Stepper Recipes

From UCSB Nanofab Wiki
Jump to navigation Jump to search

Back to Lithography Recipes.

Below is a listing of stepper lithography recipes. Stepper 1 and Stepper 2 are i-line systems with good piece handling capabilities.  Stepper 3 is a DUV (248nm) system primarily used for full 100mm wafers.  DUV resists do not work for i-line and i-line resists do not work for DUV.  Based on your sample reflectivity, absorption (or whether or not you use an ARC layer), and surface topography, the exposure time / focus offset parameters may vary.

This listing is a guideline to get you started. For critical lithography steps, you should run your own exposure and/or focus array to determine the proper parameters.

The recipes are tabulated to give you the values of the key parameters you will need to establish your recipe. Underlayers such as LOL2000 or PMGI can be used on the stepper systems.  See the underlayer datasheets for details.   Post develop bakes (not listed) are used to make the resist more etch resistant and depend on subsequent processes. Care should be taken with post development bakes as resist reflow can occur. Unless otherwise noted, all exposures are done on flat, silicon wafers.

Parameters are indicated in separate tables for each stepper system.

Stepper 1 (GCA 6300)

Multiply the GCA 6300 exposure times by 0.30 to get a starting exposure time for the GCA Autostep200 system.

Positive Resist (GCA 6300)

Unless otherwise noted, bakes are on hot plates. For recipes with CEM, the CEM is spun on after the first resist bake, exposure is then done, and the CEM is rinsed off with DI water before the PEB. CEM generally improves resolution and process tolerance at the expense of higher exposure time.

Resist Spin Cond. Bake Thickness Exposure Time Focus Offset PEB Developer Developer Time Comments
SPR955CM0.9 3 krpm/30” 95°C/60” ~ 0.9 um 1.2” 0 110°C/60” AZ300MIF 60"
SPR955CM-0.9 3 krpm/30” 95°C/60” ~ 0.9 um 3.0” 4 110°C/60” AZ300MIF 60"
SPR955CM-0.9

CEM365iS

3 krpm/30”

5 krpm/30”

95°C/90” ~ 0.9 um 2.2” -10 110°C/60” AZ300MIF 60"
SPR950-0.8 4 krpm/30” 95°C/60” ~ 0.8 um 1.0” 0 105°C/60” AZ300MIF 60"
SPR955CM-1.8 4 krpm/30” 90°C/90” ~ 1.8 um 2.3” 0 110°C/90” AZ300MIF 60"
SPR955CM-1.8 4 krpm/30” 90°C/90” ~ 1.8 um 1.7” -5 110°C/90” AZ300MIF 60"
SPR220-3.0 2.5 krpm/30” 115°C/90” ~ 2.7 um 2.4” 10 115°C/90” AZ300MIF 60"
SPR220-7.0 3.5 krpm/45” 115°C/120” ~ 7.0 um 4.5” 0 *50°C/60”

115°C/90”

AZ300MIF 120"
  • 1.0 um isolated lines; 1.25 um isolated spaces
    • Let sample sit in air for 20 minutes before PEB, step to 50°C for 60” first, then 115°C
  • See SPR220-7 Data File

Negative Resist (GCA 6300)

Unless otherwise noted, bakes are on hot plate. All flood exposures are done in broadband light using any contact aligner. Also, because the tone is negative, a shorter first exposure time will result in more undercut, which is desirable for single-layer lift-off processes. Under these conditions more develop time will also give more undercut.

Resist Spin Cond. Bake Thickness Exposure Time Focus Offset PEB Flood Developer Developer Time Comments
AZ5214 6 krpm/30” 95°C/60” ~ 1.0 um 0.2” 0 110°C/60” 60" AZ300MIF 60"
  • 0.7 um res. possible 
nLOF5510 3 krpm/30” 90°C/60” ~ 0.93 um 0.74” -6 110°C/60” 0 AZ300MIF 60"
  • 0.5 um line openings good dense or isolated
  • Use heated 1165 stripper for removal or lift-off
  • See nLOF5510 data file
nLOF2020 4 krpm/30” 110°C/60” ~ 2 um 0.55” -6 110°C/60” 0 AZ300MIF 90"
  • ~ .85 um line opening/lift-off good. Isolated mesas can be smaller.
  • Use heated 1165 stripper for removal or lift-off Sensetive to PEB temp.
  • See nLOF2020 Data File

Stepper 2 (AutoStep 200)

Positive Resist (AutoStep 200)

Unless otherwise noted, bakes are on hot plates. For recipes with CEM, the CEM is spun on after the first resist bake, exposure is then done, and the CEM is rinsed off with DI water before the PEB. CEM generally improves resolution and process tolerance at the expense of higher exposure time.

NOTE: The bolded exposure times were found by multiplying the exposure times from the GCA 6300 system by 0.30. They should be sued as a starting point. You will need to do an exposure array to get precise times for the Autostep system. In general, the resolution achievable is ~ 100 nm smaller for the Autostep200 system.

Resist Spin Cond. Bake Thickness Exposure Time Focus Offset PEB Developer Developer Time Comments
SPR955CM-0.9 3 krpm/30” 95°C/90” ~ 0.9 um 0.35” 0 110°C/90” AZ300MIF 60”
SPR955CM-0.9 3 krpm/30” 95°C/90” ~ 0.9 um 0.8” 0 110°C/90” AZ300MIF 60”
  • 0.5um holes
SPR955CM-1.8 4 krpm/30” 95°C/90” ~ 1.8 um 0.4” -1 110°C/90” AZ300MIF 60”
SPR950-0.8 4 krpm/30” 95°C/60” ~ 0.8 um 0.30” 0 105°C/60” AZ300MIF 60"
SPR220-3.0 2.5 krpm/30” 115°C/90” ~ 2.7 um 0.72” 10 115°C/90” AZ300MIF 60"
  • 0.5 um isolated lines
SPR220-7.0 3.5 krpm/45” 115°C/120” ~ 7.0 um 1.35" 0 *50°C/60”

115°C/90”

AZ300MIF 120"
  • 1.0 um isolated lines; 1.25 um isolated spaces
    • Let sample sit in air for 20 minutes before PEB, step to 50°C for 60” first, then 115°C
AZ4210 step1:500rpm/5", step2:4krpm/45” 95°C/60" ~ 2.1 um 0.75" 0 AZ400K:DI=1:4 60"
  • 2.0 um dense holes

Negative Resist (AutoStep 200)

Unless otherwise noted, bakes are on hot plate. All flood exposures are done in broadband light using any contact aligner. Also, because the tone is negative, a shorter first exposure time will result in more undercut, which is desirable for single-layer lift-off processes. Under these conditions more develop time will also give more undercut.

NOTE: The bolded exposure times were found by multiplying the exposure times from the GCA 6300 system by 0.30. They should be sued as a starting point. You will need to do an exposure array to get precise times for the Autostep system. In general, the resolution achievable is ~ 100 nm smaller for the Autostep200 system.

Resist Spin Cond. Bake Thickness Exposure Time Focus Offset PEB Flood Developer Developer Time Comments
nLOF5510 3 krpm/30” 90°C/60” ~ 0.93 um .25” -1 110°C/60” 0 AZ300MIF 60”
AZ5214 6 krpm/30” 95°C/60” ~ 1.0 um 0.06” 0 110°C/60” 60" AZ300MIF 60"
  • 0.7 um res. possible
nLOF2020 4 krpm/30” 110°C/60” ~ 2 um 0.17” -6 110°C/60” 0 AZ300MIF 90"
  • ~ .85 um line opening/lift-off good. Isolated mesas can be smaller.
  • Use heated 1165 stripper for removal or lift-off Sensetive to PEB temp.
NR9-1000PY 3 krpm/30” 135°C/180” lid down ~ 1.2 um 0.92” 0 115°C/120” lid down 0 AZ300MIF 20"

Stepper 3 (ASML DUV)

Stepper 3: Table of Contents
  1. Process Control Data
  2. Photomasks & Job Programming
  3. Anti-Reflective Coatings
  4. Positive Resist
  5. Negative Resist
  6. Other Lithography Processes
    1. Edge Bead Removal
    2. Lift-off Processes

Process Control Data

ASML CD Calibration data - Screenshot of Table
Example of Data Table with SEM's of 320nm features. Click for full data table.
ASML CD Calibration Data - Screenshot of SPC Plot
Example SPC Chart - Measured Critical Dimension "CD" versus Date. Click for charts.

Particle Checks

Our cal process enables both ~300nm critical dimension logging, and also allows us to see when a particle is present on the exposure chuck (a "hot spot").

Image showing 2 wafers with colorful rainbow patterns, both show a black spot on the right side.
Both wafers show a "hot spot" on the left side, so the tool has a particle at that location, requiring physical cleaning. (Demis D. John, 2024)

Photomasks & Job Programming

You can find info on making photomasks, CAD files for alignment marks, and other Job programming tools on the main Tool page, here:

Stepper 3 (ASML DUV) > Design Tools

Anti-Reflective Coatings

Bottom Anti-reflective coatings (aka. BARC or AR Coating) are, in general, used for the ASML stepper. LOL2000 and PMGI can also be used as under layers. BARC layers will increase the process tolerance, especially for features ≤ wavelength (248nm).

For example, without BARC you might find that 200nm features only resolve with focus of -0.2µm to -0.3µm, while with BARC the process tolerance may increase to -0.1µm to -0.4µm, which also increases tolerance to wafer flatness/PR spin uniformity.

DUV-42P-6

(replacement for AR2)

  • spin coat at 2500rpm for optimal anti-reflective properties (~60nm).
    • Mistake: this was previously written as 3500rpm.
  • Bake at 220°C for 60s on a hotplate.
  • This AR coating is removed via oxygen plasma.
  • This ARC can be etched on ICP#1, ICP#2, RIE#5, FL-ICP or even Technics PEii ashers.
  • Datasheet: DUV-42P-6

DS-K101-304

  • Spin at 1500rpm and bake at 185°C for 60sec
    • Approx 40nm for best anti-reflective properties
    • Mistake: previously written as 5000rpm, ~20nm thickness (from DS-K101-307)
  • This AR coating develops away and undercuts in AZ300MIF.
    • For isolated lines, this can cause them to lift-off by undercutting the resist.
  • Increase bake temperature to reduce undercut rate.
  • Datasheet: DS-K101-304

Positive Resist (ASML DUV)

Please see section above for anti-reflection coatings, which are usually used with the DUV Stepper.

Resist Spin Cond. Bake Thickness Exposure Dose(mj) Focus Offset PEB Developer Developer Time Comments
UV6-0.7
(replaced by UV6-0.8)
3.5 krpm/30” 135°C/60” 630nm 17 -0.2 135°C/90” AZ300MIF 45”
  • 200nm dense line/space
  • NA 0.57, Sigma 0.75
  • Eo ~ 5.5mj
  • UV6-0.7 was discontinued, we now stock UV6-0.8
UV210-0.3 5.0 krpm/30” 135°C/60” 230nm 20 -0.1 135°C/90” AZ300MIF 45"
  • 150nm dense line/space
  • Annular Illumination
  • NA 0.63, Sigma_o 0.8, Sigma_i 0.5
UV210-0.3 3.0 krpm/30” 135°C/90” 260nm 85 -0.2 135°C/90” AZ300MIF 80”
  • 170nm isolated holes
  • Annular Illumination
  • NA 0.63, Sigma_o 0.8, Sigma_i 0.5
UV26-2.5 Available but no recipes characterized 135°C/90s ~2.5 µm approx. 40 approx. +0.8 110°C/90s AZ300MiF Unknown Users must run your own development/FEM's.

Negative Resist (ASML DUV)

Please see section above for anti-reflection coatings, which are usually used with the DUV Stepper.

Resist Spin Cond. Bake Thickness Exposure Dose (mj) Focus Offset PEB Flood Developer Developer Time Comments
UVN30-0.8 3.5 krpm/30” 110°C/60” ~550nm 27 +0.15 105°C/60” Not Used AZ300MIF Approx. 20sec

(not thoroughly calibrated)

  • Replaced UVN2300, not identical. Dev time was 55-60 sec.

Other Lithography Processes (ASML DUV)