Controlling dropwise condensation is fundamental to water-harvesting systems, desalination, thermal power generation, air conditioning, distillation towers, and numerous other applications. For any of these, it is essential to design surfaces that enable droplets to grow rapidly and to be shed as quickly as possible. However, approaches based on microscale, nanoscale or molecular-scale textures suffer from intrinsic trade-offs that make it difficult to optimize both growth and transport at once. Here we present a conceptually different design approach—based on principles derived from Namib desert beetles, cacti, and pitcher plants—that synergistically combines these aspects of condensation and substantially outperforms other synthetic surfaces. Inspired by an unconventional interpretation of the role of the beetle’s bumpy surface geometry in promoting condensation, and using theoretical modelling, we show how to maximize vapour diffusion flux at the apex of convex millimetric bumps by optimizing the radius of curvature and cross-sectional shape. Integrating this apex geometry with a widening slope, analogous to cactus spines, directly couples facilitated droplet growth with fast directional transport, by creating a free-energy profile that drives the droplet down the slope before its growth rate can decrease. This coupling is further enhanced by a slippery, pitcher-plant-inspired nanocoating that facilitates feedback between coalescence-driven growth and capillary-driven motion on the way down. Bumps that are rationally designed to integrate these mechanisms are able to grow and transport large droplets even against gravity and overcome the effect of an unfavourable temperature gradient. We further observe an unprecedented sixfold-higher exponent of growth rate, faster onset, higher steady-state turnover rate, and a greater volume of water collected compared to other surfaces. We envision that this fundamental understanding and rational design strategy can be applied to a wide range of water-harvesting and phase-change heat-transfer applications.
We thank M. Khan, J. Alvarenga, D. Daniel, S. H. Kang, M. Hoang, and J. Timonen for discussions and technical assistance. This research was supported by the Department of Energy/ARPA-E award number DE-AR0000326. N.H. thanks the Research Experiences for Undergraduates programme supported by the National Science Foundation award number DMR-1420570.