Thursday, November 19, 2009

Dose matrix in electron beam lithography

In academia, PIs wants results and their grad students want lithography recipes! The newbies often want to know the exact dose which would preserve the geometrical dimensions of their CAD drawing. Telling newbies that the area dose for a particular resist thickness on some substrate fall between 80-200 µC/cm^2 generally irritates them.

They do not appreciate the dynamics of developing processes. The best way to obtain the correct dose in electron beam lithography is to start with a ballpark range and then do a dose calibration to determine the most appropriate dose.

The way to do that is to program the electron beam write to write an array of patterns, starting with an low dose and ending with a very high dose, with judicious increments, as shown below.



In the schematic, we see that the electron beam writer moved in a serpentine manner (e.g. Joe Nabity's NPGS lithography software) with underexposed patterns to overexposed patterns. By carrying some SEM imaging on these patterns, it is possible to determine the exact dose!

No comments:

Post a Comment