The innovation centres on a specially designed fibre membrane, capable of removing heat through evaporation without any need for additional energy input. Unlike conventional methods such as fans, heat sinks, or liquid pumps, the system leverages the natural capillary action of a porous membrane to draw cooling liquid across its surface, where it evaporates and extracts heat from the underlying electronics.
As the uptake of artificial intelligence (AI) and Cloud computing continues to rise, so too does the demand for data processing – and the thermal load that comes with it. Currently, cooling accounts for as much as 40% of the total energy consumption in a data centre. If growth continues unchecked, global energy use for cooling could more than double by 2030.
The UC San Diego team’s solution could help limit that impact. The membrane features a network of small, interconnected pores that optimise fluid flow and surface area for evaporation. It is positioned over microchannels that transport coolant beneath the membrane surface, allowing for passive and efficient heat dissipation.
“Compared to traditional air or liquid cooling, evaporation can dissipate higher heat flux while using less energy,” said Professor Renkun Chen of the Department of Mechanical and Aerospace Engineering at the UC San Diego Jacobs School of Engineering. He co-led the project alongside Professors Shengqiang Cai and Abhishek Saha, also of the same department. Mechanical and Aerospace Engineering Ph.D. student Tianshi Feng and postdoctoral researcher Yu Pei – both from Chen’s group – served as co-first authors on the paper.
Evaporation is already used in many common cooling applications, such as heat pipes in laptops and evaporators in air conditioning systems. However, applying it effectively to high-power electronics has posed challenges. Chen explained that earlier approaches using porous membranes often failed due to incorrect pore sizing – too small and the pores clog, too large and boiling occurs.

