Text from the interactive explanation

Engineering 3D piezoelectric ZnO tetrapod interfaces for on-demand antibacterial skin-contact wearables

Parth Pandit, Soumitra Patekar, Mohammad Imran, Julia Bally, Yogendra Kumar Mishra, Ajay Pandey

Colloids and Surfaces B · open access

Wearable sensors sit on warm, sweaty skin for hours, and bacteria love it there. Parth Pandit and colleagues made a soft polymer film studded with microscopic zinc-oxide “jacks” that turns ordinary motion into bacteria-killing chemistry, with nothing designed to leak out of it.

Drag the patch up or down to bend it , or let it walk. Watch the crystal arms light up with charge and the green bacteria on the surface.

Below the scene: switch between a day of rest and walking, hold a bend with the slider, and read the live counts. Each chapter after this one unpacks one piece of what you are seeing, using the paper’s measurements.

Motion

Resting — bacteria settle and multiply

Voltage trace (illustrative): scaled so walking peaks sit near 3–3.5 V, the level the authors measured from the film worn on a forearm and a toe while walking (Fig. 7D/E). Held bends at the paper’s 5 mm radius stretch a 0.5 mm film’s surface by about 5 % (ε = t / 2R, §2.5).

Chapter 1

Warm, damp, and crowded

A wearable sensor works best pressed tight against skin. But that snug, warm, sweaty pocket is also a perfect home for bacteria. Over hours they settle, multiply and can grow into a slimy biofilm that covers the sensor’s working surface, adds noise to its signal and can irritate the skin underneath.

How fast does it happen? The authors stuck plain polymer films (PVDF, the same plastic used later, but without any crystals) on volunteers’ inner forearms, a dry spot, and on toes, a humid one. They peeled them off after 8, 16 and 24 hours of normal activity, washed each film and counted the bacterial colonies that grew from the wash.

Slide through the day, or switch to the toe.

The usual fix: make the surface poisonous

Many antibacterial coatings release silver, copper or other biocides. They work at first, but they stop once the supply is used up, they may disturb the skin’s own friendly microbes, and lingering low doses can encourage resistance.

Release-based coating concept sketch, not data

Silver, copper or biocide stored in the coating leaks away as the time slider advances. Protection lasts only as long as the reservoir.

Reservoir: full

Freshly sterilised film: no colonies yet.

read from plot Colony counts ≈ read from Fig. 7B/C (plain PVDF control films). The 0 h point is the freshly sterilised film at the start; counts were measured at 8, 16 and 24 h. Rods on the magnified film are drawn one per recovered colony.

What if a patch carried no reservoir at all, and the energy came from you?

What you just saw

Skin-worn films collect bacteria quickly: on the humid toe, a plain film already yielded ≈97 colonies after 8 hours, ≈120 after 24. Coatings that release a germ-killer only help while their supply lasts.

Chapter 2

Squeeze a crystal, get a voltage

Some crystals do something surprising: bend or squeeze them and a voltage appears. This is piezoelectricity (“piezo” is Greek for “press”). Zinc oxide is one of these crystals. Its atoms are arranged without a centre of symmetry, so when the crystal is deformed, the centres of positive and negative charge drift apart.

The film uses zinc oxide grown as tetrapods : four needle-like arms radiating from one core, about 110° apart, like a child’s toy jack, but microscopic (note the 50 µm scale bar below).

Push an arm. Drag sideways on the crystal, or use the slider. Watch strain pile up at the arm’s root, with + charge collecting on the stretched side and − on the compressed side. Then try the same push on a sphere or a rod.

Voltmeter: 0 % of the tetrapod’s full-push output

Illustrative model of the paper’s explanation Bend exaggerated. The bars compare peak strain for the same push in a simple lever picture. The measured tetrapod-versus-nanoparticle comparison comes in Chapter 4.

Why the shape matters

A long, thin arm acts like a lever. Push its tip and the strain gathers at the root, which is where the charge appears. A round nanoparticle squeezed by the same push shares the strain evenly and barely deforms, and a rod lying flat is stiff along its length. The paper describes ordinary nanorods as having a “relatively uniform stress distribution”, while the branched tetrapod “introduces localized stress concentration.”

There is a bonus: the four arms point in four different directions, so whichever way the film bends, some arm gets levered.

Inside the crystal sketch

Each zinc (grey) sits among four oxygens (red). Unsqueezed, the centres of + and − charge sit together. Squeeze the cell, using the same Push slider, and the average position of the negative charge slides away from the positive one, creating a tiny electric dipole . Billions of these add up to a voltage.

What you just saw

Bending zinc oxide separates electric charge. The tetrapod’s long arms act as levers that concentrate strain at their roots, so an ordinary bend produces stronger local charges than the same push on a compact particle.

Chapter 3

Choosing the recipe

How many tetrapods should go into the film? More crystals mean more piezo power, but zinc oxide in large amounts can harm skin cells. The authors cast films with 0, 0.5, 1, 2, 4 and 6 % tetrapods by weight and measured two things: the voltage a small device made from each film produced when pressed, and how many human skin cells (HaCaT keratinocytes) survived 24 hours in contact with it.

Before you look: which recipe would you put against skin?

Pick one, then explore the trade-off below.

Plain PVDF: no crystals

Simulation Film cross-section; crystal count scales with loading. Top: the air-facing surface, where arms poke out. Bottom: the side cast against the glass mould.

0 wt% (plain PVDF): peak device voltage ≈ 3.8 V · skin-cell viability 100 %

read from plot Voltage ≈ read from Fig. 2F. measured Viability as reported in §3.2.1 / Fig. S6.

Under the electron microscope

Top-view SEM, scale bars 100 µm. Left: the composite’s tangled tetrapod network with arms poking out; right: the smoother plain polymer. Fig. 2A/B, Pandit et al., CC BY 4.0.

The rough, crystal-studded surface matters twice: it concentrates stress around the tetrapods when the film flexes, and it makes the film strongly water-repellent (Chapter 7).

What you just saw

A trade-off: 4 wt% tetrapods gave the strongest electrical response in the authors’ tests while keeping about 72 % of skin cells alive. Adding more hurt the cells without adding power.

Chapter 4

From charge to chemistry

How does a voltage hurt a bacterium? At the crystal’s surface, the separated charges react with water and dissolved oxygen to make reactive oxygen species (ROS): the hydroxyl radical •OH and superoxide •O₂⁻ . The dot marks an unpaired electron, which makes them extremely reactive. They survive only briefly and attack whatever they meet: membranes, proteins, DNA.

You can’t see •OH directly, so the authors used a glow trick. A probe molecule, terephthalic acid (TPA), turns into a blue-glowing molecule when it meets •OH. Shine 315 nm light on it and it glows at 425 nm. More glow means more •OH was made.

Pick a sample and scrub through 48 hours of stimulation.

Compare

Stimulus

Stimulated tetrapod film · 0 h: TPA glow ≈ 0.20 a.u.

read from plot TPA fluorescence (arbitrary units) ≈ read from Fig. 4C/E (shaking) and Fig. 5C/E (ultrasound); n = 3 films per point in the paper. Vial glow is drawn in proportion to the value.

The pattern: the glow surges in the first 2–6 hours of stimulation and then fades. •OH doesn’t accumulate; it is made and quickly consumed. Still films, plain polymer and the probe alone stay low, and tetrapod films beat nanoparticle films at every time point under both shaking and ultrasound.

A second test, the DPPH assay, reads the broader “redox environment” around the film rather than •OH itself. For stimulated tetrapod films its signal changed within 2–6 hours and then stayed put, while the •OH glow faded (Fig. 4D/G, 5D/G). The authors read the two together as an early burst of radicals plus a longer-lasting change in the chemistry around the film.

What you just saw

Motion makes the film produce •OH in a burst within hours, far more than when it sits still, and more with tetrapods than with nanoparticles. Because •OH is so short-lived, it acts right where it is made: at the crystal surface.

Chapter 5

The shake test

Now the real question: does it kill? A 2 × 2 cm piece of film was fixed inside a tube of nutrient broth holding E. coli (a million cells per millilitre). The tube was shaken on an orbital shaker at 200 rpm, placed in a 40 kHz ultrasonic bath, or left still. At set times a drop of broth was spread on an agar plate: every surviving bacterium grows into a visible dot, a colony.

Predict: shaking or ultrasound — which kills more?

Make your call, then run both stimuli below.

Shaking · 0 h: bactericidal rate ≈ 0 % (start)

real plates Agar plates photographed by the authors (Fig. 4A shaking, Fig. 5A ultrasound), CC BY 4.0. read from plot Rates ≈ from Fig. 4B/5B: colonies from the tetrapod film compared with the control film, (N₁ − N₂)/N₁.

Shaking wins, and here’s the likely reason

Piezo voltage from the film measured

Device output under shaking at 200 rpm vs ultrasound (Fig. 2H/I).

Surface charge (zeta potential) measured

Measured on the film’s precursor dispersion before and after each stimulus (Fig. 6A).

Direct mechanical deformation produced roughly ten times the piezo voltage of ultrasound, and shifted the surface charge further: from −19 mV to +6 mV, versus −7.3 mV with ultrasound. The authors link the bigger charge shift to the stronger killing. In both cases the rates declined by 48 h: to ≈85 % with shaking and ≈76 % with ultrasound.

What you just saw

Stimulated tetrapod films killed nearly all E. coli within 2–6 hours, and the bactericidal rate was still about 98 % at 24 h with shaking and about 90 % with ultrasound. Left still, the same film let bacteria grow. That is the “on-demand” part.

Chapter 6

Who did it?

Was it really the tetrapods? Maybe bending alone squashes bacteria, maybe zinc oxide is toxic whatever its shape, maybe the plastic does it. The authors tested each suspect the same way: drop bacteria on the material, bend it 200 times in 2 minutes (≈1.7 bends per second, ≈5 mm radius, ≈5 % surface stretch), then look under the electron microscope.

Animation shown sped up; the protocol was 200 bends in 2 minutes (§2.5).

Pick a suspect, then bend it.

SEM images: Fig. 3C/D (tetrapod film, scale bar 10 µm), Fig. 6C (PVDF and nanoparticle films, 20 µm) and Fig. S7 (carbon cloth), Pandit et al., CC BY 4.0. Panel letters are from the original figures.

Is zinc simply leaking out?

measured Zn²⁺ in the broth by ICP-OES: control vs stimulated (Fig. S10, §3.6). Earlier studies found that below 10 ppm zinc does not significantly affect bacterial viability, and no inhibition “halo” formed around the films on agar (Fig. S9).

Bend the suspects to reach a verdict. The case closes once you have bent the tetrapod film.

The mechanism, assembled

Here is the opening patch again, with the proposed mechanism split into layers. Turn them on and off. The polarized surface also attracts bacteria electrostatically: its surface charge shifts with stimulation (Chapter 5).

Showing: charges, ROS, contact stress

Simulation · illustrative Layers change what is drawn, not the underlying behaviour; the paper proposes that these effects act together and does not split their individual shares.

What you just saw

Bending alone, the polymer alone and zinc oxide in a compact shape did not rupture bacteria. Only tetrapods in the film did, while releasing too little zinc to matter. Charge, short-lived ROS and stress at the tips of protruding arms are proposed to work together.

Chapter 7

On real skin

Finally, people. Healthy volunteers wore 2 × 2 cm films on the inner forearm (dry) and on a toe (humid) through 24 hours of walking, jogging and office work, with plain PVDF films alongside as controls. Films came off at 8, 16 and 24 hours. Bacteria were washed off each film and grown on plates.

Forearm · 8 h — plain PVDF: ≈7 colonies · tetrapod film: ≈0

dishes drawn to match the counts; read from plot counts ≈ from Fig. 7B/C.

Chapter 1’s plain-film curve, now with its partner: the tetrapod film.

These counts are culturable bacteria recovered from the film surface, not the total load on the skin underneath. Because the tetrapod film is strongly water-repellent, fewer bacteria may also stick to it in the first place. The authors call the result preliminary evidence .

The worn film generates its own signal

peak height from plot While walking, the film produced periodic peaks of about 3–3.5 V on the forearm and toe (Fig. 7D/E). Waveform drawn for illustration.

Water beads up

Movie S1 (Pandit et al., CC BY 4.0): a water droplet meets the tetrapod film. Cropped, muted and re-encoded.

measured Water contact angle: plain PVDF 116.36 ± 0.8°, tetrapod film 146.80° (Fig. 2C/D). The authors describe the tetrapod film as superhydrophobic: droplets stay nearly spherical instead of spreading.

Ten rounds, twenty days

The same film ran ten consecutive 48 h tests, washed and dried in between. At 24 h of every round it still killed more than 95 % of bacteria (Fig. 6D; curve shape ≈ from the plot).

What you just saw

On volunteers, plain films picked up a growing bacterial load, up to ≈120 recoverable colonies on the toe, while tetrapod films yielded almost none (0–3), which the authors treat as preliminary evidence. The film also repels water, makes its own voltage during walking, and kept working for ten reuse cycles.

Finale

Back to the patch

Remember the opening scene? Now you can read it. On a rest day the film is just a quiet surface, and bacteria land and multiply. On an active day every bend levers the tetrapod arms, charge separates, short-lived ROS spark at the tips, and membranes in contact give way. It works with no reservoir to run out, powered by the wearer.

Rest day · no bending: bacteria settle and multiply

Simulation · illustrative Same model as the opening patch; not to scale. Back to the opening patch ↑

The chain of evidence

Bend → chargeBending zinc oxide separates charge, and the tetrapod’s lever-like arms concentrate the strain.Ch. 2

•OH burstAfter 2 h of shaking the tetrapod film made a TPA glow of ≈1.97 a.u. versus ≈0.98 for the same film left still.Ch. 4

Ruptured membranesAfter 200 bends, membranes collapsed only on the tetrapod film, not on carbon cloth, plain PVDF or nanoparticle film.Ch. 6

Cleaner on skinWorn films yielded ≈0–3 recoverable colonies, against up to ≈120 for plain film (preliminary evidence).Ch. 7

The authors see the film as “a promising strategy for self-cleaning, skin-contact wearable and biomedical interfaces”: think wristbands, socks and sensor patches that keep themselves cleaner as you go about your day. In their study it killed more than 98 % of E. coli under mechanical stimulation and more than 90 % under ultrasound, kept above 95 % over 20 days of reuse, and released only 1.8–2.2 ppm zinc.

As compiled in the paper’s Table 1 from the cited studies, which used different bacteria and conditions.

Sources & credits

Parth Pandit, Soumitra Patekar, Mohammad Imran, Julia Bally, Yogendra Kumar Mishra, Ajay Pandey. “Engineering 3D piezoelectric ZnO tetrapod interfaces for on-demand antibacterial skin-contact wearables.” Colloids and Surfaces B: Biointerfaces 270 (2027) 116233. Available online 7 October 2026. Open access, CC BY 4.0. https://doi.org/10.1016/j.colsurfb.2026.116233

Figures, photographs and Movie S1 from Pandit et al. (2026/2027), licensed CC BY 4.0; cropped, masked and recompressed; the video was cropped, muted and re-encoded.

Values marked ≈ were read from published plots. Values marked measured are quoted from the text or figures.

The opening patch (and its returns in Chapters 5 and 6 and the finale), the single tetrapod, the magnified film surface in Chapter 1, the film cross-section, the lab bench, the bending strip and the colony dishes are illustrative simulations, not to scale.

3D rendering uses three.js (MIT License, © 2010–2026 three.js authors).