slippage at fl at, hydrophobic interfaces ( b Ӎ 10 nm). [ 7,8 ] This lack of interfacial slippage greatly reduces fl ow rates in nanofl uidic channels thereby limiting their effi ciency for energy conversion and solute separation. A promising approach to further increase slip, particularly in slit-shaped nanoscale channels, relies on textured hydrophobic surfaces and their ability to effectively trap air when immersed in water. [ 9 ] The air layer provides both "superhydrophobicity" [ 9 ] (i.e., extreme water repellency) and a shear-free interface that greatly reduces fl uid drag [ 10 ] (shown schematically in Figure 1 ). However, conventional microtextured superhydrophobic (SH) surfaces suffer from intrinsic shortcomings that greatly limit their use in nanofl uidic applications. The most serious problem is the collapse of the air layer trapped within the texture [ 11 ] at the relatively high pressures (>1 atm) [ 12 ] required to drive the liquid fl ow in nanofl uidic devices. Although, as shown recently, [ 13,14 ] the resilience of a SH surface to water infi ltration can be improved through reducing the pattern period L to ≈10 nm, in principle this may be detrimental to slippage since theoretically b Ӎ L . [ 7 ] Nevertheless, the dependence of slippage on L has been tested only in textures with micron-scale features ( L Ӎ 1 µm), [ 15 ] and the effect of nanometer sized textures is still largely unexplored.Another long standing question is whether SH surfaces bear electric charge (shown schematically in Figure 1 ), and whether such charge may contribute to enhancing the electroosmotic fl ow, [ 16 ] or enable devices for effi cient electrokinetic energy conversion. [ 17 ] There is also no experimental consensus concerning the magnitude and even the sign of the charge accumulating at the air-water interface. [ 18 ] A recent study of colloidal electrophoresis near a SH surface reported no enhancement of electroosmotic mobility. [ 19 ] However, the surface zeta potential could not be measured due to the high interfacial roughness and the study could not conclusively assign the results to the lack of surface charge.Here we show experimentally that signifi cant interfacial drag reduction in nanoscale, slit-like channels can be obtained with hydrophobic arrays of conical textures tapering to a radius of less than 10 nanometers at their tip.
Results and DiscussionWe fabricate large-area (cm 2 ) silicon surfaces containing features with uniform size and spacing on a length scale The development of highly effi cient nanofl uidic devices necessitates means for enhancing and controlling fl uid transport under confi nement. Here, it is shown experimentally that signifi cant interfacial drag reduction in nanoscale channels can be obtained with hydrophobic arrays of conical textures tapering to a radius of less than 10 nanometers at their tip. This geometry maximizes interfacial slippage by trapping a highly resilient air layer at the solid/liquid interface. Further, it is revealed that the composite liquid/solidair interface b...