Protocol for Soft-Photolithography

Nadia ZAARI04/01/02

Biomimetic Materials Engineering Lab

Pr. Joyce WONG

Protocol for Soft-Photolithography

Opto-Electronic Component Fabrication Facility

Photonics Bldg Rm 815

Contact: Paul Mak (3-8869, )

MATERIALS:

3 inches wafers
SU-8 photoresist from Microchem (
4*4 inches glass slide
Mask printed on a transparency (high resolution printing)

Printing company

- You can ftp the Freehand file and fax the order. Ask for negative transparency and for 5040dpi. The cost is $50 and it takes about one day.

- BFS printers. They can do up to 3280 dpi but it’s cheaper: around $20.

Contact person is: Richard Crear at 617 850 6909

-Note: The features that you would like to have indented in PDMS should be white while everything else should be black. Fit design inside a circle the same size as the Si wafter you will be using.

- A cheaper printing company is which offers 8000 dpi at $44 (25 +15 shipping)

Developer

-PGMEA : polyethylene glycol methyl ether from Sigma

-Diacetone alcohol: 4 hydroxy-4-methyl-2-pentanone from VWR (product#: 31672)

Fluorosilane: (Heptafluoropropyl)trimethyl-silane from Aldrich (Product#:43,104-4)

1. COATING

Place the 3 inch wafer securely on the spin coater
Adjust the spin speed to 100 rpm and the time to 40s
Cover about ¾ of the wafer with SU8.

Spread cycle

Let the SU8 flow for a couple of minutes (SU8 is very viscous)
Ramp the spin coat from 100rpm to 500 rpm for 5 s (100rpm/s acceleration)
Spin coat at 500 rpm for 5 sec

Spin cycle

Spin coat to final speed for 30s. The final speed depends on the desired thickness.

Ex for SU8-50 to get 100m thick layer:

Use a final spin speed of 1000rpm

Product

/ Thickness
in m / Final spin
Speed in rpm
SU8-2002 / 1.5 / 3000
2 / 2000
5 / 1000
SU8-2005 / 5 / 3000
7 / 2000
15 / 1000
SU8-2010 / 10 / 3000
15 / 2000
30 / 1000
SU8-2025 / 25 / 3000
41 / 2000
75 / 1000
SU8-2035 / 35 / 3000
55 / 2000
110 / 1000
SU8-2050 / 50 / 3000
75 / 2000
165 / 1000
SU8-2075 / 75 / 3000
110 / 2000
225 / 1000

Product

/ Thickness in m / Final spin
Speed in rpm
SU8-2 / 1.5 / 3000
2 / 2000
5 / 1000
SU8-5 / 5 / 3000
7 / 2000
15 / 1000
SU8-10 / 10 / 3000
15 / 2000
30 / 1000
SU8-25 / 15 / 3000
25 / 2000
40 / 1000
SU8-50 / 40 / 3000
50 / 2000
100 / 1000
SU8-100 / 100 / 3000
150 / 2000
250 / 1000

Table 1: Thickness vs final spin speed for various SU8 (data taken from Microchem)

2. SOFT BAKE (SB)

The soft bake evaporates the solvent from the photoresist.

The soft bake time depends on the photoresist and its thickness

For best results do the soft bake in 2 steps.

Ex with SU8-50 for 100m thick layer

Step 1: 10 min @ 65C

Step 2: 30 min @ 95C (95C is above the glass transition of the SU8)

Product / Thickness
in m / Step 1:
soft bake time (min) at 65C / Step 2: soft bake time (min) at 95C
SU8-2002 / 1.5 / 60 / 120
2 / 60 / 120
5 / 60 / 120
SU8-2005 / 5 / 60 / 120
7 / 60 / 120
15 / 60 / 120
SU8-2010 / 10 / 60 / 120
15 / 60 / 120
30 / 60 / 120
SU8-2025 / 25 / 1 / 3
41 / 2 / 5
75 / 3 / 9
SU8-2035 / 35 / 2 / 5
55 / 3 / 6
110 / 5 / 20
SU8-2050 / 50 / 3 / 6
75 / 3 / 9
165 / 5 / 30
SU8-2075 / 75 / 3 / 9
110 / 5 / 20
225 / 5 / 45
Product / Thickness in m / Step 1: soft bake
time (min) at 65C / Step 2: softbake
time (min) at 95C
SU8-2 / 1.5 / 60 / 60
2 / 60 / 180
5 / 60 / 180
SU8-5 / 5 / 60 / 180
7 / 120 / 300
15 / 120 / 300
SU8-10 / 10 / 120 / 300
15 / 120 / 300
30 / 180 / 420
SU8-25 / 15 / 120 / 300
25 / 180 / 420
40 / 300 / 900
SU8-50 / 40 / 5 / 15
50 / 6 / 20
100 / 10 / 30
SU8-100 / 100 / 10 / 30
150 / 20 / 50
250 / 30 / 90

Table 2: Recommended soft-bake time from Microchem.

3. EXPOSURE TO UV

Tape the transparency with desired pattern onto the 4*4in glass slide.
Place the mask (= glass slide + transparency) in the mask aligner
Set the power to10mW/cm2
Expose for the desired amount of time.

The exposure time depends on the thickness of the SU8

Ex. for 100m thick layer of SU8-50,

The Expose dose needed is 500-650mJ/cm2 i.e. the exposure time is 50-65s.

Always add more sec. In this case an exposure of 90s will ensure a hard pattern.

Product / Thickness in m / Expose dose
in mJ/cm2 / Time of exposure in s
SU8-2 / 1.5 / 60 / 6
2 / 60 / 6
5 / 60 / 6
SU8-5 / 5 / 60 / 6
7 / 120 / 12
15 / 120 / 12
SU8-10 / 10 / 120 / 12
15 / 120 / 12
30 / 180 / 18
SU8-25 / 15 / 120 / 12
25 / 180 / 18
40 / 300 / 30
SU8-50 / 40 / 300 / 30
50 / 500 / 50
100 / 650 / 65
SU8-100 / 100 / 650 / 65
150 / 675 / 68
250 / 700 / 70
SU8-2002 / 1.5 / 115 / 12
2 / 115 / 12
5 / 115 / 12
SU8-2005 / 5 / 130 / 13
7 / 130 / 13
15 / 180 / 18
SU8-2010 / 10 / 200 / 20
15 / 200 / 20
30 / 210 / 21
SU8-2025 / 25 / 240 / 24
41 / 350 / 35
75 / 600 / 60
SU8-2035 / 35 / 350 / 35
55 / 500 / 50
110 / 650 / 65
SU8-2050 / 50 / 500 / 50
75 / 600 / 60
165 / 650 / 65
SU8-2075 / 75 / 600 / 60
110 / 650 / 65
225 / 675 / 68

Table3: Recommended exposure time.

Note: it is always better to overexpose than underexpose.

For better quality pattern (ie for the photoresist to last longer) add around 50% more sec of exposure time.

Ex with SU8-50 for 100m thick layer:

The table says 65s so expose for 90s.

4. POST EXPOSURE BAKE (PEB)

The PEB selectively cross-links the exposed portions of the SU8.

SU8 is readily crosslined and can result in a highly stress film.

To minimize stress, wafer bowing and and resist cracking, a 2 step process baking is recommended

Product / Thickness
in m / Step 1: PEB time (min) at 65C / Step 2: PEB time (min) at 95C
SU8-2002 / 1.5 / 60 / 60
2 / 60 / 60
5 / 60 / 60
SU8-2005 / 5 / 60 / 60
7 / 60 / 60
15 / 60 / 60
SU8-2010 / 10 / 60 / 120
15 / 60 / 120
30 / 60 / 120
SU8-2025 / 25 / 1 / 3
41 / 1 / 3
75 / 1 / 7
SU8-2035 / 35 / 1 / 3
55 / 1 / 5
110 / 1 / 10
SU8-2050 / 50 / 1 / 5
75 / 1 / 7
165 / 1 / 12
SU8-2075 / 75 / 1 / 7
110 / 1 / 10
225 / 1 / 15

Product

/ Thickness in m / Step 1: PEB
time (min) at 65C / Step 2: PEB
time (min) at 95C
SU8-2 / 1.5 / 60 / 60
2 / 60 / 60
5 / 60 / 60
SU8-5 / 5 / 60 / 60
7 / 60 / 60
15 / 60 / 120
SU8-10 / 10 / 60 / 120
15 / 60 / 120
30 / 60 / 180
SU8-25 / 15 / 60 / 120
25 / 60 / 180
40 / 240
SU8-50 / 40 / 2 / 4
50 / 2 / 5
100 / 30 / 10
SU8-100 / 100 / 3 / 10
150 / 12 / 15
250 / 15 / 25

Table 4: Recommended PEB time from Microchem.

5. DEVELOP

2 ways of doing that:

Using microchem’s SU-8 developer.

It is the easiest way since it only requires immersion of the wafer in the developer and a final rinse with Isopropanol (IPA)

Using Polyethylene glycol methyl ether (PGMEA), Isopropanol (IPA), Acetone and DI water

- Perform a dip in PGMEA for 3-5 min in Petri dish and agitate manually the dish

- Perform a quick dip (15s) in petri dish with IPA

- Perform a dip in PGMEA for 1-2 min in Petri dish and agitate manually the dish

- Perform a quick dip (15s) in petri dish with IPA

- Perform a dip in PGMEA for 3-5 min in Petri dish and agitate manually the dish

- Hold the wafer above a beaker and squirt acetone for 30 sec.

- Put wafer back in PGMEA. A white skum (it’s the SU8 that comes off) will appear.

- Hold the wafer above the beaker and rinse with IPA

- Squirt acetone until complete removal of SU8.

- Rinse with DI water before acetone dries on Si wafer and leaves “rainbow colors” on it.

Note: If the SU8 isn’t completely removed, a white skum will appear when rinsing with DI water.

Using Diacetone

-Dip in diacetone for a few minutes.

-Rinse with acetone.

-Rinse with water.

-Repeat as necessary.

6. CHARACTERIZATION

Follow the instructions on the Alpha-step with SHS24 method loaded.

Always perform a step down with thick SU8 coating.

Step down means the profiler goes from a high level to a low level.

The reason is that SU8 is pretty hard and when above 80m of thickness, the profiler will tend to crash on the thick layer.

Use a 100mm petri dish to store the wafer.

7. TREATMENT FOR PDMS COATING

Use ~20 ul of the Fluorosilane (located on the shelf with all the chemicals).

Place the bottle with the chemical inside a designated dessicator together with the wafer to treat.

Pull vacuum for 2-3 hours under the hood if possible.

Note: try to expose as less as possible the wafer to ambient air (if you open the petri dish with the wafer, there will likely be contamination from the particles floating in the air and therefore a lower success in patterning PDMS).