Spontaneously charged water drops induce corrosion

Metal corrosion is a serious economic and safety issue facing modern society^17,18,19. Outdoor metal products are particularly vulnerable to water drops from rain, dew, ocean waves and melting snow^{20,21,22,23,24,25}. Using anticorrosion coatings is the simplest and most effective method; however, these coatings inevitably degrade over time^26,27. Conventional understanding attributes the coating failure induced by water drops mainly to two mechanisms: physical abrasion caused by moving drops^28,29,30 and chemical degradation from acidic substances and pollutants in the drops^31,32,33.

In recent years, it has been demonstrated that water drops moving over insulating solid surfaces can spontaneously acquire a charge and deposit an opposite charge to the surface^{1,2,3,4,5,6,7,8,9}. Hereby the electric potential of charged water drops can reach several thousand volts^14,15,16, depending on the specific materials involved and the thickness of the insulating layer. Here we show a previously overlooked corrosion mechanism induced by water drops: spontaneously charged water drops, whether hitting or sliding, can cause strong electric fields, leading to dielectric breakdown and cause degradation of the protective coating (polymer or oxide layer), thereby allowing subsequent corrosion of the metal substrate. Typically, ‘degradation’ is more commonly used to describe the damage to coatings. However, according to the International Union of Pure and Applied Chemistry (IUPAC), ‘corrosion’ also refers to materials other than metals, such as ceramics and polymers³⁴. In the following discussion, we use ‘corrosion’ to refer to the damage of both coatings and metals.

As a reference, we studied electrically neutral water drops directly falling onto a sample (Fig. 1a). 35-μl water drops were continuously released at 12-s intervals from a height of 4 cm and directly dropped onto a Teflon-coated copper sample. Usually, water drops in nature, such as rainwater, are not distilled water but contain some salt, whose concentration depends on local environmental conditions^35,36. In our experiments, the water drops contained 1 mM NaCl. We made the sample by sputtering a 35-nm-thick layer of copper onto a smooth quartz plate. Then a 60-nm-thick Teflon film was deposited by dip-coating onto the copper substrate from a 1-wt% Teflon AF 1600 solution. Finally, the samples were annealed at 160 °C under vacuum for 24 h. We chose the Teflon coating because it is chemically robust. The samples were tilted by 10° so that water drops ran off before the subsequent drops hit the surface. We analysed the sample surface by reflection mode confocal microscopy after 3,000 water drops had directly hit it. This analysis revealed flat and smooth surfaces. There was no notable difference in the morphology of the sample surface before or after the drop hit it.

To demonstrate that the charge caused by slide electrification in water drops induces corrosion, while keeping other conditions unchanged, we let water drops slide down four different 50° tilted surfaces before they fell onto the Teflon-coated copper sample. We selected four surfaces that are representative of many surfaces we encounter every day (Fig. 1b). A plant leaf: Tradescantia spathacea leaves (Fig. 1b (1)). A construction material: polyvinyl chloride (PVC) foam boards of 3 mm thickness (Fig. 1b (2)). A window: polystyrene (PS) glass of 2 mm thickness (Fig. 1b (3)). PVC and PS were purchased from a local hardware store. A widely used hydrophobic surface coating: perfluorooctadecyltrichlorosilane (PFOTS) coated on quartz (Fig. 1b (4)). The PFOTS was prepared on a 1-mm-thick quartz plate by chemical vapour deposition (CVD). Depending on the specific tilted surface, water drops acquired a charge between 0.2 and 2 nC after sliding for 4 cm (Extended Data Fig. 1). Then the water drops slid off the tilted surfaces and hit the Teflon-coated copper sample, dropping from a height of 5 mm. The reflection mode confocal microscopy images showed corrosion patterns in all four cases after the impact of 3,000 water drops (Fig. 1c). The corrosion pattern was located in the area of the drops’ initial contact with the sample surface. Atomic force microscopy (AFM) revealed that the charged water drop led to the hole-like defects on the sample surface (Fig. 1d). Both the Teflon film and the underlying copper substrate were corroded.

To gain better insight into the cause of corrosion, water drops hitting the Teflon-coated copper sample were imaged from the side with a high-speed camera. Electrically neutral drops showed no sign of deformation in the drop shape before contact with the sample surface; these drops either came out of a grounded metal needle (Fig. 1e and Supplementary Video 1) or had slid over a grounded PFOTS layer on indium tin oxide (ITO) (Extended Data Fig. 2). By contrast, when water drops charged by sliding over a PFOTS-on-quartz surface and then came close to the sample surface, the bottom of the drop elongated and formed a cone shape (Fig. 1f, Extended Data Fig. 3 and Supplementary Video 2). This behaviour is similar to the formation of a Taylor cone in the electrospraying process, in which the electrostatic force acting on the liquid surface overcomes the surface tension, thus causing deformation^37,38. We attribute this cone formation of the drop to the strong electric field generated when approaching the sample, which typically serves as an indicator of sample surface corrosion.

To demonstrate that the charge-induced corrosion is not restricted to Teflon-coated copper surfaces, PS films (192 kg mol⁻¹) of thickness 60 nm, 200 nm, 1 μm and 5 μm were prepared on copper, as well as a 200-nm-thick PS film on sputtered gold by dip-coating and a 60-nm-thick SiO₂ sputter-coated film on gold. Surface corrosion similar to that shown in Fig. 1c was observed at the initial contact area after 3,000 charged drops had hit (Extended Data Fig. 4). We conclude that corrosion induced by spontaneously charged water drops on metals coated with an insulating layer is a general phenomenon.

Given that drop cone formation is strongly correlated with discharge phenomena, we placed a current amplifier at three distinct positions and monitored the respective charge changes (Fig. 2). A drop sliding 4 cm down a tilted PFOTS-on-quartz surface acquired a charge of +2.0 ± 0.021 nC (position 1, Fig. 2 inset blue bar and Extended Data Fig. 5a). When the drop hit the 60-nm Teflon-on-copper surface, most of the charge was transferred into the copper layer, +1.8 ± 0.022 nC, (position 2, Fig. 2 inset green bar and Extended Data Fig. 5b). At position 3, after leaving the Teflon-on-copper surface, the drop retained only +0.016 ± 0.003 nC (Fig. 2 inset orange bar and Extended Data Fig. 5c). These results indicate that the highest charge is exchanged between the drop and copper substrate at the moment of impact. We conclude that corrosion of the sample is caused by the discharge of water drops.

We suggest that discharge is caused by dielectric breakdown of the insulating coating when the water drop approaches the sample surface. When the electric field strength between two electrodes exceeds the dielectric strength of the medium, dielectric breakdown occurs. Charge is transferred from one electrode to another through the medium. In our case, the water drop and the metal substrate, respectively, act as two electrodes (Fig. 3a). The water drop accumulates a positive charge through slide electrification. In the metal substrate, an equivalent amount of negative image charge is induced. Together, the positive charge in the water drop and the negative charge in the metal substrate create an electric field. As the drop approaches the surface, the electric field increases. To estimate the electric field during approach, the water drop was assumed to act as a conductive sphere of radius R opposite to a conductive wall formed by the metal. Then the capacitance of the water drop changes with distance as C = 4πεε₀R(1 + 0.5ln(1 + R/h)) (ref. ³⁹). Here, ε is the relative permittivity of the intervening medium, ε₀ is the vacuum permittivity and h is the distance from the surface of the metal substrate to the bottom of the water drop. In the experiment, R is 2.0 mm. When the distance is large, the capacitance should be calculated using the relative permittivity of air; at impact, it is more appropriate to use the relative permittivity of the coating (Fig. 3b). According to U = Q/C (U: voltage between the water drop and metal, Q: charge of the water drop), the change in electric field strength can be estimated, E = U/h, as the water drop approaches the surface (Fig. 3c). The charged water drop leads to breakdown of up to a film thickness of about 10 μm for Teflon or about 50 μm for PS. This corresponds to dielectric strengths of 60 kV mm⁻¹ and 19 kV mm⁻¹, respectively. Thus, the calculations demonstrate that a water drop carrying approximately nC charge can break down most micrometre-thick insulating coatings (Extended Data Fig. 6). On the basis of Joule heating effect estimation, the damage caused to the Teflon film by a single charged drop is in the sub-μm range (Supplementary Information Discussion 1). The accumulation of Joule heating-induced localized damages (Extended Data Fig. 7) led to metal exposure and increased corrosion risk.

Our estimation of electric field strength represents a lower limit. Before the water drop touches the sample, it forms a conical structure at the bottom (Fig. 1f). This deformation would lead to a lower effective radius of curvature and thus a locally higher electric field. A more complete analysis of dielectric breakdown needs to take into consideration the coating thickness. Rather than assuming a constant dielectric strength, in reality, the dielectric breakdown strength of a coating is known to depend on the thickness⁴⁰.

As well as morphological damage, dielectric breakdown can cause chemical changes within the coating, resulting in new substances forming. Confocal microscopy images of a 200-nm-thick PS film on copper after 3,000 water drop impacts (Fig. 4a) showed the appearance of rough regions on the smooth PS surface in reflection mode. The fluorescence mode further revealed that some rough regions emitted fluorescence on 488-nm excitation, with broad emission spectra covering 520–680 nm (Extended Data Fig. 8). PS contains benzene rings and does not form a conjugated π-electron cloud system. Therefore the observed fluorescence does not originate from the PS but from the products of a reaction that have such a π-electron cloud system.

AFM-based nano-infrared (IR) analysis provided further characterization of the reaction products in the PS films^41,42. The topography image showed a roughened region corresponding to reaction products surrounded by PS (Fig. 4b). Nano-IR imaging at 1,495 cm⁻¹, corresponding to the aromatic C=C stretching of PS, showed lower absorption in the roughened area than in the surroundings, whereas imaging at 1,641 cm⁻¹ produced an inversion of the contrast, indicating chemical bonds not present in PS. IR spectra recorded from the roughened and flat areas (Fig. 4c) both exhibited the characteristic PS peaks at 1,458 cm⁻¹ and 1,495 cm⁻¹, corresponding to the aromatic C=C stretching of PS (ref. ⁴²). However, only the roughened area showed a distinct broad IR absorption band between 1,600 cm⁻¹ and 1,700 cm⁻¹, a range associated with the double-bond stretching vibration of C=C and carbonyls C=O (refs. ^43,44). These two types of double bond and the benzene rings possibly form a conjugated π-electron cloud system, thus having fluorescent properties.

To demonstrate corrosion under broader conditions, we used commercial copper foil instead of the sputtered copper on quartz and coated with a 60-nm-thick Teflon film by dip-coating. Charged drops damaged the coating through dielectric breakdown. Electrochemical impedance spectroscopy (EIS) results consistently demonstrated that the initially excellent barrier properties of the coating were degraded after 10,000 charged-drop impacts (Extended Data Fig. 9 and Extended Data Table 1). Under the same salt concentration, charged-drop impacts caused much more coating damage than continuous immersion in the bulk solution. After approximately 50,000 charged drops had hit the surface, corrosion of more than 1 mm in size occurred on the surface (Fig. 4d). Scanning electron microscopy with energy-dispersive X-ray spectroscopy (SEM-EDS) elemental analysis showed that, relative to area 1, areas 2 and 3 showed substantially increased oxygen contents and decreased fluorine and carbon contents, with chlorine also being detected in area 2 (right of Fig. 4d). The decrease in fluorine and carbon indicated damage to the Teflon film in areas 2 and 3, whereas the increase in oxygen suggested oxide-dominated corrosion products and the chlorine signal in area 2 implied the presence of chlorides.

To further analyse the corrosion products, confocal Raman spectra collected over areas 2 and 3 (Fig. 4e) identify basic cupric chloride (Cu₂(OH)₃Cl)⁴⁵ and cuprous oxide (Cu₂O)⁴⁶ as main corrosion products in area 2 and area 3, respectively. X-ray diffraction (XRD) analysis of the samples (Fig. 4f) revealed dominant diffraction peaks that correspond to metallic copper, reflecting that the corrosion products constitute a small fraction of the overall copper foil sample. But minor diffraction peaks corresponding to basic cupric chloride (Cu₂(OH)₃Cl) and cuprous oxide (Cu₂O) are also observed, consistent with the Raman results.

The above results reveal the occurrence of corrosion phenomena. Corrosion is initiated by localized dielectric breakdown of the coating induced by charged drops. As a result, the underlying metal becomes exposed to the aqueous drop environment. As more charged drops hit the surface, the defective area gradually grows (Extended Data Fig. 7). The change in wettability in the defective area promotes drop retention and enhances direct contact with the metal surface, resulting in a localized potential difference at the liquid–solid interface that drives the formation of micro-galvanic cells and subsequent electrochemical corrosion. Also, charged drops can further influence the corrosion process by inducing external electric fields that alter the local electrode polarization state and promote the formation of strongly oxidative species inside drops (detail shown in Supplementary Information Discussion 2). The corrosion mechanism of charged drops is universally applicable. We carried out a series of experiments to validate this, involving variations in drop ionic composition, ambient oxygen (Supplementary Figs. 5–14) and metal substrate (surface blistering on aluminium; Supplementary Information Discussion 3).

In the experiments described above, there were two separate stages: charging of the water drop and subsequent corrosion, which happened on two different surfaces. Corrosion caused by slide electrification can also occur on a single sample consisting of a structured substrate composed of an electrical insulator and a conductor. A rectangular mask was used to cover two-thirds of a quartz plate and a 35-nm-thick copper film was sputtered onto the uncovered area. Then a 60-nm-thick Teflon film was dip-coated onto the entire substrate surface, followed by annealing at 160 °C for 24 h in vacuum. This method produced a flat, smooth and chemically uniform Teflon surface on a hidden structured substrate.

When a water drop slid down from the top, a charge was generated in the Teflon-on-quartz area. Then the drop moved over the Teflon-on-copper area (Fig. 5a). After 3,000 water drops, a corrosion pattern appeared along the quartz–copper boundary (Fig. 5b). AFM images revealed the formation of a trench-like defect (Fig. 5c). No defects were observed after electrically neutral drops slid (Supplementary Fig. 17). In the case of the sliding drop, when the advancing contact line of the water drop reached the boundary to the hidden copper layer, a conical liquid meniscus formed and jumped ahead, leading to a local discharge of the drop (Fig. 5d). Using a PS film instead of Teflon leads to fluorescence in the boundary area (Fig. 5e). The corrosion caused by the sliding water drops was always located in the boundary area, indicating that water drop impact is not necessarily required for corrosion to occur; sliding alone can cause corrosion.

We predict that similar corrosion effects on heterogeneous or composite materials occur, in which a higher dielectric permittivity or conductive material is present in an insulating material with low dielectric permittivity as matrix (Supplementary Fig. 18). A drop sliding over such material would discharge when the advancing contact line reaches those parts with high dielectric permittivity, even if they are hidden underneath a coating.

Natural and commercial materials

Tradescantia spathacea leaves were obtained from potted plants in our office. PVC foam boards (PVC-Schaumplatte Fixmaß, Bauhaus AG, 250 × 500 × 3 mm³) and transparent PS glass (Hobbyglas Owocor, Bauhaus AG, 250 × 500 × 2 mm³) were purchased from a local hardware store. Fluorinated ethylene propylene (FEP) film (Nenull, 140 × 200 × 0.15 mm³), copper foil tape (3M 1181 Kupferband leitend, 0.04 mm, 50 mm × 16.5 m), super-hydrophobic spray (Glaco Mirror Coat Zero, 40 ml) and PS sheets (Evergreen Scale Models, 130 μm) were purchased from an online shop.

Sample preparation

Sixteen sample types were prepared:

1. (1)

   PFOTS layers on quartz. Quartz plates (75  × 25 × 1 mm³, proQuarz GmbH) were treated with O₂ plasma at 300 W for 10 min (Femto low-pressure plasma system, Diener electronic). Using CVD, 1H,1H,2H,2H-PFOTS (97%; Sigma-Aldrich) was then coated on the surface of the quartz plate. The cleaned quartz plates were subsequently placed in a vacuum glass container together with a vial containing 0.5 ml PFOTS. Then we evacuated the container to a pressure below 100 mbar, closed the pump and let it react for about 30 min. Finally, the quartz plates were rinsed with ethanol to remove any unbound silane molecules.

2. (2)

   PFOTS layers on ITO glass. ITO glass (75 × 25 × 1.1 mm³, surface resistivity 30–60 Ω sq⁻¹, Sigma-Aldrich) were treated with O₂ plasma at 300 W for 10 min (Femto low-pressure plasma system, Diener electronic). Using CVD, 1H,1H,2H,2H-PFOTS (97%; Sigma-Aldrich) was then coated on the surface of the ITO glass. The cleaned ITO glass was subsequently placed in a vacuum glass container together with a vial containing 0.5 ml PFOTS. Then we evacuated the container to a pressure below 100 mbar, closed the pump and let it react for about 30 min. Finally, the ITO glass was rinsed with ethanol to remove any unbound silane molecules.

3. (3)

   60-nm Teflon-coated copper. To obtain flat copper substrates, a 35-nm-thick copper layer was sputtered onto a quartz plate. A 60-nm-thick Teflon film was then coated onto the copper by dip-coating with a pulling speed of 10 mm min⁻¹ from 1 wt% Teflon AF 1600 (ε_r = 1.9, Sigma-Aldrich) in FC-75 (97%, Fisher Scientific). Finally, the Teflon-coated copper samples were heated in an oven at 160 °C under vacuum for 24 h. The film thickness was measured by a profiler (P-7 stylus profiler, KLA-Tencor).

4. (4)

   A 60-nm-thick PS film was coated onto the copper by dip-coating with a pulling speed of 60 mm min⁻¹ from 2 wt% PS (molecular weight 192 kg mol⁻¹, ε_r = 2.6; Sigma-Aldrich) in toluene (99.8%, Sigma-Aldrich). The sample was then heated in an oven at 120 °C under vacuum for 24 h.

5. (5)

   A 200-nm-thick PS film was coated onto copper by dip-coating with a pulling speed of 40 mm min⁻¹ from 4 wt% PS in toluene. The samples were then annealed (120 °C under vacuum for 24 h).

6. (6)

   A 1-μm-thick PS film was coated onto copper by dip-coating with a pulling speed of 20 mm min⁻¹ from 10 wt% PS in toluene. The samples were then annealed (120 °C under vacuum for 24 h).

7. (7)

   A 5-μm-thick PS film was coated onto copper by dip-coating with a pulling speed of 10 mm min⁻¹ from 20 wt% PS in toluene. The samples were then annealed (120 °C under vacuum for 24 h).

8. (8)

   200-nm PS-coated gold samples. A 35-nm gold layer was sputtered onto quartz plates. A 200-nm PS film was then coated onto the gold layer by dip-coating, as described previously.

9. (9)

   60-nm SiO₂-coated gold samples. First, a 35-nm gold layer was sputtered onto quartz plates. Then a 60-nm SiO₂ layer was sputtered onto a gold layer.

10. (10)

    60-nm Teflon-coated aluminium sample. First, a 100-nm aluminium layer was deposited onto glass slides (76.2 × 25.4 × 1 mm³, Sail Brand) by vacuum thermal evaporation. Then a 60-nm Teflon film was coated on the aluminium layer by dip-coating with a pulling speed of 10 mm min⁻¹ from a solution of 1 wt% Teflon AF 1600 (ε_r = 1.9, Sigma-Aldrich) in FC-75 (97%, Fisher Scientific). Finally, the samples were annealed (160 °C under vacuum for 24 h).

11. (11)

    PFOTS layers on aluminium. First, a 100-nm aluminium layer was deposited onto glass slides. The sample was then treated with O₂ plasma at 300 W for 10 min and PFOTS was deposited by CVD, as described previously.

12. (12)

    60-nm Teflon-coated quartz–copper samples (for sliding experiments). We used a rectangular mask to cover two-thirds of the quartz plate (75 × 25 × 1 mm³, proQuarz GmbH) and 35 nm of copper was sputtered onto the uncovered area. After removing the mask, we obtained a plate with one-third covered by Cu and two-thirds comprising the quartz surface. The boundary between the quartz and copper areas formed a sharp straight line. Finally, a 60-nm Teflon film was coated on this plate by dip-coating with a pulling speed of 10 mm min⁻¹ from 1 wt% Teflon AF 1600 (ε_r = 1.9, Sigma-Aldrich) in FC-75 (97%, Fisher Scientific). Finally, the samples were annealed (160 °C under vacuum for 24 h).

13. (13)

    200-nm PS-coated quartz–gold samples (for sliding experiments). We used the same method as above to make a substrate of two-thirds quartz and one-third 35-nm gold. A 200-nm PS film was then coated on this plate by dip-coating with a pulling speed of 40 mm min⁻¹ from 4 wt% PS in toluene. Then the samples were annealed (120 °C under vacuum for 24 h).

14. (14)

    60-nm Teflon-coated commercial copper foil samples. A 40-μm-thick single-sided copper foil tape (3M 1181 Kupferband leitend, 0.04 mm, 50 mm × 16.5 m) was cut into dimensions of 50 mm × 25 mm, stacked face to face on the adhesive side and cleaned with ethanol. A 60-nm-thick Teflon film was then coated onto the copper by dip-coating with a pulling speed of 10 mm  min⁻¹ from 1 wt% Teflon AF 1600 (ε_r = 1.9, Sigma-Aldrich) in FC-75 (97%, Fisher Scientific). Finally, the samples were heated in an oven at 160 °C under vacuum for 24 h.

15. (15)

    500-nm super-hydrophobic coating commercial Copper foil samples. A 40-μm-thick single-sided copper foil tape (3M 1181 Kupferband leitend, 0.04 mm, 50 mm × 16.5 m) was cut into dimensions of 50 mm × 25 mm, stacked face to face on the adhesive side and cleaned with ethanol. A commercial super-hydrophobic spray (Glaco Mirror Coat Zero) was applied to copper foil from a distance of approximately 0.5 m for 3–4 s. The sample was then left undisturbed for 30 min to allow the ethanol in the solution to evaporate.

16. (16)

    A 12-μm-thick PS film was coated onto copper by dip-coating with a pulling speed of 10 mm min⁻¹ from 30 wt% PS in toluene. The samples were then annealed (120 °C under vacuum for 24 h).

Water drop impact experiments

We generated water drops with a volume of 35 μl water (<1 μS cm⁻¹; Gibco, Thermo Fisher Scientific) containing 1 mM NaCl (diluted from 1 M NaCl solution, Carl Roth) through a grounded syringe needle (electrically neutral). The needle was connected to a peristaltic pump (MINIPULS 3, Gilson). Water drops were released from a height of about 4 cm from the sample surface at intervals of 12 s. More than 1,000 water drops were continuously released in each experiment. The sample was tilted by 10° so that, after impact, the water drop rolled off the surface. Each water drop always showed the same behaviour (Supplementary Fig. 19).

In the experiments of the spontaneously charged water drops that induce corrosion, the water drops first fell onto the top of a 50° tilted surface (Tradescantia spathacea leaves, PVC foam board, PS glass, FEP film or PFOTS-on-quartz plate) at intervals of 12 s from a release height of about 5 mm. Driven by gravity, water drops slid down roughly 4 cm and then left the tilted surface. During this process, the water drops undergo slide electrification, causing them to accumulate charge and deposit opposite charge onto the tilted surface. To achieve defined electrical conditions, the tilted surfaces were placed on a grounded metal plate. After the water drops left the tilted surface, they carried the electrical charge and, after falling about 5 mm, hit the Teflon-coated copper or other samples. A similar discharge cone was observed at different falling heights (Supplementary Fig. 20). As well as the 1 mM NaCl, there were more salt solutions (10 mM NaCl (Carl Roth), 100 mM NaCl (Carl Roth), 10 mM KCl (Carl Roth), 10 mM NaBr (Carl Roth), 10 mM KNO₃ (Carl Roth), 10 mM ZnSO₄ (Fluka) and also deionized water (<1 μS cm⁻¹; Gibco, Thermo Fisher Scientific) and rainwater collected from the Mainz area.

The charge deposited by the drops on the tilted surface reduced the slide electrification effect of subsequent drops. Because it takes several minutes or more for the surface to naturally return to electrical neutrality, we needed to speed up the experimental process. To do this, we used an ionizing air blower (Aerostat PC ionizing air blower, Simco-Ion), which continuously neutralized the charge on the tilted surfaces during the experiment. This ensured that each subsequent water drop had the same slide electrification effect. In practice, tilted surfaces were neutralized in the 12 s between subsequent drops. Similarly, more than 1,000 water drops were continuously released in each experiment. Surface neutralization created by the ionized air took several seconds and was negligible while drops were sliding (about 100 ms) but effective in the intervals between drops. We also tested the case without the ionizing air blower and still observed surface corrosion (Supplementary Fig. 21).

For both experiments, we used a side-view high-speed camera (Photron, FASTCAM MINI UX100, 25,000 fps, resolution 1,280 × 200, with 1× SilverTL Telecentric Lens, Edmund Optics) to observe the behaviour of water drops when they hit the Teflon-coated copper surface. For the experiment observing the evolution of surface corrosion patterns from the bottom in situ, we used a non-high-speed camera (FLIR Blackfly S) for recording.

Water drop sliding experiments

The samples for the sliding experiments were placed on a grounded metal plate tilted at 50°. 35 μl water drops containing 1 mM NaCl were released onto the top of a tilted sample from a grounded syringe needle at intervals of 12 s, with a release height of about 5 mm. We used a peristaltic pump (MINIPULS 3, Gilson) to continuously deliver solution from the tank to the syringe. During the experiment, the ionizing air blower (Aerostat PC ionizing air blower, Simco-Ion) was always turned on. In each experiment, about 3,000 water drops were released continuously. We used a bottom-view high-speed camera (Photron, Phantom TMX 7510, 40,000 fps, resolution 1,280 × 800, with 10× UPlanSApo Microscope Objective, Olympus) to observe the behaviour of the water drops when they slid down tilted samples.

Confocal laser scanning microscopy imaging

Surface morphology and fluorescent properties of samples were imaged using a confocal microscope (LSM 880, Carl Zeiss) equipped with a Zeiss Plan-Apochromat 10×/0.45 objective. An argon-ion laser (λ = 488 nm) coupled with an optical fibre to the microscope was used for excitation. The confocal observation volume was positioned on the surfaces studied. Images were recorded either in reflection or in fluorescence mode using an appropriate dichroic mirror and a spectral detection unit (Quasar, Carl Zeiss). This unit comprises a diffraction grating and a 32-channel GaAsP multianode photomultiplier array detector. While recording the fluorescence images, the maximum spectral range of the detected emission light was 517–696 nm. The emission spectra were recorded using the so-called lambda mode of the instrument. In this mode, the spectral range of 520–690 nm was distributed between the 32 channels and the signal in every channel was detected separately. The 3D image of the surface morphology (Supplementary Figs. 15 and 23) was obtained by another confocal microscope (confocal white light microscope, µsurf, NanoFocus AG) equipped with a 20×/0.46 objective, using point-by-point line scanning laser profilometry.

Charge measurements

Charge measurements were conducted using a current amplifier (response time: 0.8 ms, FEMTO DDPCA-300). We recorded the current signal of water drops or metal substrates during impact using a National Instruments data acquisition board (USB-6366 X Series).

Measurement of water drops: the current amplifier was connected to wires with a diameter of 0.7 mm made of tungsten serving as electrodes. The tungsten wire was coated with a layer of gold to enhance electrical conductivity. We measured the current of drops after sliding 4 cm on the surface (Extended Data Figs. 1 and 5a). Also, we measured the current of drops bouncing off the surface (Extended Data Fig. 5c). In both cases, the water drops touched the electrode and caused a current flow. The contact to the wire electrode corresponds to 0 ms (t₀) and the drop discharged until t₁. After t₁, there is a low positive current of about 10 nA until the rear of the drop detaches from the electrode after roughly 20 ms. This low positive current is the result of the continued deposition of negative charge on the surface as the drop moves while still in contact with the electrode. For all samples except for the Tradescantia spathacea leaves, t₁ was typically 2 ms. For the leaves, t₁ was about 10 ms. The integral of the current from t₀ to t₁ is the amount of charge generated by slide electrification. We also measured the drop current and charge under different salt concentrations, release intervals and natural rainwater (Supplementary Figs. 4, 21 and 22).

Measurement of the current of metal substrates: here the current amplifier was connected directly to the metal layer. When the water drop hit the surface, a current signal was detected as well (Extended Data Fig. 5b). The contact of the drop onto the surface corresponded to 0 ms and the drop discharged until t_1, which was about 2 ms. The integral of the current from t₀ to t₁ is the amount of charge transferred from the water drop to the metal.

AFM and AFM-IR measurements

Details of the surface morphology were studied using AFM (Dimension Icon, Bruker) in tapping mode. The developed surfaces were uniform and smooth (Supplementary Figs. 23 and 24). Commercial Cu foil had a rough surface and its roughness was slightly reduced after coating with a Teflon film (Supplementary Fig. 25). The cantilever had a nominal resonance frequency of 300 kHz and a spring constant of 26 N m⁻¹ (OTESPA, OPUS). The scan sizes were 0.5 × 0.5, 20 × 20 or 50 × 50 μm².

The products resulting from dielectric breakdown on the PS film were measured using an AFM-based nano-IR method (Vista One, Molecular Vista). The nano-IR response was recorded in photo-induced force microscopy (PiFM) mode. To obtain images of the surface topography, we excited the second eigenmode of the cantilever resonance frequency and kept the oscillation amplitude constant, using an electronic feedback circuit. The first eigenmode of the cantilever resonance frequency was used to record the nano-IR response. The incident IR light was modulated using the difference frequency between the first and second eigenmodes of the cantilever. This difference frequency was tuned to a maximum response amplitude at the first eigenmode.

SEM-EDS measurements

The determinations by SEM were performed with a HITACHI SU8000 (Hitachi High-Technologies Europe GmbH). The scanning electron microscope was coupled to an XFlash 5010 detector, an X-ray detector that allows simultaneous EDS-based elemental analyses.

Raman measurements

Raman measurements were performed using a WITec confocal Raman spectrometer (alpha300 R, 10× objective, 600/1,200 grooves mm⁻¹ grating, 5 mW) with a 532-nm excitation and 120 s integration. The pristine copper foil was included as a reference sample (Supplementary Fig. 26a).

XRD measurements

XRD measurements were performed using a Rigaku SmartLab diffractometer with a rotating Cu anode (8 keV, λ = 1.5406 Å), a Kβ filter and a HyPix-3000 detector. The sample was placed on a rotating sample stage and measured in θ–θ geometry in the range 10° < 2θ < 80° at a rate of 1° min⁻¹. The X-ray beam has a width of 1 mm. The pristine copper foil was included as a reference sample (Supplementary Fig. 26b).

EIS measurements

EIS was carried out using a Metrohm Autolab N series potentiostat (Autolab PGSTAT204) in a three-electrode configuration. Samples were mounted horizontally in a flat-cell configuration, with the exposed working area defined by an O-ring of diameter 6 mm. An Ag/AgCl electrode (3 M KCl) and a platinum wire were used as the reference and counter electrodes, respectively. The electrolyte was either 10 mM or 600 mM aqueous NaCl and measurements were performed in air at ambient temperature. Impedance spectra were recorded at open-circuit potential over a frequency range of 10⁵ Hz to 0.1 Hz using a sinusoidal perturbation of 10 mV. For 5-μm PS films, the perturbation amplitude was increased to 20 mV to improve the signal-to-noise ratio for these highly resistive samples. The EIS results demonstrated that the developed coatings greatly increase the interfacial resistance, indicating their excellent barrier properties (Supplementary Information Discussion 4). We also used a digital multimeter to test the barrier properties of the coating (Supplementary Fig. 28).

The data supporting the findings of this study are available in the paper, the Extended Data and the Supplementary Information. Source data are provided with this paper.

1. Xu, W. et al. A droplet-based electricity generator with high instantaneous power density. Nature 578, 392–396 (2020).

   Article  ADS  CAS  PubMed  Google Scholar 

2. Li, X. et al. Spontaneous charging affects the motion of sliding drops. Nat. Phys. 18, 713–719 (2022).

   Article  CAS  Google Scholar 

3. Yin, J. et al. Generating electricity by moving a droplet of ionic liquid along graphene. Nat. Nanotechnol. 9, 378–383 (2014).

   Article  ADS  CAS  PubMed  Google Scholar 

4. Sun, Q. et al. Surface charge printing for programmed droplet transport. Nat. Mater. 18, 936–941 (2019).

   Article  CAS  PubMed  Google Scholar 

5. Armiento, S., Filippeschi, C., Meder, F. & Mazzolai, B. Liquid-solid contact electrification when water droplets hit living plant leaves. Commun. Mater. 3, 79 (2022).

   Article  Google Scholar 

6. Zhang, N. et al. A universal single electrode droplet-based electricity generator (SE-DEG) for water kinetic energy harvesting. Nano Energy 82, 105735 (2021).

   Article  CAS  Google Scholar 

7. Wu, H., Mendel, N., van den Ende, D., Zhou, G. & Mugele, F. Energy harvesting from drops impacting onto charged surfaces. Phys. Rev. Lett. 125, 078301 (2020).

   Article  ADS  CAS  PubMed  Google Scholar 

8. Kwon, S.-H. et al. An effective energy harvesting method from a natural water motion active transducer. Energy Environ. Sci. 7, 3279–3283 (2014).

   Article  CAS  Google Scholar 

9. Helseth, L. E. & Guo, X. D. Fluorinated ethylene propylene thin film for water droplet energy harvesting. Renew. Energy 99, 845–851 (2016).

   Article  CAS  Google Scholar 

10. Lacks, D. J. & Shinbrot, T. Long-standing and unresolved issues in triboelectric charging. Nat. Rev. Chem. 3, 465–476 (2019).

    Article  CAS  Google Scholar 

11. Lin, S., Chen, X. & Wang, Z. L. Contact electrification at the liquid–solid interface. Chem. Rev. 122, 5209–5232 (2022).

    Article  CAS  PubMed  Google Scholar 

12. Stetten, A. Z., Golovko, D. S., Weber, S. A. L. & Butt, H.-J. Slide electrification: charging of surfaces by moving water drops. Soft Matter 15, 8667–8679 (2019).

    Article  ADS  CAS  PubMed  Google Scholar 

13. Yatsuzuka, K., Mizuno, Y. & Asano, K. Electrification phenomena of pure water droplets dripping and sliding on a polymer surface. J. Electrost. 32, 157–171 (1994).

    Article  Google Scholar 

14. Li, X., Ratschow, A. D., Hardt, S. & Butt, H.-J. Surface charge deposition by moving drops reduces contact angles. Phys. Rev. Lett. 131, 228201 (2023).

    Article  ADS  CAS  PubMed  Google Scholar 

15. Li, L. et al. Sparking potential over 1200 V by a falling water droplet. Sci. Adv. 9, eadi2993 (2023).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

16. Bista, P., Ratschow, A. D., Butt, H.-J. & Weber, S. A. L. High voltages in sliding water drops. J. Phys. Chem. Lett. 14, 11110–11116 (2023).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

17. Koch, G. H., Brongers, M. P. H., Thompson, N. G., Virmani, Y. P. & Payer, J. H. Corrosion cost and preventive strategies in the United States. NACE International http://impact.nace.org/documents/ccsupp.pdf (2002).

18. Li, X. et al. Materials science: share corrosion data. Nature 527, 441–442 (2015).

    Article  ADS  CAS  PubMed  Google Scholar 

19. Hou, B. et al. The cost of corrosion in China. npj Mater. Degrad. 1, 4 (2017).

    Article  Google Scholar 

20. Richards, J., Zhao, G., Zhang, H. & Viles, H. A controlled field experiment to investigate the deterioration of earthen heritage by wind and rain. Herit. Sci. 7, 51 (2019).

    Article  Google Scholar 

21. Chiavari, C. et al. The atmospheric corrosion of quaternary bronzes: the action of stagnant rain water. Corros. Sci. 52, 3002–3010 (2010).

    Article  CAS  Google Scholar 

22. Erkal, A., D’Ayala, D. & Sequeira, L. Assessment of wind-driven rain impact, related surface erosion and surface strength reduction of historic building materials. Build. Environ. 57, 336–348 (2012).

    Article  Google Scholar 

23. Xu, N. et al. Laboratory observation of dew formation at an early stage of atmospheric corrosion of metals. Corros. Sci. 44, 163–170 (2002).

    Article  CAS  Google Scholar 

24. Wood, R. J. K. Marine wear and tribocorrosion. Wear 376-377, 893–910 (2017).

    Article  ADS  CAS  Google Scholar 

25. Fay, L. & Shi, X. Laboratory investigation of performance and impacts of snow and ice control chemicals for winter road service. J. Cold Reg. Eng. 25, 89–114 (2011).

    Article  Google Scholar 

26. Hughes, A. E., Mol, J. M. C., Zheludkevich, M. L. & Buchheit, R. G. Active Protective Coatings: New-Generation Coatings for Metals (Springer, 2016).

27. Sørensen, P. A., Kiil, S., Dam-Johansen, K. & Weinell, C. E. Anticorrosive coatings: a review. J. Coat. Technol. Res. 6, 135–176 (2009).

    Article  Google Scholar 

28. Cheng, X., Sun, T.-P. & Gordillo, L. Drop impact dynamics: impact force and stress distributions. Annu. Rev. Fluid Mech. 54, 57–81 (2022).

    Article  ADS  Google Scholar 

29. Wood, R. J. K. Tribo-corrosion of coatings: a review. J. Phys. D 40, 5502–5521 (2007).

    Article  ADS  CAS  Google Scholar 

30. Sun, T.-P. et al. Stress distribution and surface shock wave of drop impact. Nat. Commun. 13, 1703 (2022).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

31. Bricker, O. P. & Rice, K. C. Acid rain. Annu. Rev. Earth Planet. Sci. 21, 151–174 (1993).

    Article  ADS  CAS  Google Scholar 

32. Menz, F. C. & Seip, H. M. Acid rain in Europe and the United States: an update. Environ. Sci. Policy 7, 253–265 (2004).

    Article  CAS  Google Scholar 

33. Zappi, A. et al. Factors influencing aerosol and precipitation ion chemistry in urban background of Moscow megacity. Atmos. Environ. 294, 119458 (2023).

    Article  CAS  Google Scholar 

34. Heusler, K. E., Landolt, D. & Trasatti, S. Electrochemical corrosion nomenclature (Recommendations 1988). Pure Appl. Chem. 61, 19–22 (1989).

    Article  ADS  CAS  Google Scholar 

35. Keresztesi, Á, Birsan, M.-V., Nita, I.-A., Bodor, Z. & Szép, R. Assessing the neutralisation, wet deposition and source contributions of the precipitation chemistry over Europe during 2000–2017. Environ. Sci. Eur. 31, 50 (2019).

    Article  Google Scholar 

36. Si, L. & Li, Z. Atmospheric precipitation chemistry and environmental significance in major anthropogenic regions globally. Sci. Total Environ. 926, 171830 (2024).

    Article  CAS  PubMed  Google Scholar 

37. Fernández de la Mora, J. The fluid dynamics of Taylor cones. Annu. Rev. Fluid Mech. 39, 217–243 (2007.

    Article  ADS  MathSciNet  Google Scholar 

38. Collins, R. T., Jones, J. J., Harris, M. T. & Basaran, O. A. Electrohydrodynamic tip streaming and emission of charged drops from liquid cones. Nat. Phys. 4, 149–154 (2008).

    Article  CAS  Google Scholar 

39. Crowley, J. M. Simple expressions for force and capacitance for a conductive sphere near a conductive wall. In Proc. ESA Annual Meeting on Electrostatics paper D1 (2008).

40. Zhao, L. & Liu, C. L. Review and mechanism of the thickness effect of solid dielectrics. Nanomaterials 10, 2473 (2020).

    Article  PubMed  PubMed Central  Google Scholar 

41. Nowak, D. et al. Nanoscale chemical imaging by photoinduced force microscopy. Sci. Adv. 2, e1501571 (2016).

    Article  ADS  PubMed  PubMed Central  Google Scholar 

42. Dazzi, A. & Prater, C. B. AFM-IR: technology and applications in nanoscale infrared spectroscopy and chemical imaging. Chem. Rev. 117, 5146–5173 (2017).

    Article  CAS  PubMed  Google Scholar 

43. Ţucureanu, V., Matei, A. & Avram, A. M. FTIR spectroscopy for carbon family study. Crit. Rev. Anal. Chem. 46, 502–520 (2016).

    Article  PubMed  Google Scholar 

44. Feng, X. et al. Production and method optimization of fluorescent polystyrene. J. Mol. Struct. 1243, 130746 (2021).

    Article  CAS  Google Scholar 

45. Zhang, X., Odnevall Wallinder, I. & Leygraf, C. Mechanistic studies of corrosion product flaking on copper and copper-based alloys in marine environments. Corros. Sci. 85, 15–25 (2014).

    Article  Google Scholar 

46. He, J. et al. Fast synthesis of hierarchical cuprous oxide for nonenzymatic glucose biosensors with enhanced sensitivity. J. Mater. Sci. 51, 9696–9704 (2016).

    Article  ADS  CAS  Google Scholar 

Download references

Authors and Affiliations

1. Max Planck Institute for Polymer Research, Mainz, Germany

   Zhongyuan Ni, Xiaomei Li, Aaron D. Ratschow, Lin Jian, Xiaoteng Zhou, Pravash Bista, Diego Cortes, Gunnar Glasser, Haojian Luo, Shuai Chen, Jiyao Yu, Yongkang Wang, Katrin Amann-Winkel, Kaloian Koynov, Rüdiger Berger & Hans-Jürgen Butt

2. South China Advanced Institute for Soft Matter Science and Technology, School of Emergent Soft Matter, South China University of Technology, Guangzhou, China

   Xiaomei Li

3. Department of Mechanical Engineering, Massachusetts Institute of Technology, Cambridge, MA, USA

   Xiaoteng Zhou

4. Institute for Inorganic Chemistry, University of Bonn, Bonn, Germany

   Shuai Chen

5. Institute of Physics, Johannes Gutenberg University Mainz, Mainz, Germany

   Katrin Amann-Winkel

Contributions

H.-J.B. and R.B. proposed and supervised the work. Z.N., X.L. and X.Z. prepared the samples. Z.N. and X.L. performed the drop impact and slide experiments and analysed the data. A.D.R. derived the analytical theory. Z.N., L.J. and K.K. performed confocal laser scanning microscopy experiments. Z.N. and P.B. performed the charge measurements. Z.N. and D.C. performed the bottom-view high-speed camera experiments. Z.N. and R.B. performed the AFM and AFM-IR measurements. Z.N. and G.G. performed the SEM-EDS measurements and analysed the data. Z.N., H.L. and Y.W. performed the Raman measurements and analysed the data. Z.N., S.C. and K.A.-W. performed the XRD measurements and analysed the data. Z.N. and J.Y. performed the EIS measurements and analysed the data. Z.N., R.B. and H.-J.B. prepared the manuscript. All of the authors discussed and interpreted the results together.

Corresponding authors

Correspondence to Rüdiger Berger or Hans-Jürgen Butt.

Competing interests

The authors declare no competing interests.

Peer review information

Nature thanks Kevin Golovin, Zuankai Wang and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions