For more than a century, one section of Earth’s atmosphere has remained an unbroken barrier to sustained flight. The mesosphere, stretching from around fifty to eighty-five kilometres above the surface, has been unreachable for anything that needs to remain aloft for more than minutes. It sits above the limits of balloons and aircraft, yet far below the orbits of satellites. For science and technology, this middle layer has been an open void. There has been no way to place instruments there and keep them flying in place for long-term monitoring.
That is now beginning to change. A new study has demonstrated the first flight of a macroscopic device under conditions that match natural sunlight and the low pressure of near-space. The device is not a balloon or a conventional aircraft. It is a centimetre-scale, nanofabricated structure that can remain suspended by the force of light alone. The breakthrough opens a path to filling the mesosphere gap with long-lived platforms for sensing, communications, and even interplanetary exploration.
The research team designed their flyer around a physical effect called photophoresis. In simple terms, when one side of an object in a gas is warmer than the other, gas molecules hitting the warmer side bounce away with greater momentum. In the thin air of the upper atmosphere, this imbalance produces a net force that can push the object. Most forms of photophoresis are too weak to move anything large. What makes this work stand out is the use of a related mechanism called thermal transpiration. This uses a patterned structure of microscopic holes so that gas flows from cooler to warmer regions, creating lift strong enough to balance the device’s weight.
The mesosphere is ideally suited to this approach. At lower altitudes the pressure is too high, and convection dominates heat transfer. At higher altitudes the air is too thin for significant interaction. In the mesosphere, the mean free path of gas molecules matches the size of engineered features in the device, allowing maximum efficiency. The study’s models show that the most powerful lift occurs in the sixty to eighty kilometre range, right in the middle of the mesosphere.
The structure they developed is a double-membrane disk made from alumina just one hundred nanometres thick. Between the membranes are vertical cylindrical ligaments, only a few micrometres across, that connect some of the holes in the upper and lower layers. This creates two distinct zones. In one zone, every hole in the top layer is directly linked to a hole in the bottom layer by a ligament. These areas are strong and rigid but lose some lifting power because the ligaments conduct heat and reduce the temperature difference. In the other zone, most holes are not aligned, allowing the temperature contrast between the two membranes to grow and drive more lift, but at the cost of structural stiffness. A small number of ligaments are still placed in this zone to prevent the two layers from shifting.
The bottom membrane is engineered to absorb as much sunlight as possible while minimising heat loss through radiation. The researchers achieved this by applying a stack of chromium and alumina layers designed to act as a selective solar absorber. Light from the Sun is absorbed strongly, heating the lower membrane, while infrared heat loss is restricted. The top membrane is made to be more reflective in the visible range and more emissive in the infrared so it stays cooler. This combination ensures a large temperature difference between the top and bottom surfaces.
To prove the concept, the team carried out experiments in a custom-built low-pressure chamber. They mounted the one square centimetre samples on a force transducer that could measure changes in apparent weight when light was applied. A diffused 447-nanometre laser provided the illumination, adjusted to match or exceed natural sunlight intensity. They tested in air, helium, and sulfur hexafluoride to study the effect of gas properties. As predicted, lighter gases like helium produced more lift at a given pressure because the transition regime between molecular flow and continuum flow occurred at higher pressures.
The key result came when a sample just one centimetre across and weighing one hundred and thirty micrograms was illuminated at around seven hundred and fifty watts per square metre, about fifty-five percent of the Sun’s intensity at Earth. At an air pressure of twenty-six point seven Pascals, equivalent to roughly sixty kilometres altitude, the device lifted clear of its support and hovered freely. This is the first recorded instance of a macroscopic object levitating under sunlight-equivalent conditions using photophoretic force.
The researchers also measured how lift changed with pressure, light intensity, and ligament density. Samples with the lowest ligament filling fraction produced the highest forces, confirming the importance of limiting thermal conduction between membranes. Lift scaled linearly with light intensity in the tested range, meaning that under full sunlight the effect would be even stronger. They also verified that the phenomenon worked with broadband white light from LEDs, ruling out a narrow wavelength dependence.
From their models, the team calculated how different sizes and designs would perform in the real atmosphere. A device three centimetres in radius, built with the same approach, could lift a ten milligram payload above its own five milligram weight at an altitude of seventy-five kilometres in daylight. In continuous polar summer daylight it could remain there indefinitely. Larger devices, up to eighty centimetres in radius, could carry nearly a gram under peak conditions at the equator.
Practical applications begin with atmospheric science. The mesosphere plays a role in global circulation patterns, the filtering of incoming solar radiation, and the formation of high-altitude clouds, yet it is barely sampled. Current tools like sounding rockets pass through it in minutes. Satellites can only observe it indirectly with large uncertainties. Arrays of photophoretic flyers could provide continuous, high-resolution measurements of wind, temperature, pressure, and humidity in situ. This would allow more accurate climate models and better weather forecasts.
Another application is in communications. A ten milligram payload can include a micro radio-frequency transmitter, a small solar cell, and control electronics. In daylight, such a package could send data at tens of megabits per second to ground receivers. At night the rate would drop to a few kilobits per second using stored energy. A network of these devices could act as a relay layer between ground stations and satellites or provide emergency links in remote areas. With a one hundred metre baseline phased array of ground antennas, the position of each flyer could be tracked to within a square kilometre.
The study also examines how to keep devices aloft through the night. In equatorial regions they would descend without sunlight. Power beaming from the ground using lasers or microwaves could sustain them, though this brings technical challenges such as atmospheric losses and safety considerations. Another approach is to design smaller, lighter structures optimised to absorb infrared radiation from Earth’s surface at night. These could maintain altitude but with reduced payload capacity.
Materials and durability are important factors. The mesosphere is a harsh environment, with high levels of ultraviolet radiation, reactive oxygen, and particles from meteoric dust. Alumina is robust, but coatings, electronics, and moving parts will require protection. Sulfate aerosols could damage surfaces or clog perforations over time, so long-term tests are essential before deployment.
Beyond Earth, the technology could work in the Martian atmosphere. The reduced gravity and lower pressure on Mars create conditions similar to those at seventy kilometres on Earth. Although sunlight is weaker, the balance between reduced weight and reduced lift still allows operation. Flyers could carry lightweight sensors or act as communication relays in the forty to seventy kilometre range. Seasonal periods of low wind could allow targeted missions, avoiding the high-speed zonal winds and dust storms that can dominate the planet’s weather.
The development of a working prototype involved solving both engineering and physics challenges. The team created a hybrid analytical and numerical model to predict lift based on structure size, porosity, ligament density, and optical properties. They accounted for heat transfer through conduction, radiation, and the movement of gas molecules through the perforations. The numerical part used direct simulation Monte Carlo methods to model gas-surface interactions at different pressures and flow regimes. The analytical part calculated temperature differences between the membranes for different configurations. Combining these allowed them to identify optimal designs for specific altitudes.
Fabrication was carried out using atomic layer deposition of alumina onto a patterned sacrificial silicon template. This method produces uniform, conformal coatings even over complex geometries. After deposition, the sacrificial material was removed to leave a freestanding double-membrane structure with precisely controlled thickness, hole size, and ligament placement. The optical coating on the bottom membrane was added before final assembly. The resulting structures have area densities around one gram per square metre, low enough for mesospheric lift.
In their conclusion, the authors outline what is needed to move from laboratory demonstration to operational systems. This includes integrating payloads, adding attitude control to resist winds, and developing deployment methods. High-altitude balloons could carry fleets of flyers to fifty kilometres before release. The devices must also be designed to withstand launch loads and the mechanical stresses of flight. Navigation could be achieved through simple control surfaces or microactuators that adjust the tilt, allowing the flyer to move horizontally against winds of up to ten metres per second under favourable conditions.
The successful flight of this small, perforated disk under sunlight-equivalent conditions is more than a proof of principle. It marks the first time the mesosphere has a viable, sustainable aircraft platform. With refinement, the technology could open a persistent observation and communication layer that has never before existed. From monitoring climate and weather to exploring other planets, the reach of this breakthrough extends far beyond the thin blue shell where it was first tested.
Source: Nature
Schafer, B. C., Kim, J., Sharipov, F., Hwang, G.-S., Vlassak, J. J., & Keith, D. W. (2025). Photophoretic flight of perforated structures in near-space conditions. Nature, 644, 362–369. https://doi.org/10.1038/s41586-025-09281-8
