Shape morphing of plastic films

The study addresses the challenge of creating free-standing 3D architectures from common inert plastic films that are typically used as substrates in flexible electronics (e.g., PET, PI, PTFE). Existing fabrication methods largely confine device substrates to 2D forms, and prevalent shape-morphing approaches rely on heterogeneous or stimuli-responsive materials, limiting applicability to inert homogeneous plastics. The authors propose and explore a mechanical peeling strategy to program asymmetric plastic strains in thin films, enabling transformation of 2D precursors into diverse 3D shapes. The purpose is to establish a general, predictable, and versatile method that expands 3D device architectures without requiring specialized responsive materials, thereby advancing functional 3D/4D electronics and materials.

Prior work demonstrates the value of 3D architectures in electronics for increased functionality and spatial resolution. Shape morphing has been realized using heterogeneous responsive systems such as bilayer polymers, patterned polymers, liquid crystal polymers, and gradient-structured films, enabling bending, twisting, and simple 3D devices. Compressive buckling of 2D micro/nanostructures on contractive substrates has also produced 3D forms. However, these methods are constrained to heterogeneous or responsive materials and are not suitable for creating free-standing 3D structures from inert, homogeneous plastic films like PET, PI, and PTFE. The literature on peel mechanics indicates plastic yielding can occur during peeling, suggesting a route to program residual strains for morphing, but a general strategy for controlled 3D shaping of inert films had been lacking.

• Concept and modeling: The authors develop a mechanical peeling model to describe how peeling induces asymmetric plastic strains in films, leading to shape morphing upon release. Two key control parameters are defined: peeling angle φ (supplementary dihedral angle between detached and adhered parts) and deviation angle δ (rotation from the short axis to the detaching line). A theoretical model (zero deviation angle case) predicts post-release curvature κ as a function of material properties (Young’s modulus E, hardening modulus H, yield strain ε0), film thickness h, and peeling parameters, with closed-form expressions for κh depending on r = ε/ε0. The model also yields peeling force per unit width and detachment-front conditions and is validated by 2D/3D FEA.
• Finite-element analysis (FEA): Full 3D ABAQUS simulations model detachment and relaxation of PI films, using cohesive elements for the adhesive layer (quadratic stress criterion, BK mixed-mode law) and elastic–plastic film constitutive behavior. Parameters: PI E=2.5 GPa, ν=0.35; adhesive stiffness En=52 MPa mm−1, Es=16.5 MPa mm−1, initial damage stresses tn=0.52 MPa, ts=0.55 MPa, fracture energies Gn=5.2 N m−1, Gs=14.67 N m−1. 2D plane-stress simulations validate the theoretical model for δ=0; quantitative agreement with minor deviations due to model simplifications.
• Materials and fabrication: Films studied include PET, PI, PTFE, PVP, PVB, PEO, PVDF, PVDF-HFP, PLA, PES filter membrane, weighing paper, Al and Cu foils. Adhesives include PDMS layers (Sylgard 184) with varied base:crosslinker ratios (10:1–60:1) to tune adhesion energy and viscoelasticity, and commercial tapes (Kapton, scotch). PDMS thicknesses 25–200 µm were coated by blade and cured at 80 °C for 2 h.
• Peeling experiments: A tensile machine with a slidable substrate fixes φ and controls speed; films (4–5 mm wide belts) are peeled at controlled speeds (e.g., 10 mm/s) while imaging the three-phase line. Parameters varied: adhesion energy (via PDMS ratio), peeling speed, adhesive thickness, film thickness and modulus, peeling angle φ (30°–180°), and deviation angle δ.
• Programming shapes: By modulating φ and δ during peeling, the team fabricates curls (δ=0), helices (δ≠0) with programmable chirality and pitch, spirals (vary φ), conical spirals (vary φ and δ simultaneously), polygons via discrete adhesion patterns (tape widths and intervals computed from target interior angles and film diameter), and hyperboloids via two-step orthogonal peelings with selective cutting.
• Device demonstrations: 2D Au-on-plastic circuits (Au ~70 nm by thermal evaporation after O2 plasma) are transformed into 3D (helical interconnects). P(VDF-TrFE) piezoelectric polymer films (∼16 µm with Cr/Au electrodes) are morphed into spiral cantilevers to sense vibrations; voltages are recorded under controlled excitations. Responsive bilayers (PEO/PDMS) are morphed and then actuated by humidity (4D), showing reversible shape changes depending on layer orientation. Elastomer 3D films are cast using morphed plastic templates and cured to retain 3D shapes, with optional transfer of metallic patterns.

• Mechanism and control: Peeling induces asymmetric plastic strains determined by φ (degree) and δ (orientation), enabling programmed 3D morphing of inert homogeneous films. FEA and theory quantitatively predict curvature and peeling forces.
• Parameter effects on curvature and shape:
  • Higher adhesion energy (softer, less cross-linked PDMS) increases required peeling force and post-peel curvature.
  • Higher peeling speeds increase fracture energy and peeling force, yielding greater curvature; experiments match theory.
  • Thicker adhesive layers (30–200 µm) slightly increase peeling forces (26.8 ± 1.3 to 30.3 ± 0.3 N/m) but reduce curvature (0.37 to 0.13 mm−1) due to increased separation and energy dissipation.
  • Thicker or higher-modulus films bend less during peeling, producing smaller curvatures.
  • Increasing φ from 30° to 180° increases bending and curvature; δ=0 yields tubes/curls; δ≠0 yields helices with chirality set by the sign of δ.
  • Helix pitch P scales with deviation angle and diameter (P ≈ π d tan δ), consistent with experiments; positive δ gives right-handed, negative δ left-handed helices.
• Complex 3D structures via programmed peeling: Gradually varying φ produces spiral-like shapes; varying δ yields helices with graded pitch; switching δ’s sign generates mixed-chirality helices; simultaneously varying φ and δ yields conical spirals. Discrete adhesion patterning produces polygons (triangles, quadrangles, pentagons, hexagons) with adhesive widths/intervals computed from target interior angles. Two-step orthogonal peelings plus selective cuts form hyperboloids.
• Materials generality: Method applies to polymers, metals, and composites capable of plastic deformation; also to bilayers (plastic–elastomer), though elastomer layers reduce curvature.
• 3D electronics: Au–plastic films show only slight resistance increase with curvature; 3D helical interconnect circuits function reliably under stretching, with LED rotation tracking chirality during deformation.
• Piezoelectric performance: 3D P(VDF-TrFE) spiral cantilevers sense broader frequencies and generate higher voltages than 2D films. Example: single cantilever 3_2.09_4.5 peak 53.4 mV; baseline 2D (1.5 mm × 3 cm) peak 5.7 mV; frequency bands expand (e.g., 24–92 Hz >1 mV vs 27–53 Hz). Multi-arm integration further improves performance (19–184 Hz >1 mV; max 98.2 mV).
• 4D transformation: PEO/PDMS bilayers morphed into 3D shapes undergo humidity-responsive transformations. A PEO/PDMS triangle transitions among polygonal shapes with humidity (drier: triangle→quadrilateral→pentagon→hexagon; wetter: ellipse-like). Cylindrical helices tighten under low humidity and loosen under high humidity when PEO is inward; the opposite occurs when PEO is outward.
• Template casting of 3D elastomers: Semi-cured PDMS cast on morphed plastic templates cures into free-standing 3D elastomer films; metallic patterns on templates can be transferred to yield 3D stretchable electronics.

The findings demonstrate that controlled mechanical peeling can program residual plastic strains in inert films, enabling predictable morphing from 2D to diverse free-standing 3D architectures without relying on heterogeneous or stimuli-responsive materials. By tuning peeling parameters (φ, δ, speed, adhesion energy, adhesive/film thickness, modulus), both global and local curvatures are engineered, yielding complex shapes such as helices with tunable chirality and pitch, spirals, polygons, and hyperboloids. The approach is general across materials that plastically deform and integrates naturally with device fabrication: 3D circuits maintain electrical function with minimal resistance change, while 3D piezoelectric spirals exhibit broadened frequency response and higher output, highlighting structural-performance benefits. Coupling with active materials (e.g., PEO/PDMS) extends the method to 4D morphing, where environmental stimuli further reconfigure shapes post-morphing. Overall, the strategy addresses the challenge of creating free-standing 3D structures from inert plastic substrates central to flexible electronics and opens avenues for scalable 3D/4D device architectures.

Peeling-induced asymmetric plastic deformation provides a simple, versatile route to program 3D shapes from 2D inert plastic films over millimeter-to-micrometer scales. A theoretical model and FEA accurately predict post-peel curvature and forces, enabling rational design via peeling angle and deviation angle control. The method generalizes across plastic, metallic, and composite films, supports complex geometries via programmed single/multistep peelings, and integrates with device platforms to realize functional 3D circuits and enhanced piezoelectric systems. Incorporation of responsive layers enables 4D transformations, and morphed plastic templates can cast 3D elastomeric devices with transferable circuitry. Future work could refine constitutive modeling to reduce quantitative discrepancies, expand automated control of peeling trajectories for higher design complexity, and explore integration into manufacturing workflows for large-scale 3D/4D flexible electronics.

• Theoretical simplifications in the elastoplastic constitutive model lead to small quantitative errors versus FEA/experiments; improved models could enhance prediction accuracy.
• Complex peeling sequences can pose boundary-condition and convergence challenges in direct 3D simulations, necessitating a multi-step eigenstrain strategy.
• Viscoelastic adhesive behavior and thickness influence outcomes; thicker adhesive layers reduce achievable curvature and increase energy dissipation.
• Adding elastomer layers to form bilayers decreases curvature compared to single plastic films, limiting tightness of morphing in some applications.