Electro-capillary peeling of thin films

Thin films underpin diverse technologies including flexible electronics, soft robotics, MEMS/NEMS, and biomedical devices. As films scale to micro/nanoscale, interfacial delamination and cracking threaten device reliability, motivating nondestructive methods to peel, transfer, and reuse films. Conventional mechanical peeling approaches (e.g., debonded strip and blister methods) are difficult to apply to fully attached micro/nanofilms and can induce damaging strains. Various strategies modify the bonding layer or interface chemistry, but often require complex preparations and have limited applicability across materials. The study proposes an active, simple, and broadly applicable peeling method that uses an electric field to drive a thin liquid layer to percolate and spread within the bonding layer, overcoming adhesion while minimizing film deformation. The central hypothesis is that Maxwell stress acting on charges in the liquid can be used to compete with interfacial adhesion and film tension to achieve controllable, low-strain detachment across diverse films.

Existing approaches to detach thin films include mechanical methods such as the debonded strip and blister test, and chemistry-driven methods such as hydrogen-bubble-assisted delamination of graphene on catalysts, drying-induced peeling for colloidal films, water-soluble sacrificial bonding layers, and adhesion reduction via chemical/physical surface modification. Water-based capillary peeling has shown that a liquid layer can passively detach hydrophobic films with low adhesion, but it is less effective for high-adhesion or hydrophilic systems and affords limited control over rate. Prior work demonstrates that electric fields can control liquid wettability and contact line motion via Maxwell stress and electrokinetic effects. These insights motivate using an electric field to actively drive a percolating liquid layer in the bonding interface to achieve on-demand peeling without complex interface modification, potentially extending applicability to a wider range of films and adhesion energies.

Experimental setup: A PDMS thin film was plasma-treated and bonded onto ITO glass, forming a strong, uniform bonding layer (wet adhesion ~133 mJ/m^2 by water blister test). A circular hole (diameter 2 mm) was fabricated at the film center. An electrolyte droplet (typically 20 μL of 1.0 mol/L KCl) was injected into the hole using a micropump. A DC supply applied voltage between two Pt wire electrodes: the positive electrode inserted into the droplet, the negative on the ITO surface. Upon applying voltage, the liquid percolated into the bonding layer, peeling the film radially; the peel front radius r(t) and out-of-plane lift were imaged by synchronized side- and top-view cameras. Voltage characterization: Axisymmetric and planar peeling modes were tested at 1.5–4.5 V. Lifting height and peeling length were quantified over time up to ~80 s (limited by full spreading of 20 μL in some bonding layers). Film library: PDMS films were fabricated by spin coating with thicknesses 25, 50, 100, 200, 300 μm (controlled by spin rate) and elastic moduli 1.0, 1.6, 2.4 MPa (controlled by base:curing agent ratios of 20:1, 15:1, 10:1). Functional films including hydrogel, PET, and PEN were also bonded and tested. SEM verified bonding layer morphology consistency across materials. Wet adhesion was measured by water blister tests; Young’s moduli were measured via tensile testing. Deformation measurements: Three-dimensional digital image correlation (3D DIC) characterized displacement and strain fields during peeling with 10 μm spatial resolution. Comparisons were made to strains induced by debonded strip and blister methods at matched peeling rates. Large-area implementation: A liquid replenishment strategy maintained sufficient electrolyte to sustain long-term peeling for a 5 cm diameter film; post-peel characterization included profilometry and SEM/EDS of residues. Electrolytes and solvents: Additional electrolytes (CaCl2, CuSO4, LiCl, NaCl, KOH, NaOH) across 0.1–1.0 mol/L and organic solvents (alcohol, acetone, glycerol solutions) were tested for method generality. Theoretical model: A force/energy-based model treats the film as an elastic membrane with tension T = E ε d0. Film deflection w(r) under net pressure p from Maxwell stress TM minus negligible dead weight satisfies d^2w/dr^2 = εr ε0 Ee^2 / T, with Ee ≈ U/(r^2+z1^2) and boundary conditions w|r=r0 = dw/dr|r=r0 = 0, yielding an analytical deflection profile. At onset, the critical electric force balances elastic and adhesion energies, leading to a critical peeling voltage Uc proportional to film thickness d0 and modulus E. Liquid spreading is modeled via Poiseuille flow within the interfacial gap, giving r ∝ (εr^2 ε0^2 U t)/(μ E^2 d0^2)^(1/2), predicting r ~ t, r ~ U, r ~ d0^-2, r ~ E^-2. Predictions are compared with experiments.

• Electro-capillary peeling actively detaches films by driving liquid to percolate and spread into the bonding layer under a DC electric field, significantly reducing film strain versus traditional methods (up to 86% reduction stated).
• Voltage dependence: At 1.5, 2.5, 3.5, 4.5 V, average peeling rates were 0.013, 0.023, 0.038, 0.048 mm/s, respectively (linear in voltage). Lifting heights at these voltages were ~0.23, 0.51, 0.54, 0.56 mm; lifting height increased markedly from 1.5 to 2.5 V and then plateaued.
• Time law: Peeling length r increased approximately linearly with time over the observed window (up to ~80 s for 20 μL droplets).
• Film thickness effect: Increasing PDMS film thickness from 25 to 300 μm decreased lifting height from 0.62 to 0.13 mm and peeling length from 3.61 to 0.97 mm (t ≈ 80 s). Thicker films peel more slowly and may approach negligible rates.
• Elastic modulus effect: Increasing PDMS modulus from 1.0 to 2.4 MPa reduced lifting height by 89.3% and peeling length by 64.0% over the same time window.
• Material generality: Method successfully detached hydrogel, PET, and PEN films with differing moduli and adhesion energies; relative performance depended on properties (hydrogel showed largest lift; PEN largest peeling length among the three).
• Theoretical scaling validated: Experiments support predictions r ~ t, r ~ U, r ~ d0^-2, r ~ E^-2; critical peeling voltage increases with thickness and modulus, consistent with the model.
• Low-strain peeling: 3D DIC showed maximum out-of-plane displacement ~0.152 mm and peak strain <0.00332 during electro-capillary peeling. At equal peeling rates, blister/debonded strip methods induced up to ~6× larger strains; ZnO nanorod layers suffered cracking after repeated debonded-strip peeling but remained largely intact after electro-capillary peeling.
• Practical threshold: A PDMS film (d0 = 100 μm, E = 1.0 MPa) could be peeled from ITO glass at a voltage as low as ~0.7 V, implying feasibility with simple battery power.
• Scalability and residues: With liquid replenishment, a 5 cm diameter film was fully peeled in ~8.3 min at a stable rate; KCl residue deposited on the substrate (not on the film) was removable by water rinsing; peeled films showed no residual deformation after transfer.

The study addresses the need for a simple, nondestructive, and broadly applicable method to detach micro/nanofilms by harnessing Maxwell stress to actively drive interfacial liquid spreading. Results show precise control of peeling rate and extent via applied voltage, validating the proposed physics and enabling on-demand operation. The linear r–t and r–U relations and inverse dependences on film thickness and modulus align with the theoretical model, confirming the mechanism where electric forcing competes with adhesion and membrane tension. Compared to passive capillary peeling and conventional mechanical methods, the electro-capillary approach operates across hydrophilic to hydrophobic films, wide adhesion energies (≈37–3656 mJ/m^2), and diverse electrolytes/solvents, while imposing ultralow strain that preserves sensitive surface structures (e.g., ZnO nanorods). The low critical voltages (down to ~0.7 V for 100 μm PDMS) demonstrate practical, energy-efficient implementation, and liquid replenishment supports large-area uniform peeling. Collectively, the findings establish electro-capillary peeling as a versatile technique for film transfer and device protection in flexible electronics and related applications.

This work introduces electro-capillary peeling, an active, low-strain method to detach thin films by electrically driving a liquid to percolate and spread within the bonding interface. Experiments demonstrate controllable peeling kinetics (r ~ t, r ~ U) and strong dependence on film properties (r ~ d0^-2, r ~ E^-2), validate increased critical voltage with thickness and modulus, and show broad applicability to PDMS, hydrogel, PET, and PEN films. The method significantly reduces strain versus traditional peeling, preserves surface nanostructures, operates at very low voltages, and scales to large areas with liquid replenishment. Future research could optimize interface geometries and electrode configurations to further lower critical voltages for stiffer/thicker films, extend to additional film types and bonding chemistries (including high-adhesion and water-sensitive systems via tailored solvents), refine the fluid-structure-electrostatics model (e.g., incorporating bending stiffness and 3D effects), and integrate the approach into automated transfer processes for industrial manufacturing.

• Performance diminishes for thicker or stiffer films; peeling rates decrease and may become negligible without higher voltages or modified conditions.
• Residual electrolyte (e.g., KCl) can deposit on the substrate during peeling, requiring post-process cleaning, though films remained clean.
• The baseline droplet volume (20 μL) can fully spread within some bonding layers by ~80 s, limiting observation windows unless liquid is replenished.
• The theoretical model neglects film bending stiffness and simplifies the electric field distribution; these assumptions may limit quantitative accuracy for certain geometries or materials.
• Lifting height exhibited weak dependence on voltage beyond ~2.5 V, indicating potential saturation effects and constraints on vertical deformation control.