Nickel-Catalyzed Alkene Difunctionalization as a Method for Polymerization

Abstract

Alkene dicarbofunctionalization is a rapidly emerging tool for complex molecule synthesis that installs two carbon fragments regioselectively across an alkene. This method has the potential to engineer stereo-defined polymers; yet the application of difunctionalization reactions to polymer synthesis remains unexplored. Herein, we describe the first example of a Ni-catalyzed difunctionalization of alkenes with arylboronic esters and aryl bromides innate to the alkene. The polymerization reaction proceeds regioselectively with addition of the aryl bromide to the terminal alkenyl carbon and arylboronic ester to the internal benzylic carbon. The resultant poly[p-arylene-α-(aryl)ethylene]s comprise aryl groups installed at regular intervals along the polymer backbone through chain propagation in two directions. Polymers with molecular weights generally ranging from 30–175 kDa were obtained after successful fractionation from oligomeric species. Thermal analysis of the poly[p-arylene α-(aryl)ethylene]s revealed stability up to ~399 °C, with a T_g of 90 °C, both of which are similar in value to poly(styrenes) and poly(phenylene methylene)s.

Graphical Abstract

Transition metal-catalyzed alkene difunctionalization reactions are emerging, yet prominent, methods that rapidly build molecular complexity from readily available feedstock chemicals.¹ In small molecules, these methods can assemble two to three organic units with complete regio/stereo-fidelity and functional group compatibility in a predefined and modular manner. Despite an ability to stitch together synthons leading to complex molecules, coupled with high functional group tolerance, translation of alkene difunctionalization to polymers or materials synthesis remains unexplored.

Since alkene difunctionalization methods are highly regioselective, they have the potential to address growing and unmet challenges of engineering stereo-defined polymers with diverse compositions and architectures. Methods that can sustain extensive functional group orthogonality while engendering stereoregular polymers are also limited,² with ring-opening polymerization³ and ring-opening metathesis polymerization⁴ comprising two of the more tolerant methods to achieve stereoregular polymers; however, both require cyclic monomers, limiting the scope of polymers that can be accessed. Adaption of alkene dicarbofunctionalization methods to polymerization could address these challenges in a modular way using common and commercial styrenics. In particular, this strategy would introduce a new polymerization method that allows for simultaneous incorporation of both functional monomers and side chain units (Scheme 1A). On the basis of this concept, we demonstrate the first interrogation of alkene difunctionalization for the synthesis of functionalized polymers wherein bromostyrene derivatives and arylboronic esters polymerize in a bidirectional manner to achieve poly[arylene α-(aryl)ethylene]s (PAAEs).

Scheme 1.

(A) Diagrammatic representation of alkene dicarbofunctionalization for polymerization; (B) our work: alkene 1,2-diarylation, illustrating propagation in two directions; (C) potential regioreversed polymer; and (D) competitive polymerization side product.

Prior reports featuring the synthesis of similar hydrocarbon-based polymers (e.g., poly(phenylene alkenylene)s) rely on methods such as acyclic diene metathesis polymerization in the presence of either Schrock or Grubbs initiators, followed by hydrogenation.⁵ While the resulting polymers achieved M_n values exceeding 10,000 g mol⁻¹ when using Ru-based initiators, they rely on both high temperatures and expensive initiators. Furthermore, they cannot imbue modular side chains. Methods that are versatile and able to provide access to a wide variety of unprecedented polymers in a more high-throughput, mild, and cost-effective means are needed.⁶

We hypothesized that the bifunctional monomer could be leveraged for difunctionalization with indigenous aryl bromide and exogenously-functionalized organometallic reagents, leading to a polymer chain growing along both the bromide and alkene termini (Scheme 1B). Following literature protocols for styrene 1,2-diarylation⁷ and upon extensive optimization (see SI and vide infra for details), we discovered that 5 mol % NiBr₂ catalyzed the polymerization of 4-bromostyrene with phenylboronic acid neopentylglycol ester (FG¹, FG² = H) at 100 °C in toluene, affording a poly(phenylene α-arylethylene) in 72 % yield after 8 h. The structure of the poly(phenylene arylethylene) was unambiguously confirmed using ¹H and ¹³C NMR spectroscopy (see SI), wherein the dibenzylic methine and benzylic methylene peaks of the polymer at 4.04 and 3.19 ppm were compared with that of 1,1,2-triarylethane⁸ (Figure 1).⁹ Successful polymerization was further evidenced using size-exclusion chromatography (SEC, Figure 2) wherein the polymer apparent molecular weight was found to be 59.2 kDa. SEC traces indicated a moderate polymer dispersity (Đ = 1.86), with oligomeric products also being formed (13.8 kDa; 27 % of total mass yield).

Figure 1.

Representative ¹H NMR spectra of electronically-diverse PAAEs. The inset highlights chemical shift changes of the dibenzylic proton resulting from p-H, p-OMe and p-CF₃ pendants.

Figure 2.

SEC traces demonstrating successful fractionation of targeted PAAE from oligomer through reprecipitation (solvent: THF).

Reaction of 4-bromostyrene with arylboronic esters bearing p-CF₃ (p-CF₃C₆H₄) and p-OMe (p-MeOC₆H₄) substituents generated the corresponding polymers in which p-CF₃C₆H₄ and p-MeOC₆H₄ groups were incorporated at the benzylic position. This was evidenced by ¹H NMR spectroscopy, wherein downfield and upfield shifts were observed for the methine protons in p-CF₃ and p-OMe substituted polymers, respectively, relative to the unsubstituted polymer (Figure 1, inset). This regioselectivity pattern is consistent with a Ni(0)/Ni(II) catalytic cycle for alkene difunctionalization⁷ that proceeds through initial oxidative addition of the aryl bromide, followed by migratory insertion of an alkene, and subsequent transmetalation with ArB(OR)₂ and carbon-carbon (C-C) bond-forming reductive elimination. In addition, the polymerization can also be affected with Ni(cod)₂ with similar efficacy, further supporting a Ni(0)/Ni(II) catalytic cycle.

Reaction parameters were optimized with respect to temperature, catalyst precursor, catalyst loading, reagent stoichiometry, solvent, base, boronic ester derivative, concentration, and time (Table S-1) to arrive at optimal conditions for polymerization. Extensive optimization allowed us to completely suppress the competing homopolymerization of 4-bromostyrene (Scheme 1D), which otherwise adversely affected the yield and purification of the targeted PAAE. For example, use of a large excess of base (≥ 3 equiv.) and/or the presence of oxygen generated poly(4-bromostyrene). These conditions favor the competitive alkoxide-promoted anionic polymerization, which was also observed with a variety of substituted styrenes (Table S-1).¹¹

Conditions that achieved higher percent compositions of polymer, compared to oligomer, along with reasonable molecular weights and dispersity were selected for further studies. Using optimized conditions, the resulting product is comprised of 73 % polymer with an apparent molecular weight of approximately 60 kDa and dispersity around 1.8. Furthermore, the polymer is readily fractionated from the corresponding oligomers through selective reprecipitation in ether (Figure 2).

The polymerization proceeds rapidly, reaching completion in 8 h under these optimized conditions. Moreover, 4-bromostyrene, is completely consumed within 2 hours, as evidenced by ¹H NMR spectroscopy; however, increases in product molecular weight are observed until about 8 h. We attribute further increases in molecular weight to combination of oligomer active chain ends for the duration of the catalyst lifetime (Figure S-59), after which a mixture of polymeric and oligomeric species results. This is consistent with the oligomers demonstrating lower dispersity and suggestive of the oligomer active chain ends combining as the duration of polymerization increases.

As modular methods are desirable for targeting polymer properties, we then sought to determine if control over molecular weight and dispersity could be demonstrated. While the rapid, two-direction growth of the polymer was expected to preclude tight control of dispersity, we anticipated that lower catalyst loading could result in higher molecular weight polymers at the expense of increased dispersity, as fewer catalyst species would be available to initiate polymer growth from monomers. To examine these parameters, we selected the simplest system as a model, using 4-bromostyrene and phenylboronic acid neopentylglycol ester. When decreasing catalyst loading between 20 and 0.5 mol %, this was generally observed to be the case, with a two-fold increase in M_n. However, Đ was only observed to increase until 5 mol % catalyst loading; lower catalyst loadings had negligible effects on Đ. It should be noted that low catalyst loadings of 1 mol % or less resulted in poor control. Concurrently, the percent composition of polymer versus oligomer decreased with higher catalyst loading, with 10 and 20 % catalyst loading leading to drastically lower percentages of polymer product (Figure 3), though total mass yield remained consistent.

Figure 3.

Graphical representations of: (top) the impact of catalyst loading on molecular weights and dispersity; and (bottom) mass yield and distribution of polymer and oligomer.

Unsurprisingly, reaction concentration impacted molecular weight and dispersity, with higher concentrations resulting in a thirteen-fold increase in molecular weight from 28 kDa at 0.1 M to 373 kDa under 0.5 M conditions. This is attributed to the increased proximity of catalyst species to available polymer active chain ends upon each turnover of the catalytic cycle, which likely promotes further propagation of active chains, as opposed to new chain formation. We also observed increases in dispersity with concentration, which we attribute to the influence of proximity of active species leading to more stochastic initiation, propagation, and chain combination (Figure 4).

Figure 4.

Graphical representations of the impact of concentration on: (top) molecular weights and dispersity; and (bottom) mass yield and distribution of polymer and oligomer.

Our alkene difunctionalization polymerization method also demonstrated a notable tolerance to functional groups and substitution pattern, leading to an ability to staple together a diverse array of styrenes and boronic ester-derived pendants (Table 1). We have successfully employed boronic esters functionalized in the para position with methoxy, methyl, fluoro, trifluoromethyl, and ethyl ester groups (entries 2–9). We noted that electron deficient boronic esters had deleterious effects on yields, molecular weights, and polymer:oligomer ratios compared to electron rich boronic esters. The polymerization was tolerant to electron donating substituents in the meta position (i.e., m-methyl and m-methoxy) and ortho position (i.e., o-methoxy) on the boronic ester comonomer (entries 7–9). Our method also tolerated ortho and meta substituents, like methoxy and fluoro, on the styrene (entries 10–12). Furthermore, substituents could be simultaneously varied in both the arylboronic ester and styrene as indicated by the polymerization of m-methoxyphenylboronic ester with p-bromo-o-fluorostyrene and o-methoxyphenylboronic ester with p-bromo-m-fluorostyrene (entries 11–12). However, the substitution pattern of the unfunctionalized bromostyrene significantly impacted the polymerization, likely owing to steric encumbrance around active chain ends. Meta-bromostyrene produced smaller polymers, around 28 kDa, while o-bromostyrene did not afford any polymer (entry 13).

Table 1.

Scope of polymerization by alkene difunctionalization reaction^a

Entry	Polymer	Mass Yield (%)	Composition Polymer (%)	Mn (kDa)b	Ðb	Entry	Polymer	Mass Yield (%)	Composition Polymer (%)	Mn (kDa)b	Ðb
1		72	73	59.2	1.86	8		67	57	117.5	1.84
2		67	55	45.3	1.87
3		67	43	67.7	2.11	9		62	77	174.7	2.30
4		66	38	50.8	1.70	10		67	84	79.0	2.23
5		40	100	30.7	1.96	11		35	87	78.4	1.93
6		32	49	60.2	1.98	12		44	100	54.8	1.99
7		72	32	31.7	1.48	13		39	53	32.2	1.79

Conditions: ^aReactions were run in 0.1 mmol scale in toluene at 100 °C for 8 h with 5 mol % NiBr₂, 2.3 equiv. KOEt and 2.0 equiv. ArB⁡(OR)₂ unless stated otherwise.

^bThe Mn and Ð values were determined from SEC traces of polymer/oligomer mixture.

With our methodology enabling a robust array of polymers to be achieved, we sought to explore the fundamental thermal properties of the model PAAE polymer (Table 1, Entry 1). Thermogravimetric analysis revealed an extrapolated degradation onset temperature of 399 °C, and the material exhibited a final mass loss of 72 % upon reaching 493 °C (Figure S-46). Furthermore, differential scanning calorimetry revealed a glass transition temperature (T_g) of 90 °C, and no melting events, indicating an amorphous material. Both the resultant decomposition temperature and T_g fell between reported values of poly(styrene),¹² and poly(phenylene methylene).¹³ Given the structural similarity to these models, our polymers arising from alkene difunctionalization could be used for similar applications. Furthermore, the amorphous nature of the polymer indicates the polymer is atactic. This is in good agreement with the proposed mechanism, which proceeds through primarily syncarbometallation across the alkene followed by stereoretentive reductive elimination but does not exclude anti-addition at the second and subsequent iterations of polymer chain propagation.^3a

Our PAAEs were observed to fluoresce in the presence of UV light, even at the low intensity of ambient light. This behavior is akin to poly(phenylene methylene)s that are known to fluoresce due to homoconjugation;^13a,14 thus, we sought to explore the optical properties of the PAAE. UV-Vis and emission spectroscopy revealed that the unsubstituted polymer exhibited an excitation maximum at λ = 284 nm, with a concomitant emission maximum at λ = 320 nm; the combined behaviors gave rise to the blue color. In general, all PAAEs synthesized demonstrated similar features in their UV spectra with maximum absorbances typically in the range of 235–245 nm, along with lower energy absorbances in the 260–280 nm range. The PAAEs collectively exhibited a broad emission envelope with emission maxima between 360–371 nm. In some cases (e.g., PAAEs depicted in Table 1, Entries 6 and 7), emission maxima were red-shifted to ~402–410 nm and 424–425 nm. While no significant trends were observed with respect to incorporation of electronically-donating or -withdrawing groups, other factors such as molecular weights, dispersity, and/or substitution pattern on the comonomers are likely to have subtle effects on the optical behaviors observed.

In conclusion, we have developed a new and efficient method of polymerization inspired by alkene diarylation. The resulting polymers, PAAEs, exhibit molecular weights ranging from 32–175 kDa, and concomitant dispersities ranging from 1.48–2.30. This method is tolerant to a wide range of functional groups, thereby imparting tunability and introducing the potential for facile post-polymerization modification. We have observed these materials to exhibit thermal properties reminiscent of poly(styrene) and poly(phenylene methylene), which both bear structural resemblance to our polymer. Similarly, these polymers are observed to fluoresce in the presence of UV irradiation, similar to poly(phenylene methylene)s.

With the demonstration of the proof-of-concept for this novel polymerization methodology with aryl boronic esters and styrenes that forge a para-substituted sequence, we anticipate that further expansion to other backbone sequences (e.g., featuring ortho connectivity, or a combination of ortho and meta) might exhibit higher ordered self-assemblies and desirable emergent properties. Moreover, our method thus, might be leveraged as a modular means to achieve a diverse array of polymers from styrenics that might not polymerize normally through radically-mediated polymerizations.