Fate of the RAFT End-Group in the Thermal Depolymerization of Polymethacrylates

Thermal RAFT depolymerization has recently emerged as a promising methodology for the chemical recycling of polymers. However, while much attention has been given to the regeneration of monomers, the fate of the RAFT end-group after depolymerization has been unexplored. Herein, we identify the dominant small molecules derived from the RAFT end-group of polymethacrylates. The major product was found to be a unimer (DP = 1) RAFT agent, which is not only challenging to synthesize using conventional single-unit monomer insertion strategies, but also a highly active RAFT agent for methyl methacrylate, exhibiting faster consumption and yielding polymers with lower dispersities compared to the original, commercially available 2-cyano-2-propyl dithiobenzoate. Solvent-derived molecules were also identified predominantly at the beginning of the depolymerization, thus suggesting a significant mechanistic contribution from the solvent. Notably, the formation of both the unimer and the solvent-derived products remained consistent regardless of the RAFT agent, monomer, or solvent employed.


Materials
All materials were purchased from either Sigma Aldrich or Fischer Scientific unless otherwise stated. Benzyl methacrylate (BzMA, >98.0%) was purchased from Tokyo Chemical Industries. Monomers were filtered through basic alumina before use.

Nuclear Magnetic Resonance Spectroscopy (NMR)
1 H NMR spectra were recorded on a Bruker Avance-300 and 500 MHz spectrometer using acetone-d 6 or CDCl 3 as the NMR solvent. Chemical shifts are given in ppm, downfield from tetramethylsilane (TMS) and referenced to residual solvent proton signals.

Size-Exclusion Chromatography (SEC)
SEC was measured on a Shimadzu equipment comprising a CBM-20A system controller, LC-20AD pump, SIL-20A automatic injector, 10.0 μm bead-size guard column (50 x 7.5 mm) followed by three KF-805-L columns (300 x 8 mm, bead size: 10 μm, pore size maximum: 5000 Å), SPD-20A ultraviolet detector, and RID-20A differential refractive index detector. The column temperature was maintained at 40 °C using a CTO-20A oven. The flow rate was set to 1 ml/min and with N,Ndimethylacetamide (DMAc, Acros, HPLC grade, with 0.03 w/v LiBr) as the eluent. Molecular weights were determined relative to poly(methyl methacrylate) standards wieth molecular weights ranging from 5000 to 1.5 x 10 6 g/mol (Agilent Technologies). All SEC samples were dissolved in DMAc and passed through 0.45 μm filters prior to analysis.

Mass Spectrometry (MS)
Mass spectrometry (MS) experiments were either performed on a Bruker Maxis I quadrupole-time-of-flight (QTOF) at 4500V emitter voltage with electrospray ionization (ESI) as ion source and direct injection or on a Thermo Scientific QExactive GC Orbitrap (GC-MS) with Ion-Trap (OrbiTrap) and electron impact (EI) as ion source.

Polymerization of MMA with DTB
Into a 25 ml round bottom flask, 448 mg of 2-cyanoprop-2-yl dithiobenzoate (DTB, 1.2 mmol, 1 equiv) were dissolved in 4 ml acetonitrile (ACN). A stock solution of AIBN (40 mg) was prepared in 2 ml ACN, and 1640 μL of this solution (32.8 mg, 199 μmol, 0.1 equiv) was transferred to the flask. Subsequently, 10.6 mL of MMA (10.0 g, 99.9 mmol, 50 equiv) and a stirrer bar were added, and the flask was sealed with a septum, prior to deoxygenation by nitrogen bubbling for 15 min. Polymerization was conducted in an oil bath at 70 °C for 4h with a 400-rpm stirring rates. Samples were taken every hour under a nitrogen blanket for 1 H-NMR analysis and passed through a syringe filter (0.45 μm PTFE membrane) prior to SEC analysis. Polymerization was stopped at 74 % conversion by removing the reaction from the oil bath and removing the septum. Monomer conversions were determined by NMR spectroscopy. The monomer vinyl signals were compared to the combined polymer and monomer ester signals.

Polymerization of MMA with TTC
Into a 25 ml round bottom flask, 690 mg of 2-cyanoprop-2-yl dodecyl trithiocarbonate (TTC, 1.99 mmol, 1 equiv) were dissolved in 4 ml acetonitrile (ACN). A stock solution of AIBN (40 mg) was prepared in 2 ml ACN, and 1640 μL of this solution (32.8 mg, 199 μmol, 0.1 equiv) was transferred to the flask. Subsequently, 10.6 mL of MMA (10.0 g, 99.9 mmol, 50 equiv) and a stirrer bar were added, and the flask was sealed with a septum, prior to deoxygenation by nitrogen bubbling for 15 min. Polymerization was conducted in an oil bath at 70 °C for 4h with a 400-rpm stirring rates. Samples were taken every hour under a nitrogen blanket for 1 H-NMR analysis and passed through a syringe filter (0.45 μm PTFE membrane) prior to SEC analysis. Polymerization was stopped at 54 % conversion by removing the reaction from the oil bath and removing the septum. Monomer conversions were determined by NMR spectroscopy. The monomer vinyl signals were compared to the combined polymer and monomer ester signals.

Polymerization of BzMA with DTB
Into a 25 ml round bottom flask, 255 mg of 2-cyanoprop-2-yl dithiobenzoate (DTB, 1.14 mmol, 1 equiv) were dissolved in 4 ml acetonitrile (ACN). A stock solution of AIBN (20 mg) was prepared in 1 ml ACN, and 932 μL of this solution (18.64 mg, 114 μmol, 0.1 equiv) was transferred to the flask. Subsequently, 10.6 mL of BzMA (10.6 g, 59.0 mmol, 52 equiv) and a stirrer bar were added, and the flask was sealed with a septum, prior to deoxygenation by nitrogen bubbling for 15 min. Polymerization was conducted in an oil bath at 70 °C for 4h with a 400-rpm stirring rates. Samples were taken every hour under a nitrogen blanket for 1 H-NMR analysis and passed through a syringe filter (0.45 μm PTFE membrane) prior to SEC analysis. Polymerization was stopped at 70 % conversion by removing the reaction from the oil bath and removing the septum. Monomer conversions were determined by NMR spectroscopy. The monomer vinyl signals were compared to the combined polymer and monomer methylene signal from benzyl group.

Purification of PMMA, PBzMA
Polymers were precipitated three times in a 3:2 mixture cold methanol:hexane and vacuum-filtered using a Buchner funnel. The precipitates were dried in a vacuum oven for at least 12h before use.

Depolymerization Procedure for PMMA and PBzMA
In a 125 ml Schlenk tube, 21.5 mg of PMMA was dissolved in 40 ml 1,4-dioxane (5.17 mM of MMA repeat unit). The Schlenk tube was sealed with a rubber septum and deoxygenated by nitrogen bubbling for 20 min. The Schlenk tube was then put into a 120 °C oil bath to start the reaction. The Schlenk tube was submerged into the oil bath until the surface of the solution inside was at the same height as the oil bath. To take samples, the reaction was periodically removed from the oil bath and quickly added to a cold water bath until the solution cooled to room temperature. The solution was then sampled under a nitrogen blanket. For SEC samples, 800 μL of the sample solution was blow-dried, dissolved in DMAc and passed through a syringe filter (0.45 μm PTFE membrane).

Polymerization Procedure for MMA using 1 as CTA
Into a 10 ml test tube, 5.95 mg of molecule 1 (18.5 μmol, 1 equiv) was added. A stock solution of AIBN (1.314 mg) was prepared in 4 ml ACN, and 925 μL of this solution (0.304 μg, 1.85 μmol, 0.1 equiv) was transferred to the tube. Subsequently, 0.986 mL of MMA (927 mg, 9.26 mmol, 500 equiv) and a stirrer bar were added, and the tube was sealed with a septum, prior to deoxygenation by nitrogen bubbling for 15 min.
Repolymerization was conducted in an oil bath at 70 °C for 8h with a 400-rpm stirring rates. Samples were taken every hour under a nitrogen blanket for 1 H-NMR analysis and passed through a syringe filter (0.45 μm PTFE membrane) prior to SEC analysis. Repolymerization was left until 22h and stopped at 60 % conversion by removing the reaction from the oil bath and removing the septum. Monomer conversions were determined by NMR spectroscopy. The monomer vinyl signals were compared to the combined polymer and monomer ester signals.

Polymerization Procedure for MMA using 2 as CTA
Into a 10 ml test tube, 2.2 mg of molecule 2 (9 μmol, 1 equiv) was added. A stock solution of AIBN (1.314 mg) was prepared in 4 ml ACN, and 458 μL of this solution (0.150 μg, 0.92 μmol, 0.1 equiv) was transferred to the tube. Subsequently, 0.487 mL of MMA (458 mg, 4.58 mmol, 500 equiv) and a stirrer bar were added, and the tube was sealed with a septum, prior to deoxygenation by nitrogen bubbling for 15 min. Repolymerization was conducted in an oil bath at 70 °C for 8h with a 400-rpm stirring rates. Samples were taken every hour under a nitrogen blanket for 1 H-NMR analysis and passed through a syringe filter (0.45 μm PTFE membrane) prior to SEC analysis. Repolymerization was left until 22h and stopped at 27 % conversion by removing the reaction from the oil bath and removing the septum. Increasing viscosity due to the large molecular weight led to unreliable data and thus data points after 6h were disregarded for further analysis. Monomer conversions were determined by NMR spectroscopy. The monomer vinyl signals were compared to the combined polymer and monomer ester signals.

Polymerization Procedure for MMA using 3 as CTA
Into a 10 ml test tube, 4.9 mg of molecule 3 (19 μmol, 1 equiv) was added. A stock solution of AIBN (1.314 mg) was prepared in 4 ml ACN, and 955 μL of this solution (0.314 μg, 1.91 μmol, 0.1 equiv) was transferred to the tube. Subsequently, 1.018 mL of MMA (957 mg, 9.56 mmol, 500 equiv) and a stirrer bar were added, and the tube was sealed with a septum, prior to deoxygenation by nitrogen bubbling for 15 min. Repolymerization was conducted in an oil bath at 70 °C for 8h with a 400-rpm stirring rates. Samples were taken every hour under a nitrogen blanket for 1 H-NMR analysis and passed through a syringe filter (0.45 μm PTFE membrane) prior to SEC analysis. Repolymerization was left until 22h and stopped at 57 % conversion by removing the reaction from the oil bath and removing the septum. Increasing viscosity due to the large molecular weight led to unreliable data and thus data points after 8h were disregarded for further analysis. Monomer conversions were determined by NMR spectroscopy. The monomer vinyl signals were compared to the combined polymer and monomer ester signals.

Determination of depolymerization conversion
Depolymerization conversions were determined in-situ by comparing the monomer vinyl signals to the polymer backbone -CH3 signals (simply taking a sample in dioxane and re-dissolving it in the deuterated solvent, either d 6 -acetone or CDCl 3 ). To ensure accurate conversion calculation a second sample was prepared for SEC measurement. The polymer signal (RI detector) at sampling time (t=x) was compared to the signal at start (t=0h) and thus the depolymerization conversion was calculated. Specifically, an exact same sampling volume was required for the SEC measurement (see 1.9). Depolymerization conversions from the two methods deviated by < 10%.

Determination of small molecule fraction ratio
Small molecule fraction ratio was determined by 1 H-NMR spectroscopy. Samples were taken under a nitrogen blanket and then blow-dried. The sample was re-dissolving it in the deuterated solvent (either d 6 -acetone or CDCl 3 ). The ratio was determined by comparing the unobstructed, characteristic NMR peaks from the small molecules. For molecule 1 this corresponds to signal 10a (2.56 ppm) and/or 10b (2.44 ppm) with 1H each. For molecule 2 this corresponds to signal 6 (6.12 ppm) with 1H. For molecule 3 this corresponds to signal 7 (5.61 ppm) with 2H.