It has been a slow time for the last few months as classes have been exceedingly time consuming this semester. I have made two trips to Tulsa after doing some research on the use of cold developer in order to improve line resolution. The developer MIBK:IPA and the post development rinse of pure IPA were chilled in a freezer to -15 degrees Celsius prior to the development process. The theory behind the development process is that the e-beam process causes long chain "scission" where the polymer is exposed. The main areas of exposure have longer chains and are more soluble, while the areas near by the main area of exposure still have scission but they are shorter, mostly caused by secondary and backscattered electrons. The longer the chains are the more soluble they are, but as the temperature of the developer is raised the shorter chains become more soluble as well. By lowering the temperature the shorter chains are made more insoluble and therefore the line width will increase in resolution. Below are images which show development at room temperature (RT) and at -15 degrees Celsius. Development in both cases was done for 30 seconds in both the developer and the post rinse.
 |
| E-Beam Lithography with PMMA on Silica- 2.00 nC/cm2 dose, 30 kV and 29 pA current, development performed at -15 degrees Celsius |
 |
| E-Beam Lithography with PMMA on Silica- 2.00 nC/cm2 dose, 30 kV and 29 pA current, development performed at RT (25 degrees Celsius) |
The cold development definitely showed a difference between 60 nm and 30 nm at the same dosage. There was a slight tilt to the sample that was not known until post etch which may have affected some of the etching but the resolution of both lines suggests that this was probably not the case.
Another test was run at -15 degrees Celsius which showed similar results, but even after decreasing the current to 5.6 pA the line width did not decrease as was expected, but remained the same.
 |
| E-Beam Lithography with PMMA on Silica- 2.00 nC/cm2 dose, 30 kV and 5.6 pA current, development performed at -15 degrees Celsius |
It was at this point that Joe Nabity, the designer of the system being used in Tulsa, was called (instead of just email correspondence) and our problem was shared with him. Mr. Nabity made us aware of a design parameter called the "magnification transition value", which is a magnification where the microscope must change between to relays or circuits. Being on the low end of the range can decrease resolution by a factor of two and therefore not accounting for this parameter may be the reason for our bigger line widths. On the next visit I will be determining the magnification transition value for the working distance, accelerating voltage and operating current. Then by knowing this value we can perform an etching just slightly above it and below it in order to see if this makes a difference. The pattern being used will need to be changed as well in order to make sure it fits in the magnification field. The pattern we have been using has a maximum magnification of 786x, which I believe is too small and therefore if we increase the magnification then the pattern will need to decrease in size in order to fit. The results of this trial should be ready by next week. Mr. Nabity has informed me that if this does not fix the issue then it is something more subtle.
