Insects such as water striders are able to walk effortlessly on water because of the fact that their legs are super hydrophobic.
In nature, organisms such as caterpillars, water striders and the lotus achieve super hydrophobia through a two-level structure - a hydrophobic waxy surface made super hydrophobic by the addition of microscopic hair-like structures: these structures may be covered by even smaller hairs, greatly increasing the surface area of the organism and making it impossible for water droplets to stick.
Using a superfast supercomputer at Riken (the fastest in the world when the research started in 2005), a Japanese team led by Xiao Cheng Zeng, Ameritas university professor of chemistry at the University of Nebraska-Lincoln, designed a computer simulation to perform tens of thousands of experiments that studied how surfaces behaved under many different conditions.
Zeng and his colleagues used the Riken computer to ‘rain’ virtual water droplets of different sizes and at different speeds on surfaces that had pillars of various heights and widths and with different amounts of space between the pillars.
What they learned is that there is a critical pillar height, depending on the particular structure of the pillars and their chemical properties, beyond which water droplets cannot penetrate.
If the droplet can penetrate the pillar structure and reach the waxy surface, it is in the merely hydrophobic Wenzel state (named after Robert Wenzel, who found the phenomenon in nature in 1936).
If the droplet cannot penetrate the pillars to touch the surface, the structure is in the super hydrophobic Cassie state (named after A.B.D. Cassie, who discovered it in 1942) and the droplet rolls away.
Zeng said there were three main advantages to performing the experiments on a computer rather than in a laboratory.
First, they were able to conduct thousands more repetitions than would have been possible in a lab.
Second, they didn't have to worry about variables such as dirt, temperature and air flow.
Third, they could control the size of droplets down to the exact number of molecules, whereas in a laboratory experiment the droplets would unavoidably vary by tens of thousands of molecules.