Regioselective Palladium-Catalyzed Chain-Growth Allylic Amination Polymerization of Vinyl Aziridines

Abstract

Organometallic-mediated chain growth polymerization of readily accessible chemical building blocks is responsible for important commercial and technological advances in polymer science, but the incorporation of heteroatoms into the polymer backbone through these mechanisms remains a challenge. Transition metal π-allyl complexes are well-developed organometallic intermediates for carbon-heteroatom bond formation in small-molecule catalysis yet remain underexplored in polymer science. Here, we developed a regioselective palladium-phosphoramidite-catalyzed chain-growth allylic amination polymerization of vinyl aziridines for the synthesis of novel nitrogen-rich polymers via ambiphilic π-allyl complexes. The polymerization accessed a linear microstructure with four carbons between each nitrogen, which is challenging to achieve through other chain-growth polymerization approaches. The highly regioselective allylic amination polymerization demonstrated the characteristics of a controlled polymerization and was able to achieve molar masses exceeding 20 kg mol^-1 with low dispersities (D̵ < 1.3). The identification of the polymer structure and well-defined chain ends were supported by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry, and chain extension experiments demonstrate opportunities for building more complex materials from this method. A Hammett study was performed to understand the role of the catalyst and monomer structure on regioselectivity, and the data supported a mechanism wherein regioselectivity was primarily controlled by the ligand-metal complex. Postpolymerization desulfonylation provided access to a novel polyamine that demonstrated broad anticancer activity in vitro, which highlights the benefits of unlocking novel polyamine microstructures through regioselective chain-growth allylic amination polymerization.

Figure 1.

(A) Examples of organometallic intermediates leveraged for polymer synthesis; (B) previously reported chain-growth methods for the synthesis of polyamines; (C) regioselective chain-growth allylic amination polymerization of protected vinyl aziridines developed herein.

Figure 2.

(A) Results of targeting between 5 and 40 kg mol⁻¹ using optimized conditions; number-average molar mass (M_n) was determined using SEC in THF at 1 mg/mL using a refractive index detector and is reported in relation to a polystyrene calibration curve; (B) initiator-free polymerization of 1a catalyzed by Pd-L9; (C) chain-extension experiment with poly(1a·4.4k).

Figure 3.

(A) MALDI-TOF MS of poly(1a·4.4k) highlighting the presence of the initiator; (B) repeat unit observed with poly(1a)s; (C) MALDI-TOF MS of poly(1a·6.4k), poly(1a·10.2k), and poly(1a·13.0k).

Figure 4.

(A) Exploration of monomer electronics on regioselectivity; regioselectivity was determined using ¹H NMR spectroscopy; glass transition temperature (T_g) reported from the second heating cycle at a rate of 10 °C/min; M_n was determined using SEC in DMF with 0.1% LiBr at 1 mg/mL using a refractive index detector and is reported in relation to a poly(ethylene oxide) calibration curve; (B) Hammett plot of log(L:B) vs σ_p constants and rationale for increased regioselectivity with electron-withdrawing groups.

Figure 5.

Proposed catalytic cycle of chain-growth allylic amination polymerization via π–allyl complexes both with and without an initiator.

Figure 6.

(A) Synthesis and thiol-mediated desulfonylation of poly(1e·7.2k) to produce poly(3·1.7k); (B) cell viability assays and live/dead cell imaging with poly(3·1.7k) and cisplatin. For cell viability studies, cells were incubated with cisplatin or poly(3·1.7k) in media at pH = 7.4 for 24 h, followed by addition of Alamar Blue reagent prior to analysis. For live/dead imaging, cells were incubated in media at pH = 7.4 at the IC₅₀ concentration of poly(3·1.7k) for 24 h prior to the addition of live/dead imaging reagent and imaged using a confocal laser scanning microscope. Scale bars: 50 μm.

Scheme 1.. Ligands Explored for the Polymerization of Aziridine 1a Initiated by 2a^a

^aNumber average molar mass (M_n) was determined using size exclusion chromatography (SEC) in THF at 1 mg/mL using a refractive index detector and is reported in relation to a polystyrene calibration curve. Regioselectivity was determined using ¹H NMR spectroscopy based on the ratio of integration values of resonances from 5.2 to 5.5 ppm (linear) and 4.4–4.6 ppm (branched) after accounting for relative abundance in the respective repeat units. ND = not determined.

Scheme 2.. Solvent Screen with 1a and 2a Catalyzed by Pd-L9^a

^aDetermined using ¹H NMR Spectroscopy. Number-average molar mass (M_n) was determined using SEC in THF at 1 mg/mL using a refractive index detector and is reported in relation to a polystyrene calibration curve.