Introduction

Since the development of efficient catalyst systems for the syndiospecific polymerization of styrene, for example, trichlorocyclopentadienyltitanium/methylaluminoxane (CpTiCl3/MAO), syndiotactic polystyrene (SPS) has been hailed as a new-age engineering plastic.1, 2, 3, 4, 5 SPS exhibits rapid crystallization behavior and a high melting point (∼270 °C); it also has a high heat resistance, a high chemical resistance, a low density, a high modulus of elasticity and a low dielectric constant.6 However, the brittleness of SPS must be reduced for it to be of practical use.

Copolymerization of styrene and an olefinic monomer (for instance, ethylene or propylene) is expected to be a good modification of SPS to improve its toughness. It has been reported that the typical SPS catalyst system, similar to the CpTiCl3/MAO system, is inadequate for the copolymerization protocol; the copolymerization of ethylene and styrene by the CpTiCl3/MAO catalyst system results in the production of a mixture of SPS and polyethylene (PE) or a mixture of SPS, PE and copolymer.7, 8 Recently, several researchers have reported the development of more efficient catalysts for the copolymerization of ethylene and styrene: a half-titanocene featuring an aryloxy-type ligand, a titanium complex having a pendant-type cyclopentadienyl ligand and a rare-earth metal-based catalyst.9, 10, 11, 12, 13, 14, 15, 16, 17

However, there have been only limited reports on the synthesis of copolymers containing long syndiotactic styrene–styrene sequences. Hou and co-workers18 have reported that rare-earth metal-based catalyst systems promote not only the syndiotactic homopolymerization of styrene, but also copolymerization with ethylene, in which the Cp′Sc(CH2SiTMS)2/Ph3C·B(C6F5)4 [Cp′=C5Me4(CH2TMS)] catalyst system produces a multi-block copolymer consisting of a crystalline SPS part and an ethylene-rich part. Carpentier and co-workers19 have also reported an ansa-lanthanidocene catalyst that can synthesize a copolymer containing long syndiotactic styrene–styrene sequences separated by a single or a few ethylene units. Meanwhile, it has been previously reported that the copolymeric fraction separated from the mixture obtained by the CpTiCl3/MAO system contains a similar long syndiotactic styrene–styrene sequence.8

Problems associated with the copolymerization of styrene and ethylene by a titanium catalyst appear to arise from a difference of active species (and also propagation mechanisms). It has been traditionally believed that a trivalent titanium (Ti3+) species is the predominant active species for the syndiotactic polymerization of styrene by the CpTiCl3/MAO type catalysts.20, 21, 22, 23, 24, 25, 26 However, recent studies have indicated that a tetravalent titanium (Ti4+) species is also generated and can similarly promote styrene polymerization in the CpTiCl3/MAO type catalyst system, independent of the Ti3+ active species.27, 28, 29, 30, 31 An analogous Ti4+ species is believed to be the active species in olefin polymerization by metallocene-type catalyst systems (such as Cp2TiCl2/MAO).32 Nomura and others12, 33 have reported that a half-titanocene catalyst, which has an aryloxy-type ligand, conducts the copolymerization of ethylene and styrene, resulting in effective synthesis of the random copolymer. They noted that the active species of this catalyst system might be the Ti4+ species, similar to olefin polymerization.

In addition to syndiotactic poly(ethylene-co-styrene), several papers have described copolymers containing a long styrene–styrene sequence. Okuda and co-workers34 reported the synthesis of poly(propylene-co-styrene) that consisted of isotactic styrene–styrene blocks and amorphous propylene blocks; this was synthesized by means of copolymerization with a post-metallocene-type catalyst. Additionally, Longo and co-workers35 reported the blocky stereospecific copolymerization of styrene and 1,3-butadiene by the half-titanocene-type catalyst.

We inferred from these studies that the Ti3+ in the titanium-based catalyst system is essential for the formation of the SPS structure and is also able to promote the incorporation of ethylene into the polystyrene backbone. Therefore, the trivalent titanium complex shows promise as a precursor of the copolymerization catalyst. In this paper, we report our study of the copolymerization of styrene and ethylene by a trivalent titanium complex-based catalyst system, tris(acetylacetonato)titanium (Ti(acac)3)/MAO, and we also report the characterization of the obtained polymer.

Experimental procedure

Materials

The methylaluminoxane (triisobutylaluminum-modified methylaluminoxane; MMAO) was provided by Tosoh-Finechem (Shunan, Japan) and used without further purification. Commercially obtained research-grade toluene was purified according to conventional procedures. Research-grade styrene was dried by calcium hydride and then distilled under reduced pressure before polymerization. Commercially available ethylene, provided by Japan Fine Products (Kawasaki, Japan), was dried in a dry column DC-A4 (Nikka Seiko, Tokyo, Japan). The Ti(acac)3 was synthesized by the reaction of trichlorotitanium and lithium acetylacetonate in the tetrahydrofuran and then recrystallized from distilled toluene. Other solvents and reagents were commercially obtained and used without further purification.

Polymerization

Ethylene polymerization

Prescribed amounts of toluene, the Ti(acac)3/toluene solution and the MMAO were added to a 100-ml stainless steel autoclave equipped with a magnetic stirrer. After the contents of the reactor were degassed by vacuum, a prescribed amount of gaseous ethylene was introduced at −196 °C (controlled by a liquid nitrogen bath). The batch-wise polymerization was initiated by increasing the temperature to 25 °C (controlled by a water bath equipped with a thermo controller). The polymerization was stopped by the addition of 2-propanol after the reactor was degassed. Then the reaction products were poured into a dilute HCl/methanol solution. The precipitated polymer was filtered and washed with methanol and then dried under vacuum at 60 °C for 6 h.

Styrene polymerization

Prescribed amounts of toluene, the MMAO and the distilled styrene were added to a 100-ml three-necked round-bottom glass flask equipped with a magnetic stirrer. Polymerization was initiated by addition of the Ti(acac)3/toluene solution at 25 °C, then stopped by addition of 2-propanol. The reaction products were then poured into a dilute HCl/methanol solution. The precipitated polymer was filtered and washed with methanol and then dried under vacuum at 60 °C for 6 h.

Ethylene/styrene copolymerization

Prescribed amounts of toluene, the distilled styrene and the MMAO were added to a 100-ml stainless steel autoclave equipped with a magnetic stirrer. Next, the prescribed amount of the Ti(acac)3/toluene solution was added at −196 °C (controlled by a liquid nitrogen bath). Gaseous ethylene was also introduced at −196 °C after the reactor was degassed by vacuum. The batch-wise polymerization was started by increasing the temperature to a predefined value (controlled by a water bath equipped with a thermo controller). The polymerization was stopped by the addition of 2-propanol after the reactor was degassed. The reaction products were then poured into a dilute HCl/methanol solution. The precipitated polymer was filtered and washed with methanol and then dried under vacuum at 60 °C for 6 h. The resultant polymer was Soxhlet extracted for 6 h with boiling solvent as appropriate. The soluble portion extracted was precipitated by methanol and then filtered and dried.

Polymer analysis

The number-average molecular weights (Mn) and the weight-average molecular weights (Mw) of the polymers were measured by gel-permeation chromatography (GPC) at 140 °C on a Polymer Laboratories PL GPC-220 (Polymer Laboratories, Santa Clara, CA, USA) equipped with refractive index detector, using 1,2,4-trichlorobenzene as a solvent. Calibration was done by Polymer Laboratories detector software Cirrus with a polystyrene standard. The 1H and 13C nuclear magnetic resonance (NMR) spectra of the polymers were recorded at 120 °C on a JEOL JNM-LA600 spectrometer (JEOL, Tokyo, Japan) operating at 600 and 150 MHz, respectively, in the pulse Fourier transform mode. Sample solutions were made in tetrachloroethane-d2 up to 10% by weight. The resonances of the tetrachloroethane-d2 (1H=5.91 p.p.m., 13C central peak=74.47 p.p.m.) were used as internal references. The melting points (Tm) and the heats of fusion (ΔH) were measured using a SII DSC6200 (SII, Chiba, Japan). The samples were encapsulated in aluminum pans and scanned at 10 °C min−1 up to 300 °C. The first heating data were used for the analysis in order to avoid the influence of heat degradation. The decomposition point of 5% weight loss was determined using a SII TG/DTA6200 (SII).

Results and Discussion

Table 1 shows the results of homo- and copolymerization of styrene and/or ethylene by the Ti(acac)3/MMAO catalyst system. The Ti(acac)3, trivalent titanium complex has been reported as a good precursor of the catalyst for the syndiospecific styrene polymerization,36 as shown in Run 1. Ethylene polymerization by the same catalyst system produced PE with appreciably lower activity in comparison with a typical metallocene catalyst system such as the Cp2ZrCl2/MAO. The PE obtained was revealed to be the linear type PE by 13C NMR analysis (see Supplementary Information).

Table 1 Homo- and copolymerization of ethylene and styrene by Ti(acac)3/MMAO catalyst systema
Full size table

The copolymerization of styrene and ethylene gave a powdered polymer with moderate activity similar to the styrene homopolymerization product. 1H NMR analysis of the crude product revealed that both the ethylene and the styrene units were incorporated (see Supplementary Information), regardless of whether it was a homogeneous copolymer or a mixture of the homopolymer. The styrene/ethylene ratio in the crude product was influenced by the polymerization conditions. Runs 3 and 4 indicated dependence on the MMAO concentration when the amount of Ti was fixed at 0.01 mmol. The styrene content of the crude product approximately doubled from 8.6 to 16.5 mol% by using double the amount of MMAO (2000 mmol instead of 1000 mmol), whereas the yield increased slightly from 0.24 to 0.30 g. Runs 5–9 showed a dependence on the amount of Ti(acac)3 with a fixed amount of MMAO. The yield increased correspondingly with increased amounts of Ti(acac)3, although the activity decreased slightly, perhaps due to the decreased Al/Ti ratio. The styrene content also increased by increasing the amount of Ti(acac)3, meaning that the consumption of styrene increased by considerably more than the ethylene consumption.

Table 2 shows the dependence on the polymerization temperature in the copolymerization, under the same conditions of 0.02 mmol of Ti(acac)3 and 20 mmol of MMAO for 120 min. At temperatures <20 °C, the activity was relatively low, <10 kg mol−1·Ti·h, and the styrene content of the crude product was not significantly influenced by the temperature (Runs 13–15). At temperatures >20 °C, however, the yield increased rapidly with temperature, and the styrene content also increased significantly (Runs 16–18), up to 78.5 mol% at 50 °C. The results indicate that the increase in temperature influenced the consumption of styrene more strongly than that of ethylene, resulting in the production of a highly styrene-rich crude product at higher temperatures.

Table 2 Dependence on the polymerization temperature in the copolymerization of ethylene and styrene by Ti(acac)3/MMAO catalyst systema
Full size table

To determine the polymer that comprised the crude product, GPC analysis was performed at 140 °C using 1,2,4-trichlorobenzene as a solvent. The GPC curve of homopolystyrene (Run 1) was obtained as shown in Figure 1a. However, the homo-PE (Run 2) could not be analyzed because the GPC column was blocked, perhaps due to the super-high molecular weight. The boiling 1,1,2,2-tetrachloroethane (TCE)-soluble fraction of the PE (Run 2) was measured and is shown in Figure 1b; in contrast to the polystyrene, this represents the downside peak, due to the difference in refractive indices. As shown in Figure 1c, the crude product of the copolymerization (Run 10) exhibited a bimodal distribution located at opposite sides of the baseline. This implies the existence of binary polymers: one is styrene-rich and has a lower molecular weight, and the other is ethylene-rich and has a higher molecular weight.

Figure 1
figure 1

Gel-permeation chromatography curves of the crude polymers obtained using the tris(acetylacetonate)titanium/triisobutylaluminum-modified methylaluminoxane catalyst system. (a) Polystyrene (Run 1), (b) polyethylene (Run 2, boiling 1,1,2,2-tetrachloroethane (TCE)-soluble part), (c) poly(ethylene-co-styrene; Run 10).

Full size image

The above results indicate that at least two polymerization mechanisms exist in the Ti(acac)3/MMAO catalyst system. Traditional belief holds that the syndiospecific styrene polymerization proceeds with the trivalent Ti3+ species in a typical catalyst system such as the CpTiCl3/MAO.26, 27, 28, 29, 30 However, several papers have now reported that tetravalent Ti4+ active species are valid for the random copolymerization of styrene and an olefinic monomer in a number of titanium catalysts; for example, the group 4 metallocene system.12, 34 Although the precise mechanism is not clear at present, it is plausible that the titanium of the Ti(acac)3/MMAO catalyst system adopts multiple oxidation states during polymerization, with these states promoting the propagation of different polymers.

We attempted to separate these components by Soxhlet extraction using a number of solvents: o-dichlorobenzene, TCE, xylene (mixture), methylethylketone and n-heptane. The high boiling o-dichlorobenzene largely dissolved the crude product in many cases. By contrast, a large part of the crude product was insoluble in the low boiling methylethylketone and n-heptane. TCE and the xylene proved suitable for separating the products. The GPC and 1H NMR analysis of the extracted fraction were carried out and are summarized in Figure 2 and Table 3. The GPC curves of the soluble and insoluble fractions of the product (Run 11) in the boiling TCE are shown in Figure 2a. The insoluble fraction was revealed to consist primarily of the ethylene-rich high-molecular weight polymer, whereas the soluble fraction mainly consisted of the styrene-rich low-molecular weight polymer, although the soluble fraction also contained the ethylene-rich low-molecular weight polymer as observed in Figure 1b. The boiling xylene extraction of the crude product (Run 7) resulted in similar separation, as shown in Figure 2b. The xylene-soluble fraction mainly consisted of styrene-rich low-molecular weight polymer, whereas the insoluble fraction consisted of the ethylene-rich high-molecular weight polymer and a considerable amount of styrene-rich polymer.

Figure 2
figure 2

Gel-permeation chromatography curves of the fraction obtained by Soxhlet extraction. (a) The soluble and the insoluble portions in boiling 1,1,2,2-tetrachloroethane (TCE; Run 11). (b) The soluble and the insoluble part in boiling xylene (Run 7).

Full size image
Table 3 GPC and 1H NMR data for the fractions of Soxhlet extractiona
Full size table

The microstructure of these extracted fractions, the ethylene-rich polymer and the styrene-rich polymer were investigated by 13C NMR analysis. Assignments of the resonances of the copolymer were carried out according to the previous reports.8, 18, 20, 21, 22, 23, 24, 25 The spectrum of the insoluble part, obtained by TCE extraction of the crude product in Run 11, is shown in Figure 3. In addition to continuous homogeneous monomer sequences consisting of ethylene–ethylene (30.1 p.p.m.) and a small amount of styrene–styrene (41.7 and 44.9 p.p.m.), the isolated styrene unit in the PE backbone was observed at 46.6, 37.2 and 28.0 p.p.m. This indicates that the insoluble part is mainly composed of the random copolymer of ethylene and styrene. The spectrum of the soluble part, obtained by xylene extraction of the crude product in Run 7, is shown in Figure 4. This spectrum illustrates the strong SPS backbone, along with the resonances (44.3, 45.3, 43.9, 37.7 and 27.6 p.p.m.) assignable to the specific ethylene–styrene sequences that can be attributed to the connection between the continuous styrene–styrene sequence and the ethylene monomer units (at least 2 units). The resonances assignable to the ethylene–styrene random copolymer were also weakly observable. By comparison, the traditional catalyst system, CpTiCl3/MMAO, produced only a mixture of PE and SPS under similar polymerization conditions (Run 12, see Supplementary Information).

Figure 3
figure 3

13C nuclear magnetic resonance spectrum of the insoluble fraction in boiling 1,1,2,2-tetrachloroethane (Run 11).

Full size image
Figure 4
figure 4

13C nuclear magnetic resonance spectrum of the soluble fraction in boiling xylene (Run 7).

Full size image

Table 4 lists the sequence ratio of the Soxhlet-soluble fractions, which were determined by the intensities of 13C NMR resonances. The ratio between the SPS backbone and the joint part (continuous styrene–styrene sequence adjacent to the ethylene unit), [SSS]/[SSE], was determined from the methylene carbon of the styrene unit observed at 41.7 and 43.9 p.p.m. Concerning the ethylene part, [EEE]/[ESE]/[SSE] was similarly determined from the methylene carbon at 30.1, 37.2 and 37.7 p.p.m. The ratio of the SPS backbone to the joint part can be considered as a fraction of the length of the SPS block if the block-type copolymer was formed, and can be expected to correlate with the styrene content or the properties of the soluble fraction. Figure 5a shows the relation between [SSS]/[SSE] and the styrene content of the Soxhlet-soluble fraction, plausibly indicating that the length of the styrene–styrene sequence proportionally increases with the styrene content of the copolymers.

Table 4 Sequences of the soluble fractions of Soxhlet extraction
Full size table
Figure 5
figure 5

Relation of the ratio between the syndiotactic polystyrene backbone and the continuous styrene–styrene sequence adjacent to the ethylene unit, [SSS]/[SSE], and (a) the styrene content, (b) the Tm of the Soxhlet-soluble fraction.

Full size image

The thermal properties of the extracted fractions were investigated by differential scanning calorimeter (DSC) and are listed in Table 5. For analysis of the melting point (Tm) and the heat of fusion (ΔH), the first heating data of DSC was used in order to avoid the influence of heat degradation. Figure 6 shows the DSC curves of SPS (Run 1), PE (Run 2) and the xylene-soluble fraction of the copolymerization product (Run 7). The values of Tm (and ΔH) of the PE and the SPS were 126.2 °C (167.7 mJ mg−1) and 271.5 °C (32.9 mJ mg−1), respectively. The xylene-soluble fraction exhibited a sufficiently high Tm of approximately 260 °C, which is slightly less than pure SPS, whereas the lower peak was observed at ∼115 °C. Figure 5b shows the relation between the Tm and the [SSS]/[SSE] of the Soxhlet-soluble fraction. It reveals that the Tm increases proportionally with the length of the continuous styrene sequence, which is similar to the relation between [SSS]/[SSE] and styrene content. These correlations imply that the soluble part is expected to be the block-like copolymer, which consists of the long syndiotactic styrene–styrene sequence divided by the ethylene-based section. The correlation between the length of the ethylene–ethylene sequence and the Tm seems ambiguous; however, this may be because this part contains the isolated styrene unit, or it may be because of the contamination of the other random copolymer influences in the NMR or DSC analysis.

Table 5 DSC and 13C NMR analysis data for the fractions of Soxhlet extraction
Full size table
Figure 6
figure 6

Differential scanning calorimeter curves of the polymers obtained by the tris(acetylacetonate)titanium/triisobutylaluminum-modified methylaluminoxane catalyst system. (a) Copolymer of styrene and ethylene (Run 7, the soluble fraction in boiling xylene), (b) polystyrene (Run 1) and (c) polyethylene (Run 2).

Full size image

Figure 7 shows the results of thermogravimetric/differential thermal analysis of the SPS, the PE and the copolymer extracted by xylene (Run 18). The SPS rapidly degraded at ∼400 °C. The decomposition point of 5% weight loss (Td5) of the SPS was 401.6 °C. The PE gradually decomposed from under 200 °C, with a Td5 of 271.5 °C. The copolymer exhibited a high Td5 value (398.2 °C), with rapid degradation around Td5, which was reasonably close to that of pure SPS. This indicates that the SPS-based copolymer has a high thermal stability, comparable with pure SPS.

Figure 7
figure 7

Thermogravimetric (TG) curves of the polymers obtained by the tris(acetylacetonate)titanium/triisobutylaluminum-modified methylaluminoxane catalyst system. (a) Copolymer of styrene and ethylene (Run 18, the soluble fraction in boiling xylene), (b) polystyrene (Run 1) and (c) polyethylene (PE, Run 2). SPS, syndiotactic polystyrene.

Full size image

Conclusions

We investigated the copolymerization of styrene and ethylene by a trivalent titanium complex-based catalyst system. GPC measurement revealed that the resultant polymer is a mixture of two different copolymers. 13C NMR analysis of the Soxhlet-extracted fractions of the product implied that the insoluble fractions consist of a random copolymer of ethylene and styrene, whereas the soluble parts consist of a long syndiotactic styrene–styrene sequence and an ethylene-based section. In the Soxhlet-soluble fraction, the ratio of the continuous styrene–styrene sequence to the styrene–ethylene joint sequence increased with the styrene content. The Tm of the Soxhlet-soluble fraction also increased with the [styrene–styrene]/[styrene–ethylene] ratio. These results together indicate that a block-like copolymer, which contains long SPS portions separated by an ethylene–ethylene sequence, was produced. DSC and thermogravimetric/differential thermal analysis measurements show that the SPS-based copolymer has high thermal stability comparable with pure SPS.

These results indicate that the trivalent titanium catalyst offers promising prospects for the creation of an SPS-based block copolymer. This SPS-based copolymer can be expected to show good performance in industrial applications, although further research into its physical properties, such as mechanical strength, impact strength and viscoelasticity, is necessary. We expect that further investigation toward the refinement of the catalyst system and the evaluation of mechanical properties to develop a tough SPS-based copolymer will be reported in due course.