Text from the interactive explanation
Mechanics of electroadhesion by polycation interfacial bridging: sticky electrophoresis, ionic complexation and chain entanglement
Binbin Ying, Kun-Hao Yu, Shu Yang, Jiawei Yang
Original paperElectroadhesion · the e‑GLUE model
e-GLUE network
polycation (+)
anionic network (−)
ionic pair
Hp / Hc (bridging limits)
penetration front
The lining of the gut is wet, always moving and renews itself quickly, so adhesives that only grip its surface can be cleared in under a day. The e-GLUE takes another route: a soft gel carrying long, positively charged chains that a brief electric field drives into the lining's negatively charged mesh. This paper builds the model of how that stitch forms and lets go — then tests it on the bench.
Ying, Yu, Yang & Yang · Proc. R. Soc. A 482: 20260579 (2026) · doi.org/10.1098/rspa.2026.0579
01 The cast
Five players in a stitch only nanometres deep
Ordinary mucosal adhesives bond to the surface. The authors' idea is to adhere over a depth instead. In their earlier work (Ying et al., Sci. Transl. Med. 2025, a study in pig gut) the e-GLUE held strongly to mucosa for up to 30 days. This paper asks how the stitch actually forms.
e-GLUE
A soft, electrically neutral polymer gel — in the tests, polyacrylamide — threaded with mobile, entangled polycations . The network holds still; the charged chains can move.
Polycation
A long chain with many positive charges (in the tests, PDAC), modelled as Kuhn segments of b = 3 nm with about 5 charges each. A 100 kDa chain has ≈124 segments: contour ≈372 nm, but coiled to a size le ≈ 33 nm.
Anionic network
Mucosa behaves like a mesh with fixed negative charges (a major component is anionic proteins). The bench stand-in is a polyacrylamide gel with an alginate network; the model uses 0.085 M fixed anions.
Ionic pair
A + on a chain meets a − on the mesh. Breaking one takes about 10 pN, yet pairs constantly come apart and re-form with neighbouring sites.
Tube & reptation
Surrounding entangled chains confine each chain to a virtual tube — a topological constraint, not a real pore. It can only slither along its own length: reptation.
02 Slither
Each chain moves like a snake in a tube
A chain cannot push sideways through its neighbours. It can only creep forward or back along its own path, the head choosing the way and the body following. Drag the bright head, or use the buttons. With the field on, random back-and-forth jiggling gains a slight head-first drift toward the − side.
Head 8% along the tube
03 A push too weak to break a bond
The field can't rip bonds. It tilts the odds.
At the highest field used in the experiments, 8000 V/m, the electric force on a single charge is about 10−15 N. Breaking one ionic pair takes about 10−11 N. That is roughly ten thousand times more.
≈10,000× weaker
So how do chains get through? Stickers.
Ionic pairs are not permanent. They come apart and re-form with neighbouring sites all the time. Each pair acts as a sticker that slows a chain without freezing it. The authors call this sticky electrophoresis.
In the model, the field lowers the barrier a chain must overcome to move forward, a Bell-like rule borrowed from bond mechanics: ξ(E) = ξ0·exp(−W/kBT), with W = Q·L·E. Here Q is the chain's total charge and L an activation length (the Kuhn length, 3 nm). Lower friction ξ means faster diffusion, D = kBT/(n·ξ).
MODEL · PAPER Depth reached after 15 s (Fig. 6a)
But depth barely changes: 0.32 le with no field vs 0.37 le at 7500 V/m after 15 s.
04 The interface lab
Run the stitch yourself
Pick a series, scrub the time the field has been on, then pull the layers apart. The scene places chains from a re-computation of the paper's diffusion–drift equation; the numbers and curves below are the paper's model results.
Predict first Which speeds the crossing more: tripling the field (2.5 → 7.5 kV/m) or using chains 10× shorter (100 → 10 kDa)?
Penetration 0.00 lₑ
Ionic bonds 0 mM (ideal, λ = 1)
Adhesion 0.0 kPa
Each series is shown as plotted in its own figure panel; compare values within a series. Bonds are the ideal count (every cation paired, λ = 1); the chain-length bond chart uses a log scale because 10 kDa chains reach ~80 mM.
05 Two ways out
Every bridging chain must let go of one side
Pull the layers apart and a chain straddling the interface has two exits. Route A : slide out of the e-GLUE, resisted only by water's drag along the length still in the gel. Route B : break its ionic bonds and slide out of the anionic network, resisted by drag on the short inserted part plus every bond. The chain takes whichever needs less force.
FA = 6πηv(lc − s) · FB = 6πηv·s + λ·m·s·fb/b η = 0.001 Pa·s · v = 1 mm/s · λ = 0.1 · m = 5 · fb = 10 pN · b = 3 nm
Where is the tipping point?
Setting FA = FB gives the equal-force point sp/lc = 1/(2 + λfb/[6πη(b/m)v]). With the paper's values the bond term is roughly 100 times the drag term (≈88 with these values), so a chain needs only about 1% of its length inside to prefer sliding out of the e-GLUE. In depth that is Hp = √(sp·b) ≈ 0.105 le.
No bonds (λ = 0): sp = 0.5 lc — both exits are equally hard.
Very fast pulling (v → ∞): drag dominates, sp → 0.5 lc, as if the chain were covalently bonded.
Very slow pulling (v → 0): drag vanishes, sp → 0: any inserted chain slides out of the e-GLUE rather than breaking bonds.
Hp/le = √(sp/lc) does not depend on chain length.
06 Where the strength comes from
Most of the grip comes from bonds breaking
For the baseline (100 kDa chains at 5000 V/m), the paper splits modelled adhesion into three parts. Breaking ionic bonds dominates. Sliding chains out of the e-GLUE adds a modest share that starts to matter after about 1 s. Sliding chains out of the anionic network is negligible: the inserted part is too short to cause much drag.
ionic bond breaking (route B chains)
chain slide-out of the e-GLUE (route A chains)
chain slide-out of the anionic network (≈0)
In the first ~0.7 s chains are shallower than about 0.1 le, so every bond formed would break on pulling. Later, bonds that form on chains deeper than Hp no longer break — those chains slide out of the e-GLUE instead. So the “bonds that break” curve levels off near 0.54 mM while the total keeps rising.
07 The bench test
Gel on gel, electrodes on both sides
To test the model, the authors built an e-GLUE from polyacrylamide gel with mobile PDAC (plus a little chitosan and sodium tripolyphosphate for handling). A polyacrylamide hydrogel with an anionic alginate network stood in for mucosa.
Predict Field series: the point at 6667 V/m is hidden. What adhesion did the bench measure there?
Caveat from the paper: the voltage split some water (electrolysis), making bubbles. Beyond about 6 s or about 4000 V/m, adhesion varied more — bubbles that reach the interface can open gaps and reduce bridging. Values are read from the published plot and are approximate.
08 Design trade-offs
Short chain: rapid but shallow. Long chain: slow but deep.
The model turns the e-GLUE's knobs into predictions. The paper's discussion uses them to sketch a design window — with a clear line between what was calculated or tested and what is proposed.
10 kDa chains build adhesion within ~1.5 s and then rise slowly; longer chains keep building and may overtake given more time. Raising cation concentration strengthens adhesion proportionally (10× cations → ~10× adhesion, Fig. 11d).
< 4000 V/m: weak electrophoresis — more than 6 s to reach tens of kPa.
MEASURED 4000–7000 V/m: about 10 kPa in 6 s on the bench.
> 20,000 V/m: strong currents risk heat and electrochemical reactions harmful to sensitive tissue.
Target from the discussion: the gut sees stresses in the kPa range under strong contraction, so tens of kPa gives a safety margin; hundreds of kPa may not be necessary.
(c) Ideas the authors propose — not tested here
Pulses above 20,000 V/m, with pauses that let heat dissipate, could keep chains moving while limiting heating.
Dissolve long-chain polycations near their solubility limit: more cations means more bonds, and long chains can bridge deeper.
Short chains for fast adhesion in the first seconds; long chains migrating deeper over time for retention.
Ideally chains would cross the ~100 nm mucus and epithelium layers to anchor in more stable tissue — but chains that large may be hard to find.
From the authors' earlier in vivo work, not this paper's tests: a transient activation (10–15 V for 60 s) produced minimal temperature rise and little to no observable inflammation. They also note that electrodes must stay aligned to avoid short circuits and local Joule heating.
09 What the model leaves out
Simplifications worth knowing
One chain lengthChains are modelled as all the same length (monodisperse). Real e-GLUE contains a spread of lengths (polydisperse), which would broaden the migration timing.
One average fieldThe model uses the average field across the stack rather than local, non-uniform fields.
Bonds smeared evenlyIonic bonds are spread uniformly along each chain through a single fraction λ, rather than at discrete random locations.
A gel, not tissueThe bench tests used a polyacrylamide–alginate hydrogel as a stand-in for mucosa, not living tissue.
Fitted on the same benchThe friction ξ0 and activation length L were fitted to the same bench data used to check the model (time series fitted; field series then predicted; chain series with L refitted).
BubblesWater electrolysis made bubbles; beyond ~6 s or ~4000 V/m, measured adhesion varied more.
10 Back to the stitch
Four knobs, now with predictions
Lowers chain friction, but deepens penetration only modestly: 0.32 → 0.37 le in 15 s.
Bonds and adhesion keep accumulating over 15 s; bonds are not yet saturated for 100 kDa chains.
The biggest lever: short chains bond fast but shallow; long chains bond slowly but can reach deeper.
Adhesion scales with it: about 10× more cations, about 10× more adhesion.
Long-lasting adhesion to wet, moving, self-renewing mucosa is wanted for sustained drug delivery, tissue sealing and gut-interfacing bioelectronics. That is the motivation for the e-GLUE. This paper does not test those uses. What it adds is the mechanics: a model, checked on the bench, of how a few seconds of field stitch two charged gels together and how each stitch lets go.
Source & credits
Ying B, Yu K-H, Yang S, Yang J. 2026 Mechanics of electroadhesion by polycation interfacial bridging: sticky electrophoresis, ionic complexation and chain entanglement. Proc. R. Soc. A 482: 20260579. https://doi.org/10.1098/rspa.2026.0579 · author-hosted PDF
Binbin Ying and Kun-Hao Yu contributed equally. Open access under the Creative Commons Attribution licence (CC BY 4.0). Earlier work mentioned: Ying et al., Sci. Transl. Med. 2025.
Model curves and measured values are re-plotted from the paper's figures 6, 7, 9, 10, 11 and 13 (values read from the published plots; approximate). All illustrations are original schematics drawn for this page; the molecular scene re-computes the paper's diffusion–drift equation to place chains and is not to scale.
SCHEMATIC explanatory drawing · MODEL · PAPER re-plotted from the paper's calculations · MODEL · EQUATION live evaluation of the paper's equations · MEASURED bench data · PROPOSED untested idea from the discussion
Topics in this iPaper
Engineering
materials science, mechanical engineering, chemical physics
Title page, Subject Areas, page 1
Physics
formulate a diffusion-drift model coupled with a Bell-like field-dependent chain friction
Section 3, A theoretical model, page 5
Chemistry
We determine the formation of ionic bonds by local availability of cations and anions
Abstract, page 1
electroadhesion
Mechanics of electroadhesion by polycation interfacial bridging
Title, page 1
electrophoresis
electrophoresis, ionic complexation, entanglement
Keywords, page 1
hydrogels
hydrogel with an anionic alginate network as the mucosa
Section 5, Experiments, pages 14-16
polyelectrolytes
An e-GLUE is a polymer network containing interpenetrating polycations
Abstract, page 1
reptation
The chain migrates in a reptation motion under combined diffusion and electrophoretic drift
Section 2, Model overview, page 4
mucoadhesion
Hydrogel adhesion to mucosa with prolonged retention is challenging
Section 1, Introduction, page 2