Process Library

Recipes should be submitted at bottom of this page in comments section or emailed to Neil Dilley ndilley@purdue.edu ; they need to include:

• mask type and how to prepare it (e.g., photoresist type, spin speed, bake temp/time, exposure, development)
• milling 
  • program # 
  • duty cycle (on/off times)
  • angle of substrate

From Qi Wang

Mask: Photoresist AZ9260/2000rpm/bake at 110C for 3min
Material to etch: SmNiO3
Depth: 50nm
Process used in PhaseIIJ software (program #, angle, and milling duty cycle): Program 3_-30deg_2min
A 5min ultrasonic cleaning in acetone could remove most of the photo resist. The stubborn residue near pattern edges was able to be removed by soaking in remover-PG at 80C for 1h and ultrasonic cleaning for 2min. 

From Sen Dai, deep etching ~0.5um of LiNbO3 (special SOP, ask staff)

Mask: SPR220 or any PR that fits your needs.
Our requirement is that the PR is at least ~2x thicker than the material you want to etch so that it will not be etched through.
If using SPR220:
Spin at 3000rpm, then bake 115degree 90sec. This will produce 4 um film. Expose your pattern at 350mJ/cm2. Develop using MF24A 60sec. 
Milling: program #3 (V_B=600 V, (mill 3'30" / cool 12' ) repeated 8x.

From Chengzi Huang

Mask: SPR220-4.5 as the photoresist, spin speed 5000rpm, bake 115C for 90s.

Milling: 50nm/40nm/150nm Pd/BYZ/NNO with program 3, angle -30, 3 cycles of (2 min etch +12min cooling)

When etching Si, PR residuals can be easily removed by 5s ultrasonic cleaning in acetone. When etching the above Pd/BYZ/NNO stack, post cleaning requires 1min ultrasonic cleaning + 5min branson asher O2 plasma clean.

From Steve Novakov (John Heron group, U. Michigan)

Mask: SPR 220-3.0 on the spinner with the following recipe: Step 1. 500 rpm, 500 rpm/s, 5s  Step 2: 4500 rpm, 1250 rpm/s, 35s. Bake on hotplate at 90 C for 2 minutes. This may not be 100% mandatory for projection exposure tools and large devices, but is mandatory for contact mask exposure and small devices. NOTE: The final spin speed can be varied/tuned. I prefer as thin of a resist as possible to shorten development/exposure time. However, in the case of TMD’s, having a modest edge/corner bead serves as an additional protective element, reducing pressure on the film from the mask aligner, etc.

Milling: Al2O3 |  Bi2Se3 (~ 10 nm)| MnSe (~ 5 nm)| Se  (~ 10 nm) : etch to Al2O3 substrate. Program #2, -30 degrees. Duty cycle milling on 30sec / off 30 sec, for a total milling time of 4 minutes until Bi and Se signals level off in EPD. Sample was likely milled after 2 minutes but the Bi/Se trends at EPD were tailing off slowly. 

More complete description of their litho/liftoff process is here.

Ozan Erturk (Bhave group)

PMN-PT Etching

Mask: SPR220-7 at 4000rpm, bake 110C for 30 sec hold ~1cm above the hotplate then  90s in contact.

Milling: Si/PMN-PT/ PR with program 3, angle -14, 2 cycles of ( 210 sec etch +7min cooling)  + angle -70, 2 cycles of ( 210 sec etch +7min cooling) 
This package that contains 2 cycles of -14 degree followed by -70 degree etch can be repeated as many times required to achieve the desired vertical depth.

Etched material: 1.8um  (Etch rate ~52nm/min)

Etched PR: ~1.2um (Selectivity is 1:1.5  PR : PMN-PT)

Note: selectivity significantly decreases for longer cycle times. For deeper etch depths, it is recommended to perform batches of 10 cycles with ~2hrs of cooling in between.

Si Etching

Mask: SPR220-7 at 4000rpm, bake 110C for 30 sec hold ~1cm above the hotplate then  90s in contact.
Milling: Si/PR with program 3, angle -14, 10 cycles of ( 150 sec etch +7min cooling) 

Etched material: 1.1um  (Etch rate ~45nm/min)

Etched PR: ~1.1um (Selectivity is 1:1  PR : Si)

Questions & Troubleshooting

References

Presentation: Introduction to AJA Ion Mill (Fab Forum, Feb 2019)

Process Library

Recipes should be submitted at bottom of this page in comments section or emailed to Neil Dilley ndilley@purdue.edu ; they need to include:

• mask type and how to prepare it (e.g., photoresist type, spin speed, bake temp/time, exposure, development)
• milling 
  • program # 
  • duty cycle (on/off times)
  • angle of substrate

From Qi Wang

Mask: Photoresist AZ9260/2000rpm/bake at 110C for 3min
Material to etch: SmNiO3
Depth: 50nm
Process used in PhaseIIJ software (program #, angle, and milling duty cycle): Program 3_-30deg_2min
A 5min ultrasonic cleaning in acetone could remove most of the photo resist. The stubborn residue near pattern edges was able to be removed by soaking in remover-PG at 80C for 1h and ultrasonic cleaning for 2min. 

From Sen Dai, deep etching ~0.5um of LiNbO3 (special SOP, ask staff)

Mask: SPR220 or any PR that fits your needs.
Our requirement is that the PR is at least ~2x thicker than the material you want to etch so that it will not be etched through.
If using SPR220:
Spin at 3000rpm, then bake 115degree 90sec. This will produce 4 um film. Expose your pattern at 350mJ/cm2. Develop using MF24A 60sec. 
Milling: program #3 (V_B=600 V, (mill 3'30" / cool 12' ) repeated 8x.

From Chengzi Huang

Mask: SPR220-4.5 as the photoresist, spin speed 5000rpm, bake 115C for 90s.

Milling: 50nm/40nm/150nm Pd/BYZ/NNO with program 3, angle -30, 3 cycles of (2 min etch +12min cooling)

When etching Si, PR residuals can be easily removed by 5s ultrasonic cleaning in acetone. When etching the above Pd/BYZ/NNO stack, post cleaning requires 1min ultrasonic cleaning + 5min branson asher O2 plasma clean.

From Steve Novakov (John Heron group, U. Michigan)

Mask: SPR 220-3.0 on the spinner with the following recipe: Step 1. 500 rpm, 500 rpm/s, 5s  Step 2: 4500 rpm, 1250 rpm/s, 35s. Bake on hotplate at 90 C for 2 minutes. This may not be 100% mandatory for projection exposure tools and large devices, but is mandatory for contact mask exposure and small devices. NOTE: The final spin speed can be varied/tuned. I prefer as thin of a resist as possible to shorten development/exposure time. However, in the case of TMD’s, having a modest edge/corner bead serves as an additional protective element, reducing pressure on the film from the mask aligner, etc.

Milling: Al2O3 |  Bi2Se3 (~ 10 nm)| MnSe (~ 5 nm)| Se  (~ 10 nm) : etch to Al2O3 substrate. Program #2, -30 degrees. Duty cycle milling on 30sec / off 30 sec, for a total milling time of 4 minutes until Bi and Se signals level off in EPD. Sample was likely milled after 2 minutes but the Bi/Se trends at EPD were tailing off slowly. 

More complete description of their litho/liftoff process is here.

Ozan Erturk (Bhave group)

PMN-PT Etching

Mask: SPR220-7 at 4000rpm, bake 110C for 30 sec hold ~1cm above the hotplate then  90s in contact.

Milling: Si/PMN-PT/ PR with program 3, angle -14, 2 cycles of ( 210 sec etch +7min cooling)  + angle -70, 2 cycles of ( 210 sec etch +7min cooling) 
This package that contains 2 cycles of -14 degree followed by -70 degree etch can be repeated as many times required to achieve the desired vertical depth.

Etched material: 1.8um  (Etch rate ~52nm/min)

Etched PR: ~1.2um (Selectivity is 1:1.5  PR : PMN-PT)

Note: selectivity significantly decreases for longer cycle times. For deeper etch depths, it is recommended to perform batches of 10 cycles with ~2hrs of cooling in between.

Si Etching

Mask: SPR220-7 at 4000rpm, bake 110C for 30 sec hold ~1cm above the hotplate then  90s in contact.
Milling: Si/PR with program 3, angle -14, 10 cycles of ( 150 sec etch +7min cooling) 

Etched material: 1.1um  (Etch rate ~45nm/min)

Etched PR: ~1.1um (Selectivity is 1:1  PR : Si)

Questions & Troubleshooting

References

Presentation: Introduction to AJA Ion Mill (Fab Forum, Feb 2019)