Specific Process Knowledge/Etch/DRIE-Pegasus/showerheadchange: Difference between revisions
| Line 11: | Line 11: | ||
In the new design, the gas flow resistance has been reduced by using 1 mm diameter holes in the showerhead instead. This has no impact on processes that have a continuous gas flow - i.e. processes that are not switched. However, for processes with several gas flows that switch on and off, the switching from one gas flow to another will be much more well defined. This has enabled us to run Bosch processes with shorter cycles times. Shorter cycles times means smaller scallops and hence lower roughness and better control. We have not seen any changes to the uniformity of the plasma nor any other side effect. | In the new design, the gas flow resistance has been reduced by using 1 mm diameter holes in the showerhead instead. This has no impact on processes that have a continuous gas flow - i.e. processes that are not switched. However, for processes with several gas flows that switch on and off, the switching from one gas flow to another will be much more well defined. This has enabled us to run Bosch processes with shorter cycles times. Shorter cycles times means smaller scallops and hence lower roughness and better control. We have not seen any changes to the uniformity of the plasma nor any other side effect. | ||
This modification is described by SPTS [[:File:SPTS High Flow Plenum Upgrade for Pegasus Nov 2012.pdf]] | |||
== Comparison of continuous processes == | == Comparison of continuous processes == | ||
Revision as of 13:34, 26 June 2018
Feedback to this page: click here
Modification of showerhead in december 2014
This page describes why the showerhead has been changed and what process changes have been found. SPTS calls this modification the 'High Flow Plenum upgrade'.
Motivation
The showerhead that distributes the process gasses inside the plasma source was changed. In the old design, the showerhead had a number of 300 µm diameter holes intended to distribute the gas uniformly inside the plasma source. The passage through the holes had a flow resistance that would cause the gasses coming from the MFC in a Bosch process to get mixed up - in this way working against the separation of the etch and dep cycles.
In the new design, the gas flow resistance has been reduced by using 1 mm diameter holes in the showerhead instead. This has no impact on processes that have a continuous gas flow - i.e. processes that are not switched. However, for processes with several gas flows that switch on and off, the switching from one gas flow to another will be much more well defined. This has enabled us to run Bosch processes with shorter cycles times. Shorter cycles times means smaller scallops and hence lower roughness and better control. We have not seen any changes to the uniformity of the plasma nor any other side effect.
This modification is described by SPTS File:SPTS High Flow Plenum Upgrade for Pegasus Nov 2012.pdf
Comparison of continuous processes
Black silicon recipe
| Process | Before (Old showerhead) | After (New showerhead) | |||||
|---|---|---|---|---|---|---|---|
| Name/Type | Description/parameters | Wafer ID | Comment | SEM images | Wafer ID | Comment | SEM images |
| Continuous black silicon recipe on blank wafer | 15 mins, -10 degrees, 32 mtorr, 60 sccm SF6, 55 sccm O2, 70 W platen | S004592 | Wafer centre | 
|
S004679 | Wafer centre | 
|
| S003900 | Wafer edge | 
|
S004679 | Wafer edge | 
| ||
Comparison of switched processes
Process A
| Recipe | Name | Temp. | Deposition step | Etch step | Comments | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Time | Pres. | C4F8 | SF6 | O2 | Coil | Time | Pres. | C4F8 | SF6 | O2 | Coil | Platen | Showerhead | Runs | Key words | |||
| Process A | Step1 11 cyc | 20 | 4 | 25 | 200 | 0 | 0 | 2000 | 7.0 | 25(1.5s) 90>>150 | 0 | 350(1.5s) 550 | 5 | 2800 | 120>>140(1.5s) 45 | Old | 1 | |
| Step2 44 cyc | 4 | 25 | 200 | 0 | 0 | 2000 | 7.0 | 25(1.5s) 150 | 0 | 350(1.5s) 550 | 5 | 2800 | 140(1.5s) 45 | |||||
| Step1 11 cyc | 20 | 4 | 25 | 200 | 0 | 0 | 2000 | 7.0 | 25(1.5s) 90>>150 | 0 | 350(1.5s) 550 | 5 | 2800 | 120>>140(1.5s) 45 | New | 1 | Profile improved | |
| Step2 44 cyc | 4 | 25 | 200 | 0 | 0 | 2000 | 7.0 | 25(1.5s) 150 | 0 | 350(1.5s) 550 | 5 | 2800 | 140(1.5s) 45 | |||||
SOI
| Recipe | Name | Temp. | Deposition step | Etch step | Comments | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Time | Pres. | C4F8 | SF6 | O2 | Coil | Time | Pres. | C4F8 | SF6 | O2 | Coil | Platen | Showerhead | Runs | Key words | |||
| SOI etch | SOI | 20 | 2 | 25 | 250 | 0 | 0 | 2000 | 3 | 30 | 0 | 400 | 40 | 2800 | 75 (0.025s, 75%) | Old | 1 | |
| SOI | 20 | 2 | 25 | 250 | 0 | 0 | 2000 | 3 | 30 | 0 | 400 | 40 | 2800 | 75 (0.025s, 75%) | New | 1 | OK | |
Process D
| Recipe | Name | Temp. | Deposition step | Etch step | Comments | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Time | Pres. | C4F8 | SF6 | O2 | Coil | Time | Pres. | C4F8 | SF6 | O2 | Coil | Platen | Showerhead | Runs | Key words | |||
| Process D | Original | 0 | 2 | 20 | 150 | 0 | 0 | 2000 | 2.4 | 26 | 0 | 275 | 5 | 2500 | 35 | Old | 1 | |
| Original | 0 | 2 | 20 | 150 | 0 | 0 | 2000 | 2.4 | 26 | 0 | 275 | 5 | 2500 | 35 | New | 1 | ||
| New Process D | 0 | 1 | 20 | 150 | 0 | 0 | 2000 | 3 | 26 | 0 | 275 | 5 | 2500 | 35 | New | 4 | Large undercut | |
| PrD01 | 0 | 1 | 20 | 150 | 0 | 0 | 2000 | 2.4 | 26 | 0 | 275 | 5 | 2500 | 35 | New | 2 | ||
| PrD02 | 0 | 1.1 | 20 | 150 | 0 | 0 | 2000 | 2.4 | 26 | 0 | 275 | 5 | 2500 | 35 | New | 1 | ||
| PrD-3 | 0 | 1 | 20 | 150 | 0 | 0 | 2000 | 2.5 | 26 | 0 | 275 | 5 | 2500 | 35 | New | 1 | ||
| PrD-4 | 0 | 1 | 20 | 150 | 0 | 0 | 2000 | 2.2 | 26 | 0 | 275 | 5 | 2500 | 35 | New | 1 | Best one so far! | |