Reversing Blocking Order of Trithiocarbonate‐Mediated RAFT Polymerizations Using Photocatalysis

Abstract

Many acrylic–methacrylic block copolymer sequences remain inaccessible due to synthetic limitations. Herein, photoinduced electron/energy transfer (PET) catalysis is leveraged to reverse blocking order limitations in trithiocarbonate (TTC)‐mediated reversible addition–fragmentation chain transfer (RAFT) polymerization. We synthesized poly(methyl acrylate‐b‐methyl methacrylate) by PET‐RAFT using fac‐Ir(ppy)₃, achieving predictable, linear increases in molecular weight with conversion. Kinetics studies showed that adding a tertiary amine (triethanolamine) introduced a reversible redox reaction to stabilize the TTC radical during chain extensions, leading to more uniform block copolymers (Р< 1.47) compared to block copolymers synthesized without amine (Р< 1.56). To highlight the utility of this method, triblock copolymers of poly(methyl acrylate) and poly(methyl methacrylate) blocks were investigated. The order of acrylic and methacrylic blocks impacted the physical properties of compositionally similar polymeric materials. For example, a high molecular weight triblock copolymer (P(MMA‐b‐MA‐b‐MMA), M _n = 564 kg mol⁻¹) thermoplastic elastomer showed exceptional strain (>1600%). Overall, we report (i) a new methodology to unlock synthetic access to acrylic–methacrylic block copolymers using TTCs and photocatalysis, (ii) insight into photocatalyst‐mediated radical polymerization, and (iii) synthesis of new high‐performance materials.

Acrylic‐methacrylic block copolymers have limited accessibility due to monomer order requirements in reversible‐deactivation radical polymerization techniques leading to a vast landscape of block copolymers being understudied. Using unique PET‐RAFT polymerization conditions, a controlled chain extension of poly(methyl acrylate) with methyl methacrylate could be performed resulting in a high‐molecular weight thermoplastic elastomer.

Acrylic block copolymers find wide use in lithography,^{[ 1} , ² , ^{3 ]} drug delivery,^{[ 4} , ⁵ , ^{6 ]} adhesives,^{[ 7} , ^{8 ]} and thermoplastic elastomers (TPE).^{[ 9} , ¹⁰ , ¹¹ , ¹² , ¹³ , ^{14 ]} Synthetic routes toward block copolymers include reversible deactivation radical polymerization (RDRP),^{[ 15} , ¹⁶ , ¹⁷ , ^{18 ]} anionic polymerization,^{[ 19} , ²⁰ , ^{21 ]} cationic polymerization,^{[ 22} , ²³ , ^{24 ]} and Lewis pair polymerization.^{[ 25} , ^{26 ]} However, these techniques do not enable access to many acrylic–methacrylic block copolymers as they have limitations such as blocking order requirements, minimal chain‐end fidelity, or limited monomer scope.

Blocking order requirements are dictated by the relative radical stabilities of active chain ends and restrict RDRP techniques.^{[ 27} , ^{28 ]} More stable monomers (e.g., methacrylates—3° radicals) must be polymerized before less stable monomers (e.g., acrylates—2° radicals) to avoid incomplete reinitiation. This limitation in RDRP leads to high fractions of homopolymer impurities (Figure 1a), limiting access to many discrete AB diblock and subsequent ABA/ABC triblock copolymer sequences, which are promising for polymer compatibilization,^{[ 29} , ³⁰ , ^{31 ]} biomimicry,^{[ 32} , ^{33 ]} and high‐strength materials.^{[ 34} , ^{35 ]}

Figure 1.

a) Initiation methods and subsequent polymer distributions when polymerizing against conventional blocking order with thermal, photoiniferter, or PET initiation methods. b) Reaction coordinate diagram depicting the chain transfer mechanism with an MA‐ or MMA‐capped chain end. c) DFT calculations of energy required to undergo homolytic cleavage of MA‐ and MMA‐capped chains and the amount of photothermal energy fac‐Ir(ppy)₃ can provide.^{[ 63 ].}

Few approaches exist to reverse RDRP blocking limitations. A halogen exchange reaction or photocatalysis can be used in atom transfer radical polymerization (ATRP) to chain‐extend polyacrylates with methacrylates.^{[ 36} , ^{37 ]} However, obtaining the correct combination of halogen, catalyst, ligand, and monomers for acrylic–methacrylic block copolymers can be resource‐intensive and limit access to multiblock copolymers.^{[ 38} , ³⁹ , ⁴⁰ , ^{41 ]} Additionally, achieving high molecular weight (HMW) polymers (>150 kg mol⁻¹) using ATRP requires high pressure reaction conditions, heterogeneous systems, or dual catalysis, and halogen‐exchange on HMW polymers from ATRP is not reported.^{[ 42} , ⁴³ , ⁴⁴ , ⁴⁵ , ⁴⁶ , ⁴⁷ , ^{48 ]} In reversible addition‐fragmentation chain transfer (RAFT) polymerization, ultra‐high molecular weight polymers can be readily synthesized.^{[ 48} , ^{49 ]} Direct photolysis of xanthate RAFT agents can extend polyacrylates with methacrylates.^{[ 50 ]} However, controlling molecular weights during the chain extension is challenging, only works with xanthates due to the lower carbon─sulfur bond dissociation energy (C─S BDE = 38.06 kcal mol⁻¹) compared to more commonly used trithiocarbonate RAFT agents (TTC, C‐S BDE = 41.50 kcal mol⁻¹, Figure 1a),^{[ 51 ]} and has not been reported for polymers >40 kg mol⁻¹. The limited accessibility of HMW acrylic–methacrylic block copolymers limits the ability to probe copolymer properties and applications.

Monomer‐order limitations can be bypassed during a TTC‐mediated RAFT polymerization by using minimal equivalents of a comonomer, but this yields a non‐discrete, gradient block copolymer.^{[ 52 ]} These impurities affect applications (e.g., lithography) that rely on distinct separation of polymer phases.^{[ 53} , ^{54 ]} Overall, synthetic methods to access discrete acrylic–methacrylic block copolymers using RDRP are limited. We set out to develop a method to overcome monomer restrictions with TTC‐mediated RDRP that would enable a broader range of accessible block copolymers.

Chain transfer makes reversing blocking limitations with TTCs difficult. For example, during a chain extension of poly(methyl acrylate) (PMA, macroinitiator) with methyl methacrylate (MMA), the intermediate radical species (Figure 1b) must fragment into 2 to initiate PMA chains for chain extension. However, the energy barrier to 2 (ΔG ^‡ = +16.8 kcal mol⁻¹) exceeds fragmentation into 3 (ΔG^‡ = +11.6 kcal mol⁻¹). Fragmentation into 3 leaves PMA chains un‐activated and results in free‐radical polymerization of MMA.^{[ 27 ]} To address this limitation, we expected that photoinduced electron/energy transfer (PET) catalysis using fac‐Ir(ppy)₃ will overcome blocking‐order requirements in TTC‐mediated RAFT polymerization and minimize differences in chain‐end BDE (Figure 1a). PET‐RAFT polymerization, first reported by the Boyer group,^{[ 55} , ⁵⁶ , ^{57 ]} can be performed with various photocatalysts across the visible light spectrum. Compared to photoiniferter, direct photolysis of the C─S bond on the R‐group of a RAFT polymerization chain transfer agent (CTA),^{[ 58 ]} PET‐RAFT polymerization uses a photocatalyst to cleave the C─S bond of the R‐group on the CTA, leading to a wider variety of wavelengths and initiation pathways (e.g., energy transfer, oxidative electron transfer, reductive electron transfer).^{[ 58} , ^{59 ]}

Our model density functional theory (DFT) calculations of MA and MMA chains capped with a TTC predict bond dissociation energies (BDE) of 37.7  and 31.3 kcal mol⁻¹, respectively (Figure 1c). The photocatalyst, fac‐Ir(ppy)₃,^{[ 55} , ^{60 ]} liberates 56.2 kcal mol⁻¹ of photothermal energy following excitation;^{[ 61 ]} thus, fac‐Ir(ppy)₃ could readily achieve direct C─S homolysis of both TTC‐capped species. We hypothesized that fac‐Ir(ppy)₃, which goes through an energy transfer pathway,^{[ 59} , ⁶² , ^{63 ]} would enable us to reverse conventional blocking order limitations with TTCs. In sum, we envisioned that the blocking order limitations that have plagued TTC‐mediated RDRP could be overcome using PET catalysis, enabling synthetic access to previously inaccessible and potentially useful (multi)block copolymers.

A thermally‐initiated RAFT polymerization yielded the first block of PMA (13.3 kg mol⁻¹, Р= 1.06). Next, we compared chain extension polymerizations with MMA under thermally‐initiated RAFT (purple traces), photoiniferter (pink traces), or PET‐RAFT (blue traces) polymerization conditions using size‐exclusion chromatography (SEC, Figure 2a–c). We chose polymerization conditions according to common (PET‐)RAFT polymerizations (Figure 2a, 300 equiv MMA, 5 mM [PMA], 0.1 equiv 4,4′‐azobis(4‐cyanovaleric acid) or 0.005 equiv fac‐Ir(ppy)₃ when applicable). The light‐mediated polymerizations at 25 °C yielded broad SEC traces (Figure S1); thus, we increased the temperature to 40 °C to increase the rate of activation by PET,^{[ 60 ]} leading to faster activation of chain‐ends and more uniform polymer chains, which resulted in narrower SEC traces than at 25 °C. Both thermally‐initiated RAFT and photoiniferter polymerizations yielded bimodal chromatograms. These bimodal traces, characteristic of TTC‐mediated polymerizations going against monomer blocking‐order restrictions, suggested populations of unreacted PMA, the desired P(MA‐b‐MMA) copolymers, and uncontrolled PMMA.^{[ 50 ]} The PET‐RAFT polymerization showed a monomodal shift to lower retention times with a corresponding shift in the UV–vis absorbance of the TTC (Figure S2). These results suggested that the distinct initiation mechanism of PET‐RAFT polymerization can be used to overcome monomer restrictions in TTC‐mediated polymerizations. However, the PET‐RAFT trace was broad (Р= 1.33) and did not align with expected number‐average molecular weights (M _n, M _n,theory. = 33.4 kg mol⁻¹ versus M _n,SEC = 18.4 kg mol⁻¹)

Figure 2.

Refractive index traces of PMA chain extensions with MMA using thermal, photoiniferter, or PET initiation at 5 mM a) or 2 mM b) [PMA]. c) UV–vis traces of P(MA‐b‐MMA) following chain extension at 2 mM [PMA]. Cartoon representations of bimolecular chain end activation d), bimolecular termination e), unimolecular chain end activation f), and unimolecular termination g).

To investigate how to improve molecular weight agreement, we investigated the effect of macroinitiator concentration. We expected that unimolecular and bimolecular processes affect interchain proximity and radical concentration, which impact the activation–deactivation, chain transfer, and termination reactions (Figure 2d–g).^{[ 64 ]} As termination occurs bimolecularly in RAFT polymerizations,^{[ 64 ]} we expected that decreased [PMA] would decrease the amount of termination. Decreasing the concentration of PMA in solution results in a decrease in the concentration of radical species. We expected that by decreasing [radical] by decreasing [PMA], there would be less termination. To measure the effect of [PMA], we performed polymerizations at 2 mM [PMA] (Figure 2b, 250 equiv MMA, 0.1 equiv 4,4′‐azobis(4‐cyanovaleric acid) or 0.01 equiv fac‐Ir(ppy)₃ when applicable). We compared these conditions to conventional PET‐RAFT conditions (where [PMA] = 5 mM, Figure 2a). As expected, we observed bimodal chromatograms for thermal and photoiniferter while the PET‐RAFT polymerization conditions showed a narrower chromatogram compared to when [PMA] = 5 mM.

These data suggest that higher [PMA] favors unproductive bimolecular chain transfer reactions (Figure 2d) and termination (Figure 2e) that led to higher Ð values (1.33), while decreased [PMA] led to lower Ð values (1.20) determined by multi‐angle light scattering. Following PET activation (Figure 2f), we suspect that the TTC radical degrades, contributing to termination by loss of the deactivating species (Figure 2g), as the sulfur‐centered TTC radicals resulting from PET activation can degrade or initiate new chains.^{[ 65 ]} To counteract this we introduced a competing TTC stabilization pathway.

Tertiary amines can reduce a TTC radical to a TTC anion, preventing degradation or initiation of new chains.^{[ 66} , ⁶⁷ , ⁶⁸ , ⁶⁹ , ⁷⁰ , ^{71 ]} To stabilize the TTC radical via a tertiary amine mediated reversible redox pathway, triethanolamine (TEOA) was added to the system. Experiments comparing [PMA] and fac‐Ir(ppy)₃ loading (Figures S3–S8) in the presence of TEOA showed that [PMA] = 2 mM and 0.01 equiv of fac‐Ir(ppy)₃ yielded narrow and symmetric SEC traces with increasing molar masses and low Ð values. However, the UV–vis absorbance traces from the PET‐RAFT polymerizations with TEOA showed slightly higher absorbance values compared to no TEOA (Figure S2).

We investigated polymerization kinetics to further evaluate the impact of TEOA on PET‐RAFT polymerizations. Reaction aliquots were taken during a 6 h chain extension polymerization of PMA with MMA with 0 or 1 equiv of TEOA. Without TEOA, the pseudo‐first‐order kinetics plot showed a linear relationship up to 4 h. After 4 h, we observed a minor decrease in the apparent rate, indicating a decrease in radical concentration likely due to TTC degradation/termination (Figure 3a). The M _n versus conversion plot agreed with the expected molar masses measured by ¹H nuclear magnetic resonance (NMR) spectroscopy monomer conversion and the experimental molecular weights measured by SEC using PMMA standards, but Ð increased with conversion (MMA conversion = 38 ± 2% and Р= 1.56 at 6 h, Figure 3b). The SEC refractive index (RI) traces shifted to lower elution times with increasing monomer conversion, showing a minor shoulder throughout (Figure 3c and S9). Similar kinetics were observed with the addition of 1 equiv of TEOA (Figure 3d–f and S10). However, lower Ð values persisted throughout the polymerization (MMA conversion = 39 ± 2% and Р= 1.47 at 6 h). Deconvolution of the 6 h RI traces showed that polymerizations with TEOA increased the number of extended chains from 86% to 91% (Figure S11 and S12). We observed similar kinetic data with poly(N,N‐dimethylacrylamide) (PDMA) as the macroinitiator and the difference in the starting polymer seemed to only minorly affect the efficiency of the chain extensions (Figure S13). Overall, kinetics indicated a controlled MMA chain‐extension polymerization from polyacrylates and polyacrylamides by using fac‐Ir(ppy)₃ combined with TEOA showing linear pseudo‐first‐order kinetics, linearly increasing M _n with conversion, and moderate Ð.

Figure 3.

Pseudo‐first‐order kinetics plot with 0 a) or 1 (d) equiv TEOA. M_n and Ð plot with 0 b) or 1 (e) equiv TEOA versus PMMA standards. RI traces of polymerization progress with 0 c) or 1 (f) equiv TEOA. g) UV–vis absorbance versus time with 0 and 1 equiv TEOA. h) DFT calculations of accessible routes for the TTC radical species presented in an energy coordinate diagram.

To further compare if TEOA stabilized TTCs during the PET‐RAFT polymerizations, absorbance at 365 nm in‐line with SEC was used to quantify TTC throughout the polymerization. When TEOA was present, up to a 12% improvement in TTC retention was observed compared to reactions without TEOA during the first 3 h (Figure 3g). We expected that TEOA undergoes a reversible redox reaction with the TTC radical to form the more stable TTC anion and an amine radical cation.^{[ 66} , ⁶⁷ , ⁶⁸ , ⁶⁹ , ⁷⁰ , ^{71 ]} However, a continuous decline in UV–vis absorbance occurred throughout the chain extension. The decline in TCT UV–vis absorbance indicated termination events, which are substantiated by the increase in Ð values. We investigated if additional TEOA could retain more chain‐ends in the reaction. However, using 2 or 5 equiv of TEOA resulted in decreased UV–vis absorbance compared to 1 equiv of TEOA (Figures S14–S17). These results show increasing TEOA does not improve or prolong chain‐end retention.

DFT calculations were performed to investigate the energetics for the reversible redox reaction introduced by the amine (Figure 3h). These calculations indicated that electron transfer between the TTC radical and the amine is nearly thermoneutral (ΔG = −0.6 kcal mol⁻¹) via an accessible [TTC‐amine] radical complex (ΔG = +3.2 kcal mol⁻¹). We predicted that degradation of the TTC anion via decarboxylation (ΔG ^‡ = +24.9 kcal mol⁻¹, ΔG = +12.1 kcal mol⁻¹) is much less likely than decarboxylation from the radical form (ΔG ^‡ = +19.0 kcal mol⁻¹, ΔG = −3.1 kcal mol⁻¹). Interestingly, the most likely predicted control‐limiting mechanistic pathway for the TTC radical is addition to MMA, which showed an energy barrier like MMA homopropagation (ΔG ^‡ = +14.6 kcal mol⁻¹). However, we expect this pathway to be minimized in the presence of TEOA. These data suggest that TEOA plays a minor role in limiting the degradation of the deactivating species (TTC) during the polymerization by introducing a reversible redox pathway to stabilize the TTC radical. Future studies into more reducing tertiary (multi)amines or other reducing agents capable of performing a TTC radical reduction redox reaction could further improve chain‐end retention of TTC during the chain extension.

To delineate the effects of discrete block sequence order on physical properties, we synthesized P(MMA‐b‐MA‐b‐MMA) and P(MA‐b‐MMA‐b‐MA) using a bis‐TTC to achieve bidirectional growth of block copolymers (Figure 4a). These block sequences impacted physical properties with the triblock copolymer following conventional monomer blocking order (P(MA‐b‐MMA‐b‐MA)) exhibiting brittle behavior and 8.6 MPa stress and 3.6% strain at break (Figure S18), while the triblock copolymer made by the reported PET conditions (P(MMA‐b‐MA‐b‐MMA)) exhibited elastomeric behavior and 19 MPa stress and 33% strain at break (Figure S19). The change in block sequence (P(MMA‐b‐MA‐b‐MMA) versus P(MA‐b‐MMA‐b‐MA)) resulted in over a 200% greater stress and 900% greater strain at break. These data confirm that the block sequences accessible using PET‐RAFT polymerization directly impact physical properties.

Figure 4.

a) Copolymer sequences synthesized with representative cartoons and stress versus strain curves. b) HMW block copolymer synthesized by bifunctional growth, stress versus strain curve, and images depicting the sample before, during, and after being pulled to 1001% strain.

Finally, because TTC‐mediated polymerizations provide unprecedented access to HMW acrylic polymers, we synthesized a HMW P(MMA‐b‐MA‐b‐MMA) block copolymer using a bis‐TTC (M _n = 564 kg mol⁻¹, Figure 4b). The HMW TPE did not break up to 1600% strain (1.6 MPa stress) when subjected to tensile testing, but rather thinned and slipped out of the clamps (Figure S20). To investigate if the sample had shape recovery, we subjected it to further tensile testing. The sample (starting length = 26 mm) was pulled to 1001% strain and allowed 10 min to recover. The HMW TPE fully recovered to a final length of 27 mm while remaining transparent. These results suggested this PET‐RAFT polymerization method can be used to produce a variety of acrylic/methacrylic “superelastomeric” materials.^{[ 11} , ^{72 ]}

In this study, we found that PET catalysis can reverse conventional blocking order requirements in TTC‐mediated RAFT polymerizations and provide access to HMW acrylic–methacrylic block copolymers. Using greater catalyst loadings (0.01 equiv) and lower macroinitiator concentrations (2 mM) than typical for PET‐RAFT polymerizations showed a controlled polymerization. TEOA stabilized the TTC radical through an electron transfer, corroborated by DFT calculations, and enhanced chain‐end retention compared to polymerizations without TEOA. These advances enabled access to novel block copolymer sequences with significantly different physical properties according to the block sequence, exemplified by a HMW superelastomeric P(MMA‐b‐MA‐b‐MMA) TPE. Overall, this work provides critical insight into how photocatalysis can be used to access novel high‐performance polymeric materials.

Supporting Information

The authors have cited additional references within the Supporting Information.^[73–77]

Conflict of Interests

The authors declare no conflict of interest.

Supporting information

Supporting Information

Baker J. G., Koehler S. J., Wood K. J., Troya D., Gloriod J., Anderson I. C., Gomez D. C., Figg C. A., Angew. Chem. Int. Ed.. 2025, 64, e202509029. 10.1002/anie.202509029

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.