Synthesis of Highly Cis, Syndiotactic ROMP Polymers Using Ruthenium Metathesis Catalysts

Abstract

The first example of ruthenium-mediated ring-opening metathesis polymerization (ROMP) generating highly cis, highly tactic polymers is reported. While the cis content varied from 62 to >95% depending on the monomer structure, many of the polymers synthesized displayed high tacticity (>95%). Polymerization of an enantiomerically pure 2,3-dicarboalkoxynorbornadiene revealed a syndiotactic microstructure.

The precise control of polymer microstructure resulting from the ring-opening metathesis polymerization (ROMP) of substituted norbornenes, norbornadienes, and other strained, cyclic olefins is critical for the development of polymers with well-defined characteristics.¹ These microstructures, which include various tacticities (e.g., syndiotactic, isotactic, or atactic) and double bond configurations (cis vs. trans), have a significant impact on the physical and mechanical properties of the resulted polymer.² For example, syndiotactic cis-poly(norbornene) is a crystalline polymer with a high melting point, while atactic trans-poly(norbornene) is amorphous and low-melting in comparison.³ Accordingly, the development of olefin metathesis catalysts capable of producing highly stereoregular (>95% a single structure) ROMP polymers is of great interest.

ROMP polymers with high cis content have been synthesized previously using a variety of Re-, Os-, W- and Mo-based metathesis catalysts,^4,5 as well as more recently with a Ru-derived system.⁶ However, while many of these catalysts, particularly W- and Mo-based systems, have been shown capable of controlling both the cis/trans ratio and tacticity of ROMP polymers,^4,5 only limited tacticity control has been achieved with ruthenium.⁷ In fact, it was highlighted in a recent report by Schrock and coworkers that due to the low barrier of rotation of Ru alkylidenes and consequent inability of the Ru=C bond to enforce the steric pressures necessary to give tacticity, the likelihood of developing a Ru-based metathesis catalyst able to form polymers exhibiting high tacticity appeared increasingly minimal.^5b

We recently reported on the Z-selective ruthenium metathesis catalyst 1 containing a crucial cyclometalated N-heterocyclic carbene (NHC) ligand (Figure 1), in which the Ru-C bond is formed via C-H activation induced by the addition of silver pivalate.⁸ This catalyst was shown to give on average 80–95% cis content in the ROMP of norbornene and norbornadiene derivatives,^6b thus demonstrating for the first time the cis-selective ROMP of a wide range of monomers with a single ruthenium-based metathesis catalyst. However, all of the polymers produced by catalyst 1 were atactic.

Figure 1.

Catalysts 1–4: Mes = 2,4,6-trimethylphenyl; MIPP = 2,6-methylisopropylphenyl.

Herein, we report a new series of cyclometalated catalysts (complexes 2–4) derived from C-H activation of an N-tert-butyl group. These complexes display ROMP behavior unprecedented for ruthenium-based metathesis catalysts: Not only do these catalysts yield polymers with a generally higher cis content compared to 1 (>95% in many cases), but the polymers produced are also highly syndiotactic, further demonstrating that, similar to their W- and Mo-based counterparts, Ru-based metathesis catalysts are capable of producing polymers with a wide range of specific microstructures without the use of specialized monomers or reaction conditions.

The recent development of a milder method of effecting the salt metathesis and C-H activation of Ru metathesis catalysts using sodium pivalate has enabled the synthesis of complexes with significant alterations to the chelating N-alkyl group of the NHC that were previously inaccessible.⁹ Using this new approach, we were able to prepare the less sterically encumbered N-tert-butyl catalysts 2–4 (Figure 1).¹⁰ Single-crystal X-ray diffraction of 3 confirmed cyclometallation of the N-tert-butyl substituent, as well as bidentate coordination of the pivalate ligand. It was also revealed that the N-aryl ring is positioned such that the isopropyl substituent resides on the same face as the benzylidene. Structural parameters, including bond lengths and angles, were consistent with those for 1 and its pivalate derivative.⁸

We initiated our ROMP studies by adding 2 to a solution of norbornene (5) in THF (0.25 M) at room temperature, upon which the solution rapidly became viscous.¹¹ The isolated polymer was found to be almost exclusively cis (>95%) by ¹H NMR spectroscopy (Table 1). Furthermore, the ¹³C NMR spectrum of poly-5 prepared with 2 was found to be consistent with literature reports for highly tactic poly(norbornene).¹² In comparison, performing the same reaction at room temperature with catalyst 1 gives atactic poly(norbornene) that is only 88% cis.^6b

Table 1.

Polymerization of Monomers 5–7 with Catalysts 2–4.^a

monomer	catalyst	cis, %b	yield,%c	Mn, kDad	PDId
5	2	>95	79	605e	1.41
5	3	>95	88	521e	1.49
5	4	>95	84	424e	1.45
6	2	>95	78	--f	--f
6	3	>95	54	--	--
6	4	>95	45	--	--
7	2	72	42	--	--
7	3	62	57	--	--
7	4	86	7	--	--

^aConditions were [monomer]/[initiator] = 100:1 in THF (0.25M in substrate) at RT.

^bDetermined by ¹H NMR and ¹³C NMR spectroscopy.

^cIsolated yields.

^dDetermined by gel-permeation chromatography (GPC) with a multiangle light scattering (MALS) detector.

^eThe specific refractive index increment (dn/dc) was determined to be 0.139 ± 0.005 mL/g.

^efHere and below: Not determined because insolubility of the isolated polymer in THF precluded GPC analysis.

In order to test whether the observed tacticity control was specific to norbornene, we turned to the more complex monomer 2,3-dicarbomethoxynorbornadiene (DCMNBD, 6). The tacticity of poly(DCMNBD) is readily determined by analyzing the C(7) region of the ¹³C NMR: multiple resonances correspond to an atactic polymer, such as the poly(DCMNBD) produced by 1 (Figure 3a), while a singularly tactic polymer would be expected to display only one peak in this region due to symmetry.¹³ Poly-6 produced by catalyst 2 (>95% cis; Table 1) was found to be highly tactic, in that the corresponding ¹³C NMR spectrum contained primarily one carbon resonance in the C(7) region, consistent with a tacticity of >95% for the all-cis triads (Figure 3b).

Figure 3.

The C(7) ¹³C NMR region of poly(DCMNBD) (poly-6) prepared with a) 1 (86% cis, atactic) and b) 2 (>95% cis, >95% tactic).

To probe the effect of symmetry of the N-aryl group on tacticity,¹⁴ we evaluated two catalysts (3, 4) containing a cyclometalated N-tert-butyl group similar to 2 but with an asymmetric N-aryl group (Figure 1). The geminal dimethyl backbone of 4 was installed to prevent any rotation of the N-aryl group that might be occurring in 3. Polymerization of 6 with catalysts 3 and 4 gave poly-6 that was also >95% tactic in both cases as determined by ¹³C NMR (Table 1). Poly-5 produced by these catalysts, was also highly cis (>95%), albeit slightly less tactic in both cases than poly-5 generated with 2.

As had previously been observed with 1, the experimental number-average molecular weights (M_n) for poly-5 prepared using catalysts 2–4 were significantly higher than the theoretical values (Table 1). This is indicative that k_p of 5 exceeds k_i of 2–4, which would lead to the broad PDIs observed; this is likely a result of incomplete catalyst initiation and might be expected based on the relatively low initiation rate constants of 2–4.¹⁵

To elucidate the absolute tacticities of the norbornene- and norbornadiene-derived polymers, we employed chiral monomer 7. Due to the lack of mirror planes relating the monomeric units in the resulting polymers, it is expected that if the polymers produced by 2–4 were cis, isotactic, the olefinic protons would be inequivalent (Figure 4a).¹⁶ As such, we should observe a coupling characteristic of olefinic protons by NMR spectroscopy. Conversely, in a cis, syndiotactic polymer, the cis olefinic protons would be related by a C₂ axis passing through the midpoint of the double bond and would therefore be equivalent and not coupled (Figure 4b). While poly-7 produced by catalysts 2–4 was only 62–86% cis (Table 1), the tacticity of the all-cis triads remained very high for all three catalysts (see Figures S14 and S15 in Supporting Information). Furthermore, the cis olefinic protons in the isolated polymers were uncoupled,¹⁷ strongly suggesting that poly-7 is syndiotactic in all cases (Figure 5).

Figure 4.

Olefinic protons in the two possible highly-cis, regular polymers made from an enantiomerically pure 2,3-disubstituted norbornadiene.

Figure 5.

COSY spectrum of poly-7 prepared with 2 in the olefinic proton region. The absence of olefinic coupling suggests that the polymer is syndiotactic.

One plausible explanation for the observed tacticity control is that the rate of monomer incorporation is faster than the rate of carbene epimerization.¹⁸ This is presumably in contrast to catalyst 1, which, likely due to the slow rate of propagation relative to epimerization, generates atactic polymers. This distinction might be explained by the relative differences in size and symmetry of the carbon chelate in catalysts 1 and 2. The reduced bulk of the cyclometalated N-tert-butyl group of 2 (vs. the N-adamantyl chelate in 1) likely results in fewer unfavorable steric interactions between the catalyst and the approaching bulky norbornene and norbornadiene derivatives, thus in- creasing the overall rate at which the monomers are incorporated.¹⁹ The local symmetry about the Ru-C bond in 2–4 is postulated to account for the syndioselectivity.^14,18

Finally, we sought to briefly explore the physical properties of tactic ROMP polymers in comparison to their atactic counterparts via differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA). The T_g of atactic trans-polynorbornene is 37 °C.^3a As expected, the T_g of poly-5 was significantly higher at ca. 70 °C, consistent with a higher packing order due to the increased stereoregularity of the polymer. Both the atactic, trans polymer and the syndiotactic, cis poly-5 decomposed at ca. 430 °C (see Supporting Information).

In spite of expectations to the contrary, we have demonstrated the ability of Ru-based metathesis catalysts to yield highly cis, highly tactic polymers. ROMP of a chiral norbornadiene monomer suggested that these polymers are syndiotactic. While it appears that the tacticity of these polymers is derived from the installation of a comparatively small, symmetric N-tert-butyl group, the exact role of these factors in the control of the tacticity of polymers produced by cyclometalated Ru-based systems remains to be determined.

Figure 2.

Solid-state structure of 3, with thermal ellipsoids drawn at 50% probability. For clarity, hydrogen atoms have been omitted. Selected bond lengths (Å) for 3: C1-Ru 1.932, C5-Ru 2.071, C18-Ru 1.798, O1-Ru 2.334, O2-Ru 2.202, O3-Ru 2.398.