In Simple Terms
In tiny channels filled with water and charged particles, electricity can make the walls bulge in or out. That changes how easily the particles flow, and the channel may not return to its original shape right away. The findings suggest that some tiny devices can remember past electrical signals because their structure changes.
A Closer Look at the Channel
Imagine a very narrow passage carrying water and charged particles. Now imagine that its walls are not completely rigid, but can bulge slightly when electricity is applied. That idea is at the heart of a study published in Nature Materials in 2026. The researchers directly observed a link between changes in the shape of two-dimensional nanoscale channels and the flow of ions through them. Rather than measuring current only at the two ends of a device, they combined electrical measurements with optical imaging. This let them watch the microscopic structure change while tracking the electrical response at the same time.
Memory Is Not Just in the Ions
Nanofluidic devices move ions through extremely narrow pathways. At this scale, ions do not behave as though they are in a wide container: nearby walls and interactions at the surfaces strongly affect how easily the ions can pass. As a result, the current may differ when the voltage is reversed, and the device’s response may depend on its earlier electrical states. This history-dependent behavior is known as ionic memory. It is of interest in devices designed to process information in ways inspired by some features of neural networks.
But observing memory-like behavior does not, by itself, reveal what causes it. Researchers have proposed several explanations, including chemical reactions, differences in how surfaces become wetted, specific interactions between ions and walls, and changes in the channel’s shape. Earlier studies have considered whether ion pairs or chains of charged molecules might contribute in some devices. Directly connecting these possibilities to changes inside the structure has remained challenging. The new work adds simultaneous observations of structural changes and current, showing that mechanical movement of the walls may be an important part of the memory seen in the devices studied.
A Channel Sandwiched Between Thin Layers
The team made the channels by placing layers of hexagonal boron nitride around a thin layer of graphene. This arrangement created a narrow passage for ions, and a drilled opening set the channel’s length at about 5 micrometers. The researchers examined 14 devices, since devices that look alike from the outside can differ in how tightly their layers stick together or in whether they already contain defects or bulges. Those differences can affect electrical readings.
The researchers placed the devices in a dedicated fluidic system and used a reflection microscope to capture changes in light reflected from the layers. The reflected light shifts slightly when the layers change shape or the spacing between them changes, making it possible to infer when a wall has bulged. At the same time, electrodes applied a sinusoidal alternating voltage and recorded the resulting current. A microscope with a lens designed for use in water captured a sequence of images of the structure. Synchronizing the images with the electrical measurements allowed the team to compare what was happening inside the channel with the current at each moment.
The team used an optical measure they called voltage-induced contrast. It compares the intensity of reflected light while voltage is applied with its intensity when no voltage is present. The researchers then used optical modeling to estimate the height of the bulges. These heights are not precise measurements of every bump: uncertainty about the thickness of the layers means they indicate only the approximate scale, an important limitation when interpreting their dimensions. Even so, the signal was strong enough to show that some areas changed shape at the same time as clear shifts in conductivity.
When Does Electrical Pressure Make a Bulge?
The team tested aqueous potassium chloride solutions with concentrations ranging from 1 millimolar to 1 molar, as well as 1 molar solutions of sodium chloride, lithium chloride, and calcium chloride. They also used theoretical analysis to investigate whether osmotic pressure and electrical forces associated with surface charges could overcome the attraction between the layers and make them bend. The presence of liquid alone does not pull the layers apart; an outward force must be strong enough to overcome their resistance to separation.
At low voltage, current and voltage were roughly proportional, and the researchers saw no noticeable electrical deformation. However, some devices carried more current than theoretical calculations predicted. Pre-existing bulges or areas where the layers were not tightly attached could help explain this: such spots provide a wider effective space for ions to move through, even before the voltage causes any further change.
When the voltage exceeded about 0.2 volts in one of the devices examined, a bulge appeared or an existing one grew. This coincided with a departure from the linear relationship between voltage and current and with electrical hysteresis that depended on the direction of the cycle. In other words, the device no longer followed the same path as the voltage rose and then fell. The channel’s shape did not necessarily return to its starting point along the same path, and widening or narrowing the passage changed the space available for ions to flow through.
Two Kinds of Bulges, Two Different Responses
Analysis revealed two patterns, distinguished by how the bulges responded to voltage polarity. In the bidirectional pattern, a bulge could expand under both positive and negative voltage. The model links this behavior to concentration polarization: as voltage redistributes ions, the average salt concentration inside the channel rises, creating an osmotic difference between the liquid trapped inside it and the liquid in the reservoir. That difference generates pressure pushing the wall outward, whether the polarity is positive or negative under the conditions studied.
The unidirectional pattern, by contrast, responds mainly to one polarity. This is consistent with an explanation based on electrical forces acting on charges at the channel surfaces. Under one voltage direction, those forces can push the surrounding layers outward. These patterns do not imply that only one response occurs in these devices. Instead, they suggest that surface structure, adhesion, and charge can determine which mechanism has the stronger influence in a particular region.
The researchers analyzed 50 bulges in nine devices and observed differences between the two responses. Most bulges collapsed when the voltage returned to zero, but some deformations persisted during slower voltage cycles. This retention was particularly evident at frequencies between 10 and 30 millihertz in one device. The lasting response offers a physical explanation for how an electrical effect can remain after the stimulus ends: the channel’s structure can retain a state shaped by its voltage history.
A Small Shape Change Can Transform the Readout
The structural changes were more than details visible under a microscope; they were associated with substantial shifts in conductivity. In one device, conductivity rose by about 50% in less than a second after a particular bulge changed shape. In another device, the current-voltage curves differed from one cycle to the next. The researchers attributed this behavior to two regions where bulges were developing. When the bulges changed shape consistently, the response was more repeatable. Irregular changes, by contrast, produced greater variation from cycle to cycle.
These observations help explain why devices made from the same materials and arranged in the same way may perform differently. Current may depend not only on chemical composition but also on how tightly the layers adhere, the surface charges, the channel’s shape, bulges that formed earlier, and the speed at which the voltage changes. The models showed that osmotic and electrical pressures can overcome relatively weak adhesion between layers. They also showed that differences in how a bulge grows and recedes create distinct paths for its height and for conductivity. This gives rise to hysteresis: the current at a particular voltage can depend on whether the voltage is rising or falling.
What the Results Show—and What They Don’t
The study provides experimental evidence that mechanical movement and electrical response are coupled in the channels the team examined. Voltage does more than move ions; it can also reshape the passage they travel through. This matters when building models of nanofluidic devices: calculations focused only on ion interactions may not be enough if the walls can deform. A complete picture should also account for adhesion between layers, pressure forces, and changes in geometry over time.
However, the observations do not prove that all ionic memory is caused by mechanical bulging. The researchers noted that purely ionic mechanisms could cause a change in refractive index of less than 1%, a shift that may fall below the optical method’s detection limit. Other explanations therefore remain possible in systems where no visible deformation occurs or where the imaging tools cannot resolve it. The work shows that deformation was present and influential in the devices that were imaged—not that it explains every nanofluidic system.
The study, by Sourav K. F. Varunse and colleagues, published in Nature Materials in 2026, also points to a possible engineering strategy. If the location and repeatability of bulges depend on how strongly the layers adhere, introducing low-adhesion regions in selected places might steer deformation to known locations. This is a design proposal, not a final result proving that the devices are ready for use. But it suggests a way to make bulging more repeatable and reduce variation between samples.
The Bottom Line
These experiments show that the memory of some ion channels may be partly built into their shape. An electrical signal can create or enlarge a liquid-filled bulge, changing the channel’s cross-section and altering how ions move. Some of that change can persist even after the voltage returns to zero. By combining imaging and electrical measurements in devices with channels about 5 micrometers long, the researchers linked tiny deformations to current hysteresis and sharp changes in conductivity, including a 50% increase in less than a second. The mechanism depends on the properties of each device, and optical measurements cannot rule out every other ionic explanation. Still, controlling adhesion, surface charges, geometry, and frequency could help make future nanofluidic memory devices more reliable—and may support the development of electrically controlled nanoscale valves and ion-based information processing.