terminal hydrides are presented. The chemical shift of the NMR signals of most of these bridged hydrides is reported in the range of −1 to −2 ppm, whereas for the terminal hydrides, a chemical shift above 2 ppm was obtained.36 In addition, Bochmann and co-workers found a cationic dihydrido-bridged heterodinuclear complex of the monohydride species of hafnocene 2 together with DIBAL-H after the preactivation reaction with 100 equiv of TIBA and the addition of [CPh3][B(C6F5)4].34 Hence, we attribute the observed signal at −1.81 ppm to a bridged hydride species of the complex. The very narrow signal can be attributed to the fact of a missing fast exchange reaction of the bridged hydride species between the two sites of the complex, which is most presumably a result of the different steric hindrance on these sites. This behavior was previously published by Baldwin et al. for complexes possessing differently crowded sites on the metal atom.36 In the case of zirconocene 1, the absence of the educt signal indicates completion of the preactivation reaction already at low Zr/Al ratios (Zr/Al = 1:10). However, besides this

of M/Al (1:10, 1:20). The spectrum of hafnocene 2 with an M/Al ratio of 1:10 shows, besides the signals of TIBA together with its impurities (isobutene, DIBAL-H), relevant signals below 0 ppm and in the range between 5.5 and 6.3 ppm (Figure 5). Three signals of this magnified region, always with a ratio of 1:1:1, can be assigned to the monoalkylated product (continuous and dotted line). The monoalkylation reaction of the dichloro precursor generates a metallocene with four different substituents. Hence, the methylene group attached to the metal center is diastereotopic, affording two different signals (dotted line) with a 2J coupling constant of 14 Hz and two different 3J coupling constants of 4 and 8 Hz. These two signals always appear in combination with a singlet at 6.14 ppm that can be assigned to the hydrogen atom of the indenyl moiety. Furthermore, after a reaction time of 1 h, a residual educt of roughly 10% (5.85 ppm) can be detected. Besides these signals, an additional singlet was observed at −1.81 ppm. The intensity of this singlet rises with higher amounts of TIBA. Baldwin et al. investigated several bridged metallocenes according to their reaction with DIBAL-H. As a result, D


Organometallics

Article

Figure 5. Magnified region of the 1H NMR spectrum of the preactivation reaction of TIBA with hafnocene 2 (M/Al = 1:10; T = 60 °C; c(Hf) = 2.6 × 10−2 mol/L).

Figure 6. Influence of the polymerization temperature on the molecular weights of the PPs produced with complexes 1 and 2 at similar propene concentrations (1.2 ± 0.1 mol·L−1) (propene concentrations were calculated using the equation reported by Busico et al.28).

observation, all obtained spectra of zirconocene 1 and hafnocene 2 activated with TIBA are, especially in the region below 0 ppm, similar with minor variations of the integral ratios and slightly different shifted signals. NMR spectroscopy reveals similar reaction products for the two metallocenes 1 and 2, whereas the UV−vis investigations clearly indicate a crucial difference for the formed precursors. This is most presumably due to the higher concentrations applied during NMR measurements. In addition, for higher amounts of TIBA (e.g., 200 equiv), useful data are not accessible by standard NMR measurements. Consequently, 1H NMR spectroscopy cannot reveal the different precursor structures of metallocenes 1 and 2 formed during the reaction with 200 equiv of TIBA. Molecular Weight Dependency. As already mentioned above, at elevated polymerization temperatures, hafnocene 2 shows unexpected lower molecular weights compared to zirconocene 1 for the produced polypropylenes. Usually, hafnocenes are reported to afford polymers with higher

molecular weights compared to the corresponding zirconocene.7,11,29,37 This is also observed for metallocenes 1 and 2.26 At high propene concentrations and low polymerization temperatures, molecular weights of up to 5 000 000 g/mol are reported for the dimethylated hafnocene 2 activated with [CPh3][B(C6F5)4], whereas under similar polymerization conditions with the dimethylated zirconocene 1, PP with a molecular weight below 1 000 000 g/mol is formed.27 Increasing polymerization temperatures result, as expected, in polymers with decreased molecular weights for both metallocenes due to higher rates of chain release reactions at these temperatures (Figure 6). To exclude an influence of propene concentration at different polymerization temperatures, experiments were performed at similar propene concentrations. When two metallocenes 1 and 2 are compared, a significant difference for the decline of the molecular weights with increasing Tp is obvious. At lower Tp, hafnocene 2 produces PP with higher molecular weights compared to zirconocene 1. At a specific temperature (in this case, between 40 and 50 °C), the E


Organometallics

Article

Figure 7. End group analysis via 1H NMR spectroscopy (at 100 °C in C6D6Br) of two polymers produced with zirconocene 1 (bottom) or hafnocene 2 (top) (Tp = 80 °C; p = ppropene + pAr = 7.6 bar, with pAr = 1.9 ± 0.2 bar).

Figure 8. Dependency of the degree of polymerization (Pn) from propene concentration (cpropene) in the propene polymerization with complexes 1 and 2 at Tp = 60 °C, compared to SiMe2(BenzInd)2ZrCl2, a metallocene known in the literature, at Tp = 50 °C (Pn was determined via GPC; propene concentrations were calculated for all depicted metallocenes using the equation reported by Busico et al.28).

molecular weight of PP formed with zirconocene 1 becomes higher compared to the PP produced with hafnocene 2 at identical conditions. This much more pronounced decrease of the molecular weight for the hafnocene 2 must be a reason for different transfer reactions taking place for the two metals and will be, therefore, discussed in more detail in the next section. Transfer Reactions. For the polymerization of propene, different transfer reactions are known. These are β-hydride transfer, which can be uni- or bimolecular, transfer to the cocatalyst (e.g., TMA present in MAO), and β-methyl transfer.38,39 β-Hydride transfer affords vinylidene chain ends, transfer to the cocatalyst leads to saturated chain ends after aqueous workup, and β-methyl transfer gives allylic chain ends. To investigate the influence of the cocatalyst, different aluminum compounds were synthesized and compared concerning their influence on the propene polymerization

behavior of hafnocene 2. On the one hand, steric encumbrance of the aluminum derivatives was increased, and on the other hand, aluminum agents with lower Lewis acidity were used. In both cases, the probability for a transfer of the polymer chain to the cocatalyst is reduced. For this purpose, AlNp3 (Np = neopentyl) and MeAl(BHT)2 (BHT = 2,6-di-tert-butyl-4methylphenolate) were synthesized. However, in the polymerization reaction of propene with hafnocene 2 and these cocatalysts, no significant change of the molecular weights was observed. Hence, the transfer of the polymer chain to the cocatalyst can be neglected as the main reason for the significant decrease of the molecular weight. Because of the low molecular weights of the polymers produced at elevated temperatures (60 °C and above), further information about chain-release reactions are accessible via 1H NMR end group analysis. The olefinic region of the 1H NMR F


Organometallics

Article

Table 2. Comparison of Literature Known Hafnocene and Zirconocene Systems, Showing, Dependent on Their Transfer Reactions, a Different Behavior at Elevated Tpd39,44 entry

complex

cocatalyst

Tpa

Pn

1 2 3 4 5 6 7 8 9

Cp2ZrCl2 Cp2HfCl2 Cp*2ZrCl2 Cp*2HfCl2 Et(Flu)2ZrCl2 Et(Flu)2HfCl2 Et(Flu)2ZrCl2 1 2

MAO MAO MAO MAO MAO MAO MAO TIBA/Ie TIBA/Ie

50 50 50 50 50 50 50 60 60

17 140 4.5 3.4 1700 1600 315 430f 140f

β-Meb

β-Hb

Altb,c

91.1 98

100 100 7.9 2

1 traces

75 64g 95g

25 34g 4g

a In °C. bIn %. cAlt: transfer to aluminum. dPolymerization conditions: in liquid propene for entries 1−6; p = ppropene − a = 4 bar for entry 7; p = ppropene + pAr = 4.0 bar, with pAr = 1.9 ± 0.2 bar for entries 8 and 9. eI: [CPh3][B(C6F5)4]. fPn determined by GPC. gDifference to 100% is due to the formation of internal double bonds.

transfer reactions (entries 3−6, Table 2). For the metallocenes that are less sterically hindered (entries 1 and 2), the bimolecular β-hydride transfer reaction is dominant and polymers with a higher Pn are formed by the hafnocene. Besides the generally similar behavior of metallocenes 1 and 2 and already published metallocenes, a more significant difference of Pn and a higher selectivity toward the allylic chain ends for the hafnocene 2 compared to the zirconocene 1 is obvious (entries 3−6, 8, and 9, Table 2). One explanation for the crucial difference of metallocenes 1 and 2 compared to other reported complexes is a higher difference of the hafnocene's activation barriers for the β-methyl and the βhydride transfer reaction. Compared to the zirconocene, this results at higher polymerization temperatures for the hafnocene complex in a relative to propagation more significantly increased β-methyl transfer reaction rate. However, for these C1-symmetric metallocenes, also a combination of their characteristic back-skip polymerization mechanism and a site-dependent transfer reaction could account for this behavior and will be, therefore, discussed in the following. In the case of the C2v-symmetric complexes (entries 1−7, Table 2), both active sites of these species have a similar steric encumbrance, whereas the C1-symmetric complexes are differently crowded on their active sites. Because increasing steric hindrance by substituents on the Cp moiety is known to favor β-methyl over β-hydride transfer reactions, a different preference for the transfer reaction depending on the site of the complex is reasonable (Figure 9).45 A polymer chain situated on the less sterically hindered site (blue-marked) is assumed to undergo a β-hydride transfer reaction due to its similarity to rac-SiMe2(2-Me-Ind)2ZrCl2 and rac-C2H4(Ind)2ZrCl2, which are both known to predominantly undergo a βhydride transfer reaction.39 In contrast, a polymer chain coordinated to the sterically more hindered site (red-marked)

spectra of two polymers, one produced with zirconocene 1, the other with hafnocene 2 under the same conditions, is depicted in Figure 7. Zirconocene 1 produces PP with two main termination products that are the vinylidene- and the allylterminated polymer chains with a ratio of 1:0.8 under the applied conditions. The vinylidene end group is generated via a β-hydride and the allyl end group via a β-methyl transfer reaction. In addition, two small singlets, one at 4.95 ppm, the other at 4.86 ppm, are present. These two signals can be assigned to polymer chain derivatives with internal double bonds and are reported to be formed via isomerization reactions.39 In contrast, hafnocene 2 shows a very selective transfer reaction toward the allylic chain ends. Only minor amounts of the vinylidene-terminated polymer chains can be detected. Because of the fact that, in the literature, for the β-methyl transfer reaction, besides unimolecular, bimolecular pathways are also reported, we were interested in whether the β-methyl transfer reaction was uni- or bimolecular for the investigated metallocenes 1 and 2.40−42 Therefore, polymerization reactions with increasing propene concentrations were conducted (Figure 8). For comparison, SiMe2(BenzInd)2ZrCl2, a complex known to predominantly undergo a bimolecular β-hydride transfer reaction, is also plotted.43 In the case of SiMe2(BenzInd)2ZrCl2, the degree of polymerization (Pn) is only slightly increased with propene concentration, indicating the influence of the bimolecular transfer reaction. Both complexes 1, and 2 show an increased Pn with raised propene concentration and, therefore, a much more pronounced unimolecular transfer reaction. A bimolecular pathway for the β-methyl transfer reaction plays a minor role or is irrelevant for these complexes at least under applied conditions. NMR and molecularity investigations clearly reveal the unimolecular and fast β-methyl transfer reaction in the case of hafnocene 2 to be the reason for the significant decreased molecular weights of PP produced at elevated and high polymerization temperatures. The observed trend of a fast β-methyl transfer reaction at high polymerization temperatures affording polymers with similar or lower Mw in the case of hafnocenes compared to the corresponding zirconocenes can also be found in the literature for other metallocenes (Table 2). Unfortunately, reported data concerning the end group analysis are few, especially in the case of hafnocenes. However, for all shown examples at elevated Tp, similar or slightly lower molecular weights are obtained in the case of hafnocenes, which are known to undergo β-methyl

Figure 9. Site dependency of the chain-transfer reactions for metallocenes 1 and 2. G


Organometallics

Article

Table 3. Formation of PP-MMs with Hafnocene 2 Activated with TIBA/[CPh3][B(C6F5)4] in Toluene under Varying Conditions

Table 5. Polymers Obtained by Copolymerization Reactions of Ethene with Different PP-MMs by Hafnocene 2 after Activation with TIBA/[CPh3][B(C6F5)4] in Toluene

macromonomers

tpa

Tpb

pc

Mnd

Sallyle

productivityf

polymer

Mwa

Tmb

MM contentc

productivityd

MM1 MM2 MM3

13 53 51

70 60 50

4.6 4.0 4.0

3900 7200 32 000

93 91 >98

19 000 5000 5000

PE-1 PE-g-PP-1 PE-g-PP-2

900 000 900 000 1 100 000

133 129 128

16 23

1000 400 3200

In minutes. bIn °C. cp = pAr + pethene in bar, with pAr = 1.4 ± 0.1 bar. In g/mol determined via 1H NMR spectroscopy, assuming an unsaturated chain end for each polymer chain. eSelectivity toward allylic chain ends in percent. fIn kg (PP)/(mol(Hf) h).

In g/mol. bIn °C. cIn wt %. dIn kg (PP)/(mol(Hf) h), calculated on the polymer yield after Soxhlet extraction.

a

a

d

was determined to be 7000 g/mol and, for syndiotactic PP, below 3500 g/mol.50,51 To see if the highly selective β-methyl transfer reaction in the case of hafnocene 2 can be used as a versatile tool for the formation of MMs in a broad range of Mn, polymerization reactions of propene at varying conditions were conducted (Table 3). All of the MMs produced show a high content of allylic chain ends of at least 91%. Furthermore, the molecular weight of the MMs can be varied over a broad range up to an Mn of 32 000 g/mol, which is not the limit. In the literature, for the metallocene-based formation of polypropylene MMs, fulfilling the requirements of an MM for formation of LCB copolymers, there are only few examples. Isotactic PP-MMs were obtained with rac-dimethylsilanediylbis(2-methyl-4-phenylindenyl)zirconium dichloride, affording an allylic chain end selectivity between 75 and 81% with an Mn ranging from 15 000 to 55 000 g/mol.52,53 In the case of atactic PP-MMs, bis(9-fluorenyl)zirconium dichloride is reported to generate MMs with an Mn in the range of 3000 up to 8000 g/mol together with an allylic chain end selectivity of 68−82%, respectively.39,44 Syndiotactic PP-MMs are accessible by bis(phenoxyimine)titanium dichloride complexes.54 Recently, a copolymer of ethene and propene produced by a Cp*2ZrCl2/MAO catalyst was reported with high selectivity toward allylic chain ends.55,56 Molecular weights of these MMs can be increased with ethene concentration (8000−25 500 g/mol). Comparing these reports to results obtained with metallocene 2 reveals the evident advantage of hafnocene 2 in terms of producing MMs with high allylic chain end selectivity at a broad range of Mn. In addition, the formed MMs afford high solubility in various organic solvents. Therefore, non incorporated MMs can easily be removed from the copolymer by Soxhlet extraction, for example, from polymers with a PE or iPP backbone. The copolymerization reaction of ethene with MM2 or MM3 using hafnocene 2 afforded PP-g-PE copolymers PE-g-PP-1, -2 (Tables 4 and 5). The incorporation level of the MM was determined by 1H NMR spectroscopy after nonincorporated MM was removed by Soxhlet extraction with pentane. 1H NMR spectra showed no more olefinic signals, indicating complete removal of unreacted

is assumed to predominantly undergo a unimolecular β-methyl transfer reaction, an analogue to the depicted C2v-symmetric complex.39,44 An explanation for the higher selectivity toward the allylic chain ends in the case of hafnocene 2 can be given via the chain back-skip polymerization mechanism of these C1-symmetric metallocenes (Scheme 1). According to the assumed site dependency and the polymerization mechanism, the amount of β-methyl transfer reactions correlates with the relative to propagation existing rate of the chain back-skip from the sterically hindered to the opposite site of the complex. Hence, the higher amount of the β-methyl transfer reaction product for hafnocene 2 can be explained by a relative to propagation slower chain back-skip reaction at the sterically more hindered site of the complex. This assumption is further supported by kinetic investigations of the site epimerization of different hafnocenes and zirconocenes showing a slower rate in the case of hafnocenes.46 Additionally, the relative to propagation, slower rate of the chain back-skip for hafnocene 2 was also verified by polymerization experiments of the investigated C1symmetric complexes 1 and 2. Polypropylenes produced with hafnocene 2 have a lower mmmm-pentad concentration compared to the polypropylenes formed with zirconocene 1. This is in accordance with the above depicted polymerization mechanism (Scheme 1) in the case of a, relative to propagation, slower rate for the chain back-skip with hafnocene 2. Macromonomers and Their Applicability. For the incorporation of polypropylene oligomers into a polymer chain (e.g., PE) via coordinative insertion polymerization, the macromonomers (MMs) applied must have an accessible double bond. Hence, allylic are highly preferred over vinylidenic polymer chain end groups. Depending on the amount and length of incorporated MMs, different copolymers, such as comblike or long-chain-branched (LCB) polymers, can be obtained. The latter can be used as an additive for metallocene produced PP in order to improve the melt processability (e.g., higher melt strength and shear thinning).47−49 MMs for the formation of LCB polymers are typically considered to be longer than 2.5 times of the entanglement molecular weight (Mne). For iPP and aPP, Mne

Table 4. Conditions of the Copolymerization Reactions of Ethene with Different PP-MMs by Hafnocene 2 after Activation with TIBA/[CPh3][B(C6F5)4] in Toluene polymer

MM

MM amounta

Mn (MM)b

c (MM)c

pd

Tpe

tpf

PE-1 PE-g-PP-1 PE-g-PP-2

MM2 MM3

5 15

7200 32 000

2.3 × 10−3 1.6 × 10−3

2 2 2

30 30 30

6 48 22

a

In grams. bIn g/mol determined via 1H NMR spectroscopy, assuming an unsaturated chain end for each polymer chain. cIn mol/L. dp = pAr + pethene in bar, with pAr = 1.4 ± 0.1 bar. eIn °C. fIn minutes. H


Organometallics

Article

mentioned, all chemicals are commercially available, purchased from Aldrich, Acros, ABCR, or VWR, and used without further purification. All deuterated solvents for NMR measurements were obtained from Deutero. Dry solvents, such as toluene, diethyl ether, tetrahydrofuran, n-pentane, and dichloromethane, were obtained from an MBraun MBSPS-800 solvent purification system. Dioxane and n-hexane were obtained by filtration over Al2O3 and stored over a 3 Ǻ molecular sieve. rac-[1-(9-η5-Fluorenyl)-2-(5,6-cyclopenta-2-methyl-1-η5-indenyl)ethane]zirconium dichloride (1),26,57,58rac-[1-(9-η5-fluorenyl)-2-(5,6cyclopenta-2-methyl-1-η 5 -indenyl)ethane]hafnium dichloride (2),26,57,58rac-[1-(9-η5-fluorenyl)-2-(5,6-cyclopenta-2-methyl-1-η5indenyl)ethane]dimethylhafnium,27 (BHT)2AlMe,59trineopentylaluminum, 60and [CPh3][B(C6F5)4]61were prepared according to published procedures. All 1H NMR (300 MHz) and 13C NMR (75 MHz) measurements were performed on a Bruker ARX-300. Chemical shifts were referenced to signals from CDCl3 (7.26 and 77.16 ppm for 1H and 13C, respectively), CD2Cl2 (5.32 and 53.84 ppm for 1H and 13C, respectively), and toluene (2.08, 6.97, 7.01, 7.09 ppm; and 137.48, 128.87, 127.96, 125.13, 20.43 ppm for 1H and 13 C, respectively). The polypropylenes and the copolymers produced with the C1-symmetric complexes were measured at 100 °C in bromobenzene, and the chemical shift was referenced to the residual solvent signal (furthest upfield peak being 7.00 ppm for 1H). For UV−vis investigations, a device from Varian (Cary 50) equipped with a Peltier thermostatted cell holder, allowing temperature control and stirring of the sample with a magnetic stirring bar, was used. To perform the kinetic investigations under argon, a 3.5 mL modified quartz glass cell equipped with a Schlenk connection was used. The spectra were recorded in the wavelength range between 600 and 300 nm. After equilibration of the complex solution at the desired temperature, the preactivation agent (TIBA) was added via the septum of the cell, while, at the same time, the measurement was started. The addition of the TIBA solution was performed using a gastight Hamilton syringe with a Teflon tipped plunger. TIBA was injected as a 1.1 M toluene solution. Concentrations of the reaction solutions after addition of TIBA solution were in the range of 8 × 10−4 to 9 × 10−4 mol/L. NMR investigations of the preactivation reaction of complexes 1 and 2 with TIBA were performed in screw cap NMR tubes equipped with a Teflon sealing. Used solutions of TIBA in deuterated toluene had a concentration of 2.0 mol/L. After addition of the aluminum compound complex, concentrations were in the range of 1.9 × 10−2 to 2.9 × 10−2 mol/L. The NMR tube was placed into an oil bath already heated to the desired temperature and allowed to react for 1 h. Subsequently, a 1H NMR spectrum was recorded. Polymerization reactions were performed in a 1 L Büchi steel autoclave at constant pressure and temperature. To remove water and residual polymer of former polymerization reactions, the autoclave was charged with 300 mL of dry toluene and 3 mL of a 1.1 M TIBA solution in toluene and stirred for 1 h at 90 °C. For polymerization reactions, the autoclave was charged with 280 mL of toluene and 0.5 mL of a 1.1 M TIBA solution in toluene. For preactivation, 5 μmol of the metallocene was dissolved in 10 mL of toluene, 200 equiv of TIBA was added, and, subsequently, the mixture was stirred for 1 h at 60 °C. The preactivated complex solution was added via syringe into the autoclave. After equilibration of the reaction mixture at the desired polymerization temperature, the argon system pressure was adjusted to 1.9 ± 0.2 bar (formation of ePP) or 1.4 ± 0.1 bar (formation of MMs). Afterward, the autoclave was pressurized with propene up to the desired polymerization pressure. After equilibration of pressure and temperature, the polymerization was started by adding a trityltetrakis(pentafluorophenyl)borate solution (5 equiv, in 10 mL of dry toluene) via a pressure buret into the autoclave. The propene consumption was measured by a gas flow meter (Bronkhorst F-111C-HA-33P). The pressure (Bronkhorst pressure controller P-602C-FAC-33P) and temperature (Thermo Scientific HAAKE DynaMax) were kept constant during the entire polymerization procedure. Pressure, temperature, and propene consumption were also monitored and recorded online during the whole polymerization process. Polymerization reactions were quenched by adding methanol via the pressure

MM. Additionally, the highly sensitive RI detector during the GPC measurement showed a monomodal GPC trace without any shoulder after Soxhlet extraction, whereas a bimodal GPC trace was observed before extraction. MM contents up to 23 wt % show a good incorporation of the MMs during the copolymerization with ethene using hafnocene 2. The melting transitions of the copolymers are in a similar range (PP-g-PE-1, -2), being as expected lower compared to the corresponding homopolymerized PE (PE-1).

■

CONCLUSIONS In this report, detailed investigations of C1-symmetric metallocenes 1 and 2 containing different metal centers (Zr and Hf, respectively) were conducted. Significant differences concerning activation reactions and polymerization properties were shown to occur. UV−vis studies show a monoalkylation for zirconocene 1 already at low amounts of TIBA and a fast consecutive reaction at high amounts. In contrast, for hafnocene 2, only at high amounts of TIBA, a quantitative monoalkylation reaction can be observed. Additionally, performed NMR investigations show similar reaction products for 1 and 2 at low amounts of TIBA. This is in disagreement with obtained results by UV−vis spectroscopy and can be rationalized by higher concentrations in NMR samples. As a matter of fact, UV−vis investigations conducted at similar concentrations compared to preactivation and polymerization reactions can thus be used as a reliable tool for the interpretation of obtained results. Therefore, the observed lower activities for the hafnocene 2 are assumed to be due to different catalyst precursors formed during the preactivation with TIBA. Another main difference between 1 and 2 is a more significant decrease of the Mw of the PP produced with increasing Tp in the case of 2 compared to 1. The reason for this was shown to be a fast and unimolecular β-methyl transfer reaction for the hafnocene 2. Because hafnocenes, known to undergo predominantly a β-methyl transfer reaction, are also reported with similar to slightly lower molecular weights at higher polymerization temperatures compared to the corresponding zirconocenes, this behavior seems to be an intrinsic effect of the metal. Furthermore, the increased selectivity toward the β-methyl transfer reaction in the case of hafnocene 2 compared to zirconocene 1 is assumed to be the result of its higher activation barrier difference of the β-methyl and the βhydride transfer reaction. To rationalize the higher selectivity difference compared to catalysts reported in the literature, this effect can be assumed to be even more pronounced for complexes 1 and 2. In addition, we proposed a site-dependent transfer mechanism influenced by the characteristic ligand structure, which would also be reasonable to explain the increased selectivity of hafnocene 2 for the β-methyl transfer reaction compared to zirconocene 1. Furthermore, the high selectivity for the β-methyl transfer was verified to be a versatile tool for the formation of MMs, which are accessible for the coordinative insertion polymerization with metallocenes. Copolymerization reactions of these MMs with ethene and hafnocene 2 afforded PE-g-PP copolymers, being new with respect to the microstructure of the side chains, which is intermediate between isotactic and atactic.

■

EXPERIMENTAL SECTION

All manipulations with air- and moisture-sensitive compounds were carried out either under argon using standard Schlenk techniques or in an argon (4.8) filled glovebox from MBraun. If not otherwise I


Organometallics

Article

buret. Subsequently, the content of the autoclave was poured into acidified methanol. The polymer was filtered off, exhaustively washed with methanol, and dried in vacuum at 80 °C for several hours. The workup for the MMs afforded a slightly adapted procedure. Especially the MM with lower Mn only poorly precipitated after pouring the content of the autoclave into the acidified methanol. After some hours, a second sticky, colorless phase at the bottom of the beaker was formed. The solvent was decanted, and the MM was exhaustively washed with methanol. To reduce the water content, the MM was dissolved several times in dry toluene, which was removed in vacuum at 80 °C. For the copolymerization reactions with ethene the MM was added into the autoclave together with 2 mL of TIBA and stirred for some time before the preactivated catalyst was added. Afterwards the polymerization reaction was performed in the above-mentioned manner. Residual MM was removed from the copolymerized polymers (PE-g-PPs) by Soxhlet extraction with n-pentane within 24 h.

■

(21) Bruce, M. D.; Coates, G. W.; Hauptman, E.; Waymouth, R. M.; Ziller, J. W. J. Am. Chem. Soc. 1997, 119, 11174−11182. (22) Carlson, E. D.; Krejchi, M. T.; Shah, C. D.; Terakawa, T.; Waymouth, R. M.; Fuller, G. G. Macromolecules 1998, 31, 5343−5351. (23) Kravchenko, R.; Masood, A.; Waymouth, R. M.; Myers, C. L. J. Am. Chem. Soc. 1998, 120, 2039−2046. (24) Hauptman, E.; Waymouth, R. M.; Ziller, J. W. J. Am. Chem. Soc. 1995, 117, 11586−11587. (25) Rieger, B.; Jany, G.; Fawzi, R.; Steimann, M. Organometallics 1994, 13, 647−653. (26) Dietrich, U.; Hackmann, M.; Rieger, B.; Klinga, M.; Leskelä, M. J. Am. Chem. Soc. 1999, 121, 4348−4355. (27) Rieger, B.; Troll, C.; Preuschen, J. Macromolecules 2002, 35, 5742−5743. (28) Busico, V.; Brita, D.; Caporaso, L.; Cipullo, R.; Vacatello, M. Macromolecules 1997, 30, 3971−3977. (29) Ewen, J. A.; Haspeslach, L.; Atwood, J. L.; Zhang, H. J. Am. Chem. Soc. 1987, 109, 6544−6545. (30) Wieser, U.; Brintzinger, H. H. UV/VIS Studies on the Activation of Zirconocene-Based Olefin-Polymerization Catalysts. In Organometallic Catalysts and Olefin Polymerization; Blom, R., Follestad, A., Rytter, A., Tilset, M., Ystenes, M., Eds.; Springer: Berlin, 2001; p 3. (31) Coevoet, D.; Cramail, H.; Deffieux, A. Macromol. Chem. Phys. 1998, 199, 1451−1457. (32) Giannetti, E.; Nicoletti, G. M.; Mazzocchi, R. J. Polym. Sci., Part A: Polym. Chem. 1985, 23, 2117−2134. (33) Bryliakov, K. P.; Talsi, E. P.; Semikolenova, N. V.; Zakharov, V. A.; Brand, J.; Alonso-Moreno, C.; Bochmann, M. J. Organomet. Chem. 2007, 692, 859−868. (34) Bryliakov, K. P.; Talsi, E. P.; Voskoboynikov, A. Z.; Lancaster, S. J.; Bochmann, M. Organometallics 2008, 27, 6333−6342. (35) Busico, V.; Cipullo, R.; Pellecchia, R.; Talarico, G.; Razavi, A. Macromolecules 2009, 42, 1789−1791. (36) Baldwin, S. M.; Bercaw, J. E.; Brintzinger, H. H. J. Am. Chem. Soc. 2008, 130, 17423−17433. (37) Rieger, B.; Troll, C.; Preuschen, J. Macromolecules 2002, 35, 5742−5743. (38) Resconi, L.; Cavallo, L.; Fait, A.; Piemontesi, F. Chem. Rev. 2000, 100, 1253−1345. (39) Resconi, L.; Camurati, I.; Sudmeijer, O. Top. Catal. 1999, 7, 145−163. (40) Suzuki, Y.; Yasumoto, T.; Mashima, K.; Okuda, J. J. Am. Chem. Soc. 2006, 128, 13017−13025. (41) Yang, P.; Baird, M. C. Organometallics 2005, 24, 6005−6012. (42) Yang, P.; Baird, M. C. Organometallics 2005, 24, 6013−6018. (43) Stehling, U.; Diebold, J.; Kirsten, R.; Roll, W.; Brintzinger, H. H.; Jungling, S.; Mülhaupt, R.; Langhauser, F. Organometallics 1994, 13, 964−970. (44) Resconi, L.; Jones, R. L.; Rheingold, A. L.; Yap, G. P. A. Organometallics 1996, 15, 998−1005. (45) Resconi, L.; Piemontesi, F.; Franciscono, G.; Abis, L.; Fiorani, T. J. Am. Chem. Soc. 1992, 114, 1025−1032. (46) Song, F. Q.; Lancaster, S. J.; Cannon, R. D.; Schormann, M.; Humphrey, S. M.; Zuccaccia, C.; Macchioni, A.; Bochmann, M. Organometallics 2005, 24, 1315−1328. (47) Khadir, A.; Gauthier, M. Highly branched (arborescent) polymers as melt processing additives. In ANTEC’97: Plastics Saving Planet Earth, Conference Proceedings; Society of Plastics Engineers: Brookfield, CT, 1997; Vol. 1−3, pp 3732−3736. (48) Beigzadeh, D.; Soares, J. B. P.; Duever, T. A. Macromol. Symp. 2001, 173, 179−194. (49) Chum, P. S.; Kruper, W. J.; Guest, M. J. Adv. Mater. 2000, 12, 1759−1767. (50) Eckstein, A.; Suhm, J.; Friedrich, C.; Maier, R. D.; Sassmannshausen, J.; Bochmann, M.; Mülhaupt, R. Macromolecules 1998, 31, 1335−1340. (51) Liu, C. Y.; Yu, J.; He, J. S.; Liu, W.; Sun, C. Y.; Jing, Z. H. Macromolecules 2004, 37, 9279−9282.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS This work was supported by Borealis Polyolefine GmbH. REFERENCES

(1) Natta, G. J. Polym. Sci. 1959, 34, 531−549. (2) Sinn, H.; Kaminsky, W.; Vollmer, H. J.; Woldt, R. Angew. Chem., Int. Ed. Engl. 1980, 19, 390−392. (3) Sinn, H.; Kaminsky, W. Adv. Organomet. Chem. 1980, 18, 99− 149. (4) Ewen, J. A. J. Am. Chem. Soc. 1984, 106, 6355−6364. (5) Kaminsky, W.; Kulper, K.; Brintzinger, H. H.; Wild, F. Angew. Chem., Int. Ed. Engl. 1985, 24, 507−508. (6) Ewen, J. A.; Elder, M. J.; Jones, R. L.; Haspeslagh, L.; Atwood, J. L.; Bott, S. G.; Robinson, K. Makromol. Chem., Macromol. Symp. 1991, 48−49, 253−295. (7) Ewen, J. A.; Jones, R. L.; Razavi, A.; Ferrara, J. D. J. Am. Chem. Soc. 1988, 110, 6255−6256. (8) Mallin, D. T.; Rausch, M. D.; Lin, Y. G.; Dong, S. Z.; Chien, J. C. W. J. Am. Chem. Soc. 1990, 112, 2030−2031. (9) Llinas, G. H.; Dong, S. H.; Mallin, D. T.; Rausch, M. D.; Lin, Y. G.; Winter, H. H.; Chien, J. C. W. Macromolecules 1992, 25, 1242− 1253. (10) Chien, J. C. W.; Llinas, G. H.; Rausch, M. D.; Lin, Y. G.; Winter, H. H.; Atwood, J. L.; Bott, S. G. J. Polym. Sci., Part A: Polym. Chem. 1992, 30, 2601−2617. (11) Gauthier, W. J.; Corrigan, J. F.; Taylor, N. J.; Collins, S. Macromolecules 1995, 28, 3771−3778. (12) Gauthier, W. J.; Collins, S. Macromolecules 1995, 28, 3779− 3786. (13) Gauthier, W. J.; Collins, S. Macromol. Symp. 1995, 98, 223−231. (14) Bravakis, A. M.; Bailey, L. E.; Pigeon, M.; Collins, S. Macromolecules 1998, 31, 1000−1009. (15) Nele, M.; Mohammed, M.; Xin, S.; Collins, S.; Dias, M. L.; Pinto, J. C. Macromolecules 2001, 34, 3830−3841. (16) Mohammed, M.; Nele, M.; Al-Humydi, A.; Xin, S.; Stapleton, R. A.; Collins, S. J. Am. Chem. Soc. 2003, 125, 7930−7941. (17) Al-Humydi, A.; Garrison, J. C.; Youngs, W. J.; Collins, S. Organometallics 2004, 24, 193−196. (18) Nele, M.; Pinto, J. C.; Mohammed, M.; Collins, S. J. Polym. Sci., Part A: Polym. Chem. 2005, 43, 1797−1810. (19) Al-Humydi, A.; Garrison, J. C.; Mohammed, M.; Youngs, W. J.; Collins, S. Polyhedron 2005, 24, 1234−1249. (20) Coates, G. W.; Waymouth, R. M. Science 1995, 267, 217−219. J


Organometallics

Article

(52) Weng, W. Q.; Markel, E. J.; Dekmezian, A. H. Macromol. Rapid Commun. 2001, 22, 1488−1492. (53) Weng, W. Q.; Markel, E. J.; Dekmezian, A. H. Macromol. Rapid Commun. 2000, 21, 1103−1107. (54) Cherian, A. E.; Lobkovsky, E. B.; Coates, G. W. Macromolecules 2005, 38, 6259−6268. (55) Stadler, F. J.; Arikan-Conley, B.; Kaschta, J.; Kaminsky, W.; Munstedt, H. Macromolecules 2011, 44, 5053−5063. (56) Arikan, B.; Stadler, F. J.; Kaschta, J.; Münstedt, H.; Kaminsky, W. Macromol. Rapid Commun. 2007, 28, 1472−1478. (57) Giardello, M. A.; Eisen, M. S.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1995, 117, 12114−12129. (58) Kukral, J.; Lehmus, P.; Feifel, T.; Troll, C.; Rieger, B. Organometallics 2000, 19, 3767−3775. (59) Stapleton, R. A.; Al-Humydi, A.; Chai, J. F.; Galan, B. R.; Collins, S. Organometallics 2006, 25, 5083−5092. (60) Beachley, O. T.; Victoriano, L. Organometallics 1988, 7, 63−67. (61) Chien, J. C. W.; Tsai, W. M.; Rausch, M. D. J. Am. Chem. Soc. 1991, 113, 8570−8571.

K


Polymerization Behavior of C1-Symmetric Metallocenes (M = Zr, Hf

Recommend Documents