Charged Raindrops Punch Through Coatings Meant to Stop Rust

Electrically charged raindrops can bore through Teflon and rust the metal underneath, according to a Nature paper from the Max Planck Institute for Polymer Research in Mainz. Neutral drops in the same rig left the film smooth.

The group had spent years measuring that charge in hopes of turning rain into a trickle of electricity. The new tests show the same spark can open a path for rust.

How Charged Drops Punch Through Coatings

Electrically charged raindrops can force a protective film to fail by dielectric breakdown, then rust the metal. In the Mainz tests, drops that picked up 0.2 to 2 nanocoulombs on everyday surfaces pitted 60-nanometre Teflon on copper after 3,000 hits, while uncharged drops left the same film smooth.

Zhongyuan Ni, a PhD candidate at the institute, is first author of the study, charged drops induce coating corrosion, published on 26 August 2026. Hans-Jürgen Butt, a physicist at the same lab, proposed the work with Rüdiger Berger and supervised it. Chemical & Engineering News, which interviewed Ni, reported that the team first looked after only a few dozen drops and saw nothing. Ni then let the drops keep falling.

We found that even with small amounts of charge, they can still cause damage. It was quite surprising.

Zhongyuan Ni, PhD candidate, Max Planck Institute for Polymer Research, quoted by Chemical & Engineering News

Each lab drop was 35 microlitres of water with 1 mM sodium chloride, released every 12 seconds. For the charged runs, a drop slid about 4 cm down a surface tilted at 50 degrees, then fell 5 mm onto Teflon-coated copper tilted at 10 degrees so the water would run off. The Teflon was Teflon AF 1600, dip-coated onto 35-nanometre sputtered copper and baked at 160 °C under vacuum for 24 hours. After 3,000 charged hits, confocal images showed pits at the first contact point. Atomic force microscopy found holes several nanometres deep, deep enough to pass through the 60-nanometre film into the copper.

High-speed video caught the difference before the drop even landed. An uncharged drop kept a smooth belly. A charged drop stretched into a cone, the same shape seen in electrospray work when electric force beats surface tension. On PFOTS-coated quartz the drop arrived carrying +2.0 nanocoulombs. At impact, +1.8 nanocoulombs moved into the copper. The drop left with +0.016 nanocoulombs. The paper treats that dump of charge as the start of the damage. Ni told Chemical & Engineering News that after thousands of drops, the water looked as if it were pinning on the surface, and the copper under the microscope looked like scraped skin.

Nature research paper: Spontaneously charged water drops induce corrosionhttps://t.co/2fcxya0rsM

— nature (@Nature) August 26, 2026

Leaves, PVC and Fluorinated Glass Load the Drops

The charge does not come from the sky in these runs. It comes from the last few centimetres of slide. Water moving on an insulating solid picks up charge and leaves the opposite charge on the path, a process the lab calls slide electrification, analogous to rubbing a balloon on hair. The Nature abstract notes that the potential of drops charged this way can exceed 1 kV, and in some materials several thousand volts.

A 2024 Physics explainer of work by Aaron D. Ratschow and colleagues, who is also on the new paper, laid out charging needs a water-repelling surface. The coatings sold to shed rain are the surfaces that load rain with charge. That is the twist the wire copy left sitting on the bench.

PVC and polystyrene came from a hardware store. PFOTS, perfluorooctadecyltrichlorosilane, was grown on 1 mm quartz by chemical vapour deposition. The paper also used FEP film as a charging track. The threshold that worried the first readers of Nature’s announcement is not exotic. Drops at the low end of that 0.2 nanocoulomb range still marked the film.

The damage was not a Teflon quirk. Polystyrene films of 60 nm, 200 nm, 1 micrometre and 5 micrometres on copper all showed the same kind of scar after 3,000 charged hits. So did 200 nm polystyrene on gold and 60 nm silica on gold. Butt’s team wrote that charged drops hitting or sliding can set up strong fields, break the coat as an insulator, and then let the metal corrode. The 5 micrometre plastic is still a lab skin, not a ship’s paint, but it is almost a hundred times thicker than the 60 nm Teflon and it still failed.

Mainz Spent a Decade Chasing Rain Power

Slide charging was not discovered this week. The same department spent years trying to make it useful as a tiny generator. In 2019 Amy Stetten, then a postdoc with Butt and Stefan Weber, slid drops down hydrophobic glass, measured how charge grew with slide length, and built a model of fast charge laid down by each drop against slow leak-off behind it. The institute’s 19 December 2019 release said the group hoped to harvest electrical power from raindrops for low-power sensors in wet, isolated places, and to design surfaces that separate charge more effectively.

1. 21 November 2019: Soft Matter paper by Stetten, Golovko, Weber and Butt reports reproducible slide charging on PFOTS-treated glass, with drops going positive.
2. 19 December 2019: Max Planck Institute for Polymer Research pitches the effect as hydrovoltaics for small sensors in rainy sites.
3. 14 April 2022: Nature Physics paper led by Xiaomei Li, with Pravash Bista, Stetten, Berger, Weber and Butt, shows electrostatic drag slowing drops on insulating hydrophobic plates.
4. 31 May 2024: Physics covers Ratschow’s theory that the charging step requires a water-repelling surface.
5. 26 August 2026: Nature publishes the corrosion result, with Ni and Li running the drop-impact and slide experiments.

Scientific American, speaking to Butt on publication day, recorded him still circling a basic puzzle: why water is slow on a smooth plate when friction looks negligible. “For a liquid, there is no force which could break a bond,” Butt said. Physicists had not expected liquids to charge the way two solids do. He also said people have wanted to collect rain to create electricity, but that has not worked on a large scale. The 2026 paper is the same physics pointed at metal instead of at a sensor battery.

A Conductive Layer Left the Teflon Intact

The cleanest control was not a tougher polymer. It was a way to stop the drop charging in the first place. When PFOTS sat on a grounded indium tin oxide (ITO) coating on glass, the drop that left that plate did not form a cone as it neared the Teflon-copper. After 3,000 hits the Teflon looked as it started. The fluorinated chemistry was the same. The conductive pad underneath was not. Neutral drops from a grounded metal needle also left the sample unharmed.

Guangwen Zhou, a mechanical engineering professor at Binghamton University who was not on the paper, told Chemical & Engineering News by email that the electrical state of a drop, not only its chemistry, acidity or knock, can set where corrosion begins. He said makers could raise the dielectric strength of coats, or design surfaces that cut charging by changing chemistry, wettability or electrical properties of the path the drop travels. The ITO run is already one version of that second path: bleed the charge before the drop lands.

The paper is not a claim that this is the only way outdoor metal dies. Ni and colleagues told Chemical & Engineering News that sun, wind and other weather still count. The finding is that charge is an extra route that coating tests built around acid and abrasion would miss.

Cuprous Oxide After 50,000 Drop Hits

Three thousand hits opened nanometre pits. Keep going and the chemistry of the wound becomes ordinary rust, just arrived by an unusual door. After 50,000 charged impacts of 35 microlitre drops with 10 mM sodium chloride on 60 nm Teflon-coated copper foil, scanning electron microscopy showed scars more than 1 mm across. Energy-dispersive spectroscopy found less fluorine and carbon in the damaged patches, more oxygen, and chlorine in one of them.

PIT CHEMISTRY AFTER 50,000 HITS

• Fluorine and carbon: Both fall in the damaged zones, which matches loss of the Teflon film.
• Oxygen and chlorine: Both rise, pointing to oxides plus some chlorides rather than a clean cut.
• Area 2: Raman spectra match basic copper chloride.
• Area 3: Raman spectra match cuprous oxide, with X-ray diffraction showing the same two products on a copper background.

On 200 nm polystyrene, confocal fluorescence and nano-infrared maps found reaction products with strong infrared bands between 1,600 and 1,700 cm-1, consistent with C=C or C=O groups. Charged-drop hits also did more coating damage than sitting in the same salt bath. One Teflon-copper piece took 10,000 charged hits from drops that had slid on FEP. A twin soaked for 4 hours in 10 mM sodium chloride, about the total contact time of 50,000 drop hits, and the drops won. Methods list rainwater collected in the Mainz area among the liquids used, along with deionized water and salts such as potassium chloride, sodium bromide, potassium nitrate and zinc sulfate.

Paint Specs Still Ignore Drop Charge

The paper’s list of things that might suffer is broad: cultural heritage, buildings, ships, cars, other metal parts. Chemical & Engineering News used the Eiffel Tower as a colour example, noting that the 330 m iron structure is repainted by hand about every 7 years in a job that can run more than a year. That cycle is not a result from the Mainz rig. It is a reminder that outdoor metal is already on a short paint clock, and that clock was set without this mechanism on the spec sheet.

Preet Singh, a corrosion researcher at the Georgia Institute of Technology who was not involved, told Scientific American that infrastructure breakdown is a trillion-dollar problem and that better control would save money and lives. He said reliable products, from pharmaceuticals to petrochemical plant to implants, need tight hold on corrosion. The NACE International IMPACT study, still the figure most industry desks use, put the US$2.5 trillion global corrosion cost at 3.4% of world GDP on a 2013 basis, with US$375 billion to US$875 billion a year marked as recoverable if known controls were applied. Those numbers do not include this drop-charge path, because no one had isolated it.

WHAT WE KNOW

• Lab films: 60 nm Teflon and polystyrene from 60 nm up to 5 micrometres failed under charged drops, including coats on gold.
• Charge dump: About 1.8 nC of a 2.0 nC PFOTS-quartz drop moved into the copper at impact.
• Kill-switch: The same PFOTS chemistry on grounded ITO produced no cone and no pits after 3,000 hits.

WHAT IS UNCONFIRMED

• Factory paint: Automotive and marine coats are tens of micrometres thick, well above the 5 micrometre plastic that failed here, and were not in the Nature sample set.
• Real rain on a car: Lab drops were 35 microlitres on a 12-second clock; outdoor size, salt and rate still need field tests, as Zhou noted.
• Heritage share: How much of a monument’s paint loss is this mechanism versus acid, grit and sun is not measured.

The methods section lists 16 sample types, including a commercial super-hydrophobic spray on copper foil, which shows the lab was already looking past homemade Teflon. Further work, the authors wrote, has to find how to suppress this corrosion and how to build coats that can take a charged drop. Until those tests exist, a water-repelling finish can be read two ways at once: it sheds the drop, and it may be the reason the drop arrives armed.

After 50,000 hits the scars were millimetre-wide and packed with copper oxides. The same drop, sent across the same fluorinated chemistry on a grounded ITO pad, left the Teflon looking as it started.

Frequently Asked Questions

How Do Sliding Raindrops Pick Up Charge?

Stetten’s 2019 Soft Matter study found the charge on successive drops saturates as more drops pass, as the wait between them grows, and as the slide gets longer, because the solid has a limited storage capacity and slowly leaks the charge it holds. Drops on PFOTS glass went positive and left a negative trail, the same sign pattern the 2026 corrosion runs used.

Did the Tests Use Real Rainwater?

Yes. Alongside 1 mM and 10 mM sodium chloride, the methods used rainwater collected from the Mainz area, deionized water under 1 μS cm-1, and other salts including 10 mM potassium chloride, sodium bromide, potassium nitrate and zinc sulfate, plus 100 mM sodium chloride.

Did Gold Coatings Fail as Well as Copper?

They did in the extended-data set. A 200-nanometre polystyrene film on sputtered gold and a 60-nanometre silica film on gold both showed corrosion at the first contact patch after 3,000 charged drops that had slid on PFOTS-on-quartz, so the effect is not a copper-only oddity.

How Was the Lab Teflon Film Made?

The team sputtered 35 nanometres of copper onto smooth quartz, dip-coated a 60-nanometre Teflon AF 1600 film from a 1-wt% solution, and annealed the stack at 160 °C under vacuum for 24 hours. That is a model film for microscopy, not a sprayed car clear-coat.

Who Did the Drop Experiments in Mainz?

Ni and Xiaomei Li ran the impact and slide experiments and analysed the data. Ni, Li and Xiaoteng Zhou prepared the samples. Butt and Berger proposed and supervised the work at the Max Planck Institute for Polymer Research, with Aaron D. Ratschow among the co-authors.