Rejuvenating liquid crystal elastomers for self-growth

Inspired by plant growth, soft growing robots can extend their bodies to navigate unstructured environments and typically rely on pneumatic eversion, pressurized elongating tubes, chain-locking mechanisms, or additive manufacturing. While such systems advance assembly and manufacturing, there remains a need for materials that can spontaneously and independently grow without external stimuli or energy input. Liquid crystal elastomers (LCEs) are promising shape-shifting polymers due to large actuation strain and muscle-like energy density. Conventional LCE actuation relies on the liquid crystal–isotropic phase transition triggered by external stimuli. A rare exception is a specific LCE system previously reported to self-grow at room temperature without external input, but only freshly prepared samples in their initial state can self-grow; non-fresh (annealed or previously grown) samples cannot. This requirement to synthesize fresh materials from monomers for each use limits practicality. Rejuvenation is challenging because it must disrupt π-π interactions formed in the stable state and rearrange the crosslinked network topology. Fresh self-growing LCEs are intrinsic vitrimers containing ester bonds and the volatile transesterification catalyst dipropylamine (DPA). Non-fresh LCEs lack catalyst due to DPA volatility and are therefore unexchangeable. Even non-fresh LCEs with catalyst may still fail to self-grow if the network topology is not reset. The authors propose converting permanently crosslinked non-fresh LCEs into dynamic vitrimers by swelling with transesterification catalysts. During swelling, solvent disrupts π-π interactions to temporarily revert the network toward the initial state while catalyzed transesterification rearranges the topology, enabling rejuvenation, erasing growth history, and restoring on-demand self-growth and actuation.

The paper situates its work within soft growing robotics and shape-shifting polymers: prior artificial growth strategies include pneumatically driven everting skins, pressurized elongating tubes, chain-locking blocks, and additive manufacturing. LCEs are highlighted for large actuation strain and energy density, with typical actuation requiring external stimuli via LC–isotropic transitions. Only one LCE-based polymer system was recently reported to self-grow autonomously at room temperature. Fresh self-growing LCEs function as vitrimers due to ester bonds and the presence of a transesterification catalyst (DPA), enabling network rearrangement via dynamic covalent chemistry. However, DPA’s volatility leads to catalyst loss in non-fresh samples, leaving a permanently crosslinked, non-exchangeable network that cannot self-grow. Vitrimer literature is referenced to support the concept of dynamic topology rearrangement through exchange reactions. The work addresses a gap: enabling non-fresh, annealed or previously grown LCEs to regain the initial state required for self-growth without synthesizing new material from monomers.

Materials: RM257 (LC diacrylate monomer), EDDET (dithiol chain extender), PETMP (tetra-thiol crosslinker), DPA (base catalyst), TBD and acetic acid (for neutralized TBD, nTBD), and dichloromethane (DCM) as solvent. Synthesis of self-growing LCEs (F-LCE): Thiol–Michael addition using stoichiometric acrylate and thiol functionalities. RM257 (1 mmol), EDDET (0.6 mmol), PETMP (0.2 mmol) dissolved in 5 mL DCM, mixed and catalyzed with DPA (15 µL). The mixture was cast in a PTFE mold (4 cm × 4 cm × 0.5 cm) and cured at room temperature for 12 h. FTIR confirmed full conversion (loss of RM257 C=C at 1637 cm⁻¹ and thiol S–H at 2568 cm⁻¹). Gel content ≈97.7%. Preparation of non-fresh states: Annealed LCE (A-LCE) obtained by drying F-LCE in vacuum at 120 °C for 48 h. Grown LCE (G-LCE) obtained by pre-stretching F-LCE by 30%, fixing both ends on a glass slide, and allowing self-growth in an incubator at 30 ± 0.5 °C. Rejuvenation (R-LCE): Non-fresh A-LCE or G-LCE immersed in DCM containing nTBD (3 mg mL⁻¹) for 48 h at ambient temperature to swell and introduce catalyst, then solvent removal. nTBD loading in LCE ≈0.84 wt%. FTIR of annealed R-LCE showed an N–H stretch at 1646 cm⁻¹, evidencing nTBD incorporation. After extended swelling, A-LCE shape was similar, whereas G-LCE reverted to original oriented state, erasing growth history. Regrowth and actuation: After rejuvenation, samples were pre-stretched by 30% and fixed at both ends, then allowed to self-grow at room temperature (≈30 °C incubator used for controlled conditions). Regrown R-LCEs became soft actuators at room temperature. Orientation and growth history were evaluated by POM and XRD. Selective and local rejuvenation: Half-part rejuvenation performed by immersing only half the sample in DCM with nTBD (3 mg mL⁻¹) for 48 h while immobilized, then solvent evaporation, 30% pre-stretch, fixation, and self-growth at 30 °C. Local rejuvenation achieved by dropping DCM+nTBD (3 mg mL⁻¹) onto selected regions, then briefly exposing to DCM vapor to maintain overall slight swelling, followed by solvent evaporation, pre-stretching, and growth. Characterization: DSC (TA Q2000) for Tg and isotropic transition temperature (Ti), TGA (TA Q50) for thermal stability, DMA (TA Q-800) for tensile tests and room-temperature stress relaxation, rheometry (TA AR-G2) for stress relaxation (100–160 °C) and creep resistance (80–140 °C, 600 min), X-ray diffraction (Bruker D8, Cu Kα, λ=1.54 Å) for orientation and d-spacing, POM (Nikon ECLIPSE LV100POL) for birefringence, FTIR (PerkinElmer Spectrum 100). Stress relaxation times were analyzed via Arrhenius plots and fitted with the KWW model. Conditions during DSC for transesterification studies were limited to ≤100 °C to avoid aromatic ester heterolytic exchange. Controls and comparisons: Pure solvent swelling (no catalyst) tested and found insufficient for rejuvenation; catalyst comparison conducted between nTBD and DPA at 3 mg mL⁻¹ and varying swelling times (24, 48 h). Swelling time (1–48 h) and catalyst concentration (0.3–3 mg mL⁻¹) were varied to study effects on elongation and actuation strains.

- Rejuvenation restores self-growth: Swelling non-fresh LCEs (A-LCE, G-LCE) in DCM with nTBD (3 mg mL⁻¹, 48 h) reintroduced dynamic covalency and disrupted π-π interactions, reverting samples to an initial-like state that self-grows at room temperature after 30% pre-stretch and fixation. Growth history of G-LCE was erased, as evidenced by POM and isotropic XRD rings post-rejuvenation. - Thermal and mechanical parity with fresh LCEs: F-LCE and R-LCE showed similar Tg (≈ -2.9 °C vs ≈ -6.6 °C) and Ti (≈ 74 °C vs ≈ 69–74 °C depending on cycle). TGA 5% weight-loss temperatures: 342.5 °C (annealed F-LCE) and 325.9 °C (annealed R-LCE). Mechanical properties of A-LCE vs annealed R-LCE: fracture stress 2.56 vs 2.25 MPa; fracture elongation 409% vs 390%. XRD d-spacings similar (peak at 2θ ≈ 20.0°, d ≈ 4.44 Å; after annealing peaks at 2θ ≈ 19.5°, d ≈ 4.55 Å). Order parameters S for oriented states after regrowth: 0.61 (G-LCE) and 0.59 (regrown R-LCE). - Creep resistance: At 80 °C negligible deformation; at 90 °C only 0.36% strain over 600 min; creep onset apparent at 120–140 °C. - Repeatability: Rejuvenation–regrowth cycles repeated ≥5 times on one sample with consistent actuation strains. Original G-LCE actuation strain 51.0%; after rejuvenations: 49.3%, 47.1%, 48.7%, 50.2%. - Process parameters: • Swelling time: Elongation strain rose with time: 1 h 38.8%, 6 h 45.7%, 12 h 48.2%, 24 h 51.5%, 48 h 52.1%. Actuation strain: 38.7%, 42.7%, 47.2%, 49.9%, 51.5% respectively. Minimal increase between 24 and 48 h indicates near-equilibrium transesterification. • Catalyst concentration (48 h): nTBD at 0.3, 1.5, 3 mg mL⁻¹ yielded elongation 47.0%, 51.5%, 52.1% and actuation 43.3%, 45.4%, 51.5%. • Catalyst type (3 mg mL⁻¹): nTBD outperformed DPA. For 24/48 h swelling, elongation with nTBD: 51.5%/52.1%; with DPA: 41.6%/46.2%. Actuation after 48 h: 51.5% (nTBD) vs 32.6% (DPA), attributed to DPA volatility reducing alignment fixation. - Network dynamics: Stress relaxation followed Arrhenius behavior with activation energy 104.3 kJ mol⁻¹ and topology freezing transition temperature ~57.1 °C; KWW analysis indicated complete relaxation at long times due to exchange reactions. FTIR evidenced catalyst uptake; DSC around 74 °C Ti showed aromatic esters did not undergo heterolytic exchange under rejuvenation conditions; NMR of RM257 with nTBD showed no acrylate consumption. - Local and programmable growth: Selective swelling enabled half-sample growth and spatially programmed growth in patterned regions (T-shaped and plus-shaped), transforming 2D films into 3D actuators through localized rejuvenation and subsequent regrowth. - Catalyst retention: nTBD remained functional in annealed R-LCE after 50 °C annealing and for >6 months, enabling subsequent cycles with reduced catalyst or even solvent-only swelling due to residual catalyst.

The work addresses a key barrier to deploying self-growing LCEs: only freshly prepared materials could previously self-grow, preventing practical storage and reuse. By leveraging synergistic physical and chemical effects during swelling—solvent-mediated disruption of π-π interactions to temporarily isotropize the LC network and catalyst-enabled transesterification to rearrange network topology—the authors reset non-fresh LCEs (annealed or previously grown) to an initial-like state. This rejuvenation restores autonomous self-growth and actuation at room temperature, erases prior growth history, and permits repeated cycles without significant loss in actuation strain or mechanical integrity. Parameter studies identify how swelling duration, catalyst concentration, and catalyst identity govern elongation and actuation outcomes, with nTBD notably superior to DPA due to lower volatility and sustained catalytic action. Rheological analysis confirms dynamic network behavior consistent with vitrimer-like exchange and a topology freezing transition compatible with room-temperature stabilization of programmed alignment. Demonstrations of selective, localized rejuvenation show fine spatial control of growth and shape morphing, enabling reprogramming of actuation modes and conversion from planar to complex 3D forms. Collectively, these findings enable on-demand reuse, storage, and programmable growth of LCEs, advancing fully untethered soft robots that do not require external stimuli or power during growth.

A swelling-enabled rejuvenation strategy converts non-fresh LCEs into dynamic, self-growing materials by combining solvent-induced disruption of mesogenic π-π interactions with catalyst-driven transesterification. Rejuvenated LCEs recover the initial state required for spontaneous room-temperature growth, erase prior actuation histories, and can repeatedly regrow into stable soft actuators with actuation strains around 50% and properties comparable to fresh LCEs. The approach supports spatially selective rejuvenation for local growth and programmable actuation, enabling reversible transitions from 2D to 3D structures. This deepens understanding of self-growing LCE mechanisms and removes practical barriers by allowing long-term storage and on-demand deployment. The strategy positions rejuvenated LCEs as promising structural materials for untethered autonomous soft growing robotics, particularly in energy-limited scenarios such as exploration and rescue. Future work could further optimize catalyst systems, solvent processes, and spatial patterning methods to expand complexity, speed, and robustness of programmable growth.

- Rejuvenation requires solvent swelling and a transesterification catalyst; pure solvent alone temporarily disrupts ordering but does not produce permanent rejuvenation or regrowth. - Catalyst identity matters: DPA’s volatility limits its effectiveness in fixing alignment, yielding lower actuation strains compared to nTBD. - Process sensitivity to conditions: Sufficient swelling time (≈24–48 h) and adequate catalyst concentration (up to 3 mg mL⁻¹) are needed to approach maximal elongation/actuation; insufficient parameters reduce performance. - Thermal limits: While creep is minimal at 80–90 °C, noticeable creep occurs at ≥120 °C; thermal processing must avoid conditions that could trigger unwanted reactions (e.g., aromatic ester heterolytic exchange), necessitating lower-temperature drying (e.g., 50 °C for R-LCE). - Pre-stretch and fixation are required to realize self-growth and subsequent actuator formation, adding procedural steps and fixturing requirements. - Although nTBD retention aids subsequent cycles, catalyst leaching or distribution heterogeneity over many cycles was not exhaustively quantified.