A renewably sourced, circular photopolymer resin for additive manufacturing

Vat photopolymerization enables rapid fabrication of complex 3D parts but uses mostly petroleum-derived acrylate/epoxy resins that cure into crosslinked thermosets which are difficult to recycle. Existing dynamic or recyclable photopolymer systems typically require adding new reactive components upon recycling (open-loop), altering composition and creating a snowballing effect. The authors hypothesize that using strained cyclic disulfides (lipoates) as the photoactive units can provide a resin where the dynamic bond is formed during photopolymerization and later depolymerized back to the original monomers, enabling genuine closed-loop 3D printing with renewably sourced components. The goal is to create a high-resolution, modular resin platform derived from lipoic acid that prints well, can be efficiently deconstructed, and reprinted multiple times while maintaining composition and properties comparable to commercial resins.

Prior recyclable/vitrimeric photopolymer approaches incorporate dynamic covalent bonds (for example thiol-ene, thioester, vitrimer chemistries) but typically require addition of extra photoactive species during regeneration, leading to composition drift and a snowballing need for more material. Reversible cycloadditions and thiol-ene-based 2D photosets demonstrate some recyclability but face issues for high-fidelity vat photopolymerization, including complex monomer synthesis, requirements for high-energy/continuous irradiation, incomplete depolymerization, and property disparity after recycling. Strained cyclic disulfides (such as 1,2-dithiolanes from lipoic acid) are known to polymerize via radical methods and show dynamic behavior including depolymerization. Previous methods with internal dynamic bonds struggled with orthogonality between photopolymerization and dynamic exchange, limiting closed-loop recycling at scale. The authors build on work showing ring-opening polymerization and dynamic exchange of lipoates, proposing them as renewable, safer alternatives to (meth)acrylates for circular vat photopolymerization.

Monomer synthesis and resin formulation: Lipoic acid was esterified to produce a multivalent crosslinker (isosorbide dilipoate, IsoLp2) and a monofunctional reactive diluent (menthyl lipoate, MenLp1) using EDC coupling; alternative greener synthesis via bulk acid-catalyzed Fischer esterification was also demonstrated. Mixed lipoate formulations (e.g., MenLp1-IsoLp2 at 30:70 wt%) showed better storage stability than individual components. Additional lipoate diluents from renewable alcohols were synthesized (e.g., ethyl lipoate, EtLp1; guaiacyl lipoate, GuaLp1; stearyl lipoate, SteaLp1) and a three-armed crosslinker (glyceryl trilipoate, GlyLp3) to probe modularity.

Printing and curing: Resins were printed on a commercial DLP printer (405 nm). Printing resolution was assessed using a custom array of squares, walls, and bridges; complex benchmark parts (3DBenchy) were also printed. For rapid material screening, 2D photosets (~0.5 mm thick) were cured on glass slides with 1 wt% photoinitiator (ethyl (2,4,6-trimethylbenzoyl) phenylphosphinate). Recycled resins for circular-printing tests were supplemented with photoinitiator (1.5 wt% for first recycle, 2.5 wt% for second) to normalize curing kinetics.

Depolymerization pathways: (1) Base-catalyzed chemical depolymerization: pulverized prints were treated in MeTHF with 1 mol% (relative to disulfide) P1-t-Bu phosphazene and thiophenol (equimolar) at 80 °C for 3 h under N2, yielding depolymerized resin. (2) Catalyst-free thermal depolymerization: reflux in DMF for 2 h, giving generally high and consistent recovery. (3) Hydrolytic depolymerization: NaOH in H2O/DMF to cleave esters and regenerate original alcohols and lipoic acid for re-synthesis of virgin resin.

Characterization: Composition and purity by 1H NMR; molecular weight profiles by SEC (CHCl3 + 0.5% NEt3, PS standards); UV–vis spectroscopy; FTIR; mechanical testing (uniaxial tensile, UTS and Young’s modulus); DSC (Tg, Tm); TGA (Td at 5% weight loss); DMA (rubbery plateau, tan δ); photorheology (gel point under irradiation); z-depth cure profiling and x–y feature fidelity using optical profilometry and image analysis. Structure–property mapping was performed by varying diluent:crosslinker ratios (MenLp1-IsoLp2 from 90:10 to 10:90 wt%) and by substituting alternative lipoate monomers and crosslinkers (e.g., EtLp1, GuaLp1, SteaLp1; IsoLp2 vs GlyLp3). Circular printing was validated over two recycle cycles using EtLp1-GlyLp3 (31:69 wt%).

• Printing performance: MenLp1-IsoLp2 (30:70 wt%) printed on an off-the-shelf 405 nm DLP with high fidelity. Smallest reproducible wall: ~100 µm (≈3 pixels) at 25 s per 50 µm layer. Overhanging bridge showed modest cure-through of 113 ± 7% at 20 s. Complex 3DBenchy printed at 35 s per layer with a build rate of 5.1 mm h−1 (excluding peeling).
• Chemical depolymerization and recovery: Catalytic depolymerization (MeTHF, 1 mol% P1-t-Bu:thiophenol, 80 °C, 3 h, N2) dissolved prints and recovered resin in up to 98% yield. SEC of recycled resin matched initial resin with minor oligomers; 1H NMR indicated ~85 mol% monomeric purity. The MenLp1:IsoLp2 ratio after depolymerization was 31:69 wt% (≈1:1.53 mol), within experimental error of the initial formulation.
• Structure–property tunability: For MenLp1-IsoLp2 2D photosets, increasing crosslinker content raised mechanical properties and Tg. Examples: 30 wt% IsoLp2: UTS 0.9 ± 0.1 MPa, E 0.014 ± 0.002 MPa; 90 wt% IsoLp2: UTS 15.6 ± 0.9 MPa, E 3.3 ± 0.2 MPa; Tg spanned −30 to 20 °C. All networks showed Td > 190 °C.
• Alternative monomers/crosslinkers: GuaLp1-IsoLp2 gave properties similar to MenLp1-IsoLp2; EtLp1-IsoLp2 lowered Tg (e.g., Tg 16 °C) and strength; SteaLp1-IsoLp2 showed a near-ambient Tm but was brittle. GlyLp3-based networks generally provided larger rubbery plateaus (≥100 °C) and improved dimensional stability; EtLp1-GlyLp3 exhibited an extended rubbery plateau (~10–160 °C).
• Depolymerization scope: Catalytic depolymerization of 2D photosets yielded 24–96% (unoptimized); 3D-printed EtLp1-GlyLp3 afforded 97% recycled resin yield with 96% cyclic disulfide content. Catalyst-free thermal depolymerization in DMF was broadly effective for 2D photosets (≥80% yields) and for 3D EtLp1-GlyLp3 (91% yield, 96% cyclic disulfide content).
• Closed-loop cycles: EtLp1-GlyLp3 (31:69 wt%) underwent two recycle cycles via thermal depolymerization with high yields (91% first, 94% second). Cyclic disulfide content: 96% (first), 94% (second). Resin compositions remained consistent (EtLp1:GlyLp3 ≈ 32:68 wt% first, 31:69 wt% second). SEC showed similar molecular weights with small oligomer fractions (>1,000 g mol−1). UV–vis of recycled resins showed a shoulder near 290 nm consistent with oligomers.
• Curing kinetics and print fidelity post-recycling: Photorheology indicated rapid gelation; with photoinitiator added, all resins reached gel point within 1.4 s and exhibited similar plateau moduli. Z-depth cure profiles were comparable across pristine and recycled resins. X–y accuracy remained high with a slight decrease after the second recycle. Minimal cure-through along z (123 ± 10 to 129 ± 21%).
• Health and safety and sustainability: Lipoate-based resins avoid sensitizing (meth)acrylates and are derived from renewable feedstocks; lipoic acid is already produced at scale. Hydrolytic depolymerization regenerates monomers for potentially infinite closed-loop recycling.

The work demonstrates that integrating strained cyclic disulfides (lipoates) as the photoreactive units achieves the key orthogonality needed for circular vat photopolymerization: the dynamic bond is created during photopolymerization and can be selectively reversed to regenerate the original monomers or resin mixtures. High-resolution printing performance, efficient depolymerization with high recovery, and preservation of resin composition and curing behavior over multiple recycling cycles directly address the long-standing challenge of closed-loop recycling for photopolymer resins. The platform is modular, enabling tuning of mechanical and thermal properties across orders of magnitude and approaching or matching ranges of commercial flexible/elastic resins. The combination of renewable sourcing, potential biodegradability, and the ability to chemically recycle or regenerate monomers suggests significant environmental and health advantages over conventional acrylate/epoxy systems. Minor oligomer formation during depolymerization slightly affects absorption and curing rates but can be mitigated by adding photoinitiator, without compromising bulk properties. Overall, the results validate a practical pathway toward circular additive manufacturing for vat-photopolymerized materials.

This study introduces a renewably sourced lipoate-based resin platform for vat photopolymerization that enables high-fidelity 3D printing, efficient depolymerization, and circular reuse. The materials exhibit tunable mechanics and thermal behavior, with properties comparable to commercial soft resins, and maintain resin composition and print performance across at least two recycle cycles. The approach offers health and sustainability benefits versus (meth)acrylate resins and provides multiple depolymerization routes, including monomer-regenerating hydrolysis for potential infinite closed-loop use. Future work should focus on eliminating oligomeric byproducts to improve depolymerization orthogonality, further enhancing mechanical performance (e.g., via non-covalent reinforcement such as hydrogen bonding or metal coordination), optimizing z-resolution with benign opaquing agents, and scaling greener monomer syntheses to minimize environmental, health, and safety concerns.

• Minor oligomeric contaminants persist after depolymerization, slightly altering UV–vis absorption and curing rates; added photoinitiator is needed to match pristine curing kinetics.
• The phosphazene-catalyzed depolymerization is not universal across all formulations and phosphazenes present cytotoxicity concerns; greener catalytic systems are desirable.
• 2D photosets showed variable depolymerization yields (24–96%), potentially due to overcuring/photo-oxidation of disulfides; process control and formulation optimization may be needed.
• Storage and synthesis considerations: IsoLp2 can gel upon concentration or thermal exposure; EDC coupling and chlorinated solvents raise EHS concerns though alternative Fisher esterification routes were demonstrated.
• Some formulations (e.g., SteaLp1-IsoLp2) yielded brittle materials with near-ambient Tm, limiting their mechanical applicability without further modification.