The effect of surface roughness on interfacial measurements

The ideal interface would be one which was smooth on an atomic scale. The only interfaces which can be produced which are as smooth as it is theoretically possible are interfaces between a single crystal of a high melting metal, e.g. platinum, and a deformable polymer. All interfaces between two solids will tend to be very rough on an atomic scale. The question therefore arises - How do real interfaces differ from idealised smooth interfaces  

One important way in which they differ is that currents (either ac or dc) at real interfaces are not uniform across the interface. Therefore, a measured macroscopic dc current density i (A cm ) will not in general be a microscopic current density i on a small part of the interface (say 1 pm by 1 pm) on a rough or non-uniform electrode. In the formation of each interface it is necessary therefore to ensure that the surface is as smooth as is reasonably practicable. In most cases codes of best practice have been evolved and should generally be followed unless radical improvements are possible. In this way results should at least be comparable from one laboratory to another. 

One of the major causes of the distorted behaviour seen when rough electrodes are used is due to the fact that different parts of the interface have different bulk resistance pathways to them, in some cases causing transmission line type behaviour. 

---