• NMOS Fabrication Process Description


  •   
  • FileName: nmos_fab_process_desc.pdf [preview-online]
    • Abstract: 1NMOS Fabrication Process DescriptionLast modified by Alex Chediak on March 6, 2000. Please send comments and suggestions to him.Process Flow (Process Overview)Week 1: Starting MaterialsWeek 2: Initial Oxidation - 5200 Å

Download the ebook

1
NMOS Fabrication Process Description
Last modified by Alex Chediak on March 6, 2000. Please send comments and suggestions to him.
Process Flow (Process Overview)
Week 1: Starting Materials
Week 2: Initial Oxidation - 5200 Å
Week 3: Active Area Photolithography
Week 4: Gate Oxidation - 800 Å
Week 5: Poly-Si Deposition - In Microlab
Week 6: Gate Photolithography
Week 7: Source-Drain Deposition (N+)
Week 7: Source-Drain (N+) Drive and Intermediate Oxidation
Week 8: Contact Cut
Week 9: Metallization
Week 10: Metal Definition
Week 1: Starting Materials. (No lab this week.)
• Wafers
1. 3" p-type silicon wafers with a resistivity of 14-16 ohm-cm and crystal
orientation. In addition to work wafers, each section will receive one wafer to be used as
a control during week #4.
2. Blanket Implant: 3.0x1012 /cm2, B11, 60 KeV.
• Masks (4" x 4" Chrome Plates)
1. ACTV = Defines the Active Area (Dark Field)
2. POLY = Defines the Gate (Light Field)
3. CONT = Defines the Contacts (Dark Field)
4. METL = Defines the Metal (Light Field)
Week 1
2
Week 2: Initial Oxidation - 5200 Å. (Performed by TA.)
5. Standard clean your work wafer
1. Mix piranha solution as follows: (Piranha is used to remove metals and organics.)
1. Measure out 5 parts of H2SO4 in Pyrex beaker
2. Very slowly add 1 part of H2O2. Note: This mixture is self heating. When cool it
may be refreshed by very slowly adding 1 part of H2O2.
2. Immerse wafers in piranha solution for 10 minutes. Piranha removes organic
contaminants by oxidizing (burning) them, and metals by forming soluble complexes.
3. Pour 500 ml of 10:1 BHF into a plastic beaker.
4. Rinse the wafers off in DI water for 10 seconds in each of 3 dedicated rinse beakers
successively, then spray rinse.
5. Dip in 10:1 BHF for 20 seconds. This "HF dip" removes the native oxide.
6. Rinse in DI water for 10 seconds in each of 3 dedicated rinse beakers successively, spray
rinse
7. Place in Polymetrics rinser until resistivity meter indicates 10-15 M or for 10 minutes.
Note: Clean, bare silicon is hydrophobic and metallic in appearance.
8. Spin dry.
6. Oxidize wafers at 1050 °C for 5-70-5 minutes (dry-wet-dry) O2.
7. Measure oxide thickness. It should be approximately 5200 Å.
Week 2
3
Week 3: Active Area Photolithography
8. Standard resist coating
1. If practical, spin wafers immediately after high temperature treatment. Otherwise dehydrate
the wafers in an oven for 20 minutes at 120 °C.
2. Place wafers in HMDS vapor for 2 minutes or more minutes. HMDS promotes adhesion
between oxides (e.g., native silicon dioxide) and photoresist.
3. Lift the black lever directly under the front of the spinner.
4. Make sure the vacuum on the hot plate is not being used. If it is then wait until the vacuum
is turned off. Using the spinner when the hot plate vacuum is being used can cause the
wafer to fly off the spinner and break!
5. Verify the spinner is set to the right spin speed and time by loading and spinning a dummy
wafer. The wafer should spin at 3000 RPM for 30 seconds.
6. To start the spinner use the large button on the foot pedal. To stop the spinner use the
small button on the foot pedal.
7. Get a bottle of OCG 825 photoresist from the blue box next to the spinner.
8. Dispense one eye dropper of resist on the wafer and start a spin cycle.
9. Soft bake at 90 °C for 1 minute on the hot plate. Soft baking evaporates most of the solvent
in the PR so that it is not sticky.
1. Turn on the hot plate vacuum (black lever) if the spinner is not being used.
2. Place the wafer on the hot plate.
3. After the wafer is sucked down to the hot plate, turn OFF the hot-plate vacuum so that
the spinner may be used.
4. Center the wafer on the hot plate with a metal tweezer. (Placing the wafer on the hot
plate without the vacuum on will cause the wafer to slide all over the hot plate. If the
wafer slides off the wafer may break.)
Week 3
4
9. Standard photomasking: Mask #1 (ACTV)
1. Following Kasper Mask Aligner Instructions, align wafers to mask and expose. Typical
exposure times are 30 to 60 seconds. Shorter times are needed when the lamp is newer
and for more-reflective substrates such as aluminum. Cover wafer box with Al foil when
carrying exposed wafer into etching room.
2. Pour 500 ml of premixed KTI 934 1:2 developer solution into a beaker.
3. Turn off the white lights in the etching room, develop and etch only in yellow light. The
yellow filters filter out shorter, more energetic wavelengths of light (e.g. green, blue,
violet, and UV), which could cause exposure of the entire wafer.
4. Dip the exposed wafer into the developer for typically 40-60 seconds (ask the TA for the
current optimal exposure time; it takes more time if the developer has been used more)
and agitate slowly. Observe the reddish liquid forming near the wafer. It is the dissolved
photoresist.
5. Rinse in DI water for 10 seconds in each of 3 dedicated rinse beakers successively, then
spray rinse.
NOTE: DO NOT use the same set of 3 rinse beakers for more than one kind of rinse. For
example, do not use the developer rinse beakers for rinsing after an HF etch. Instead prepare
another set of 3 dedicated rinsing beakers.
6. Spin dry. The spin dryer does NOT have a vacuum interlock so turning off the vacuum
before the spinner has come to a rest will cause the wafer to fly off and break. This event
is a major source of wafer breakage in the 143 lab. DO NOT LET THIS HAPPEN TO
YOU! BE PATIENT!
7. Inspect under microscope with yellow filter. Measure line widths of test pattern [1] with
filar eye piece. Record data.
8. If not completely developed repeat Step 4 for 15 seconds, then Steps 5-7.
9. Hard bake the photoresist in an oven for 20 minutes at 120 °C. Hard baking cross-
polymerizes the PR polymer, making the the photoresist physically hard, more adhesive,
and less permeable to chemicals.
10. Label wafers with diamond scribe. Label using small letters near the flat. Do not scrib off
the edge of the wafer as this will cause the wafer to break. DO NOT LABEL YOUR
WAFER ACROSS THE CENTER AS IT WILL DESTROY YOUR DEVICES.
Week 3
5
11. Oxide etching and inspection
1. Pour 500 ml of buffered 5:1 HF (BHF) into a plastic beaker. Buffered HF is a mixture
of NH4F (ammonium fluoride) and HF 5:1. Its etch rate of thermal SiO2 is ~1000
Å/min. at 25 °C. BHF is used rather than plain diluted HF because the buffer keeps the
strength and thus the etch rate closer to constant.
2. Determine the etching time according to the oxide thickness to be etched plus a 10 %
overetch. The overetch is performed for process latitude (i.e., the oxide thickness and the
etch rate both may vary across a wafer and among wafers).
3. Dip wafers in water (Rinse tank 1 is acceptable) to wet the surface. Because of the
surface tension of BHF, air bubbles can sometimes get trapped on the wafer surface if
the surface is completely dry, leading to localized areas where the oxide is not removed.
4. Dip wafer in buffered 5:1 HF for the length of time determined in Step 2. Etching is
complete when the etchant "beads" on bare Si; (i.e., a hydrophobic surface is detected.)
5. Rinse in DI water for 10 seconds in each of 3 dedicated rinse beakers successively, then
spray rinse only (no Polymetrics).
6. Spin dry.
7. Inspect etching for completion under microscope.
12. Do the standard resist strip
1. Dip wafer in acetone for 2 minutes.
2. Rinse in DI water for 10 seconds in each of 3 dedicated rinse beakers successively, then
spray rinse.
3. Spin dry.
13. Measure line widths of test pattern [1] with filar eyepiece in microscope.
Week 3
6
Week 4: Gate Oxidation - 800 Å
Gate oxidation is done in tube #3 (bottom) at 1100 °C following a TCA (C2H3Cl3) clean. Gases
used are N2 and O2.
14. TCA Clean Furnace. (Performed by TA.)
1. When you arrive, the furnace will already be on and stabilized.
2. Confirm temperature controller is set to the required temperature. For 1100 °C the
controllers should be about +30.0, 87.0, +00.0. The center value adjusts the
temperature of the central zone of the furnace. The left and right values set the
temperature of the load and gas-input zones relative to the center region.
3. Check the temperature in the "hot zone" using the thermo-couple.
4. Open the sliding exhaust (scavenger) door on the side of the loading vestibule. Close
the others if they are not being used.
5. Remove the glass end cap from tube (holding it with an insulating glove) and place it
on the fire brick on the counter.
6. Load an empty boat into the "hot zone". Push boat in no faster than 1" every 10
second. Pushing the boat in faster can cause the boat to crack due to thermal stress.
7. Replace the glass end cap. Make sure gas outlet is pointing towards the exhaust
(scavenger) door.
8. Turn on O2 flow at maximum rate (15+ cm, steel ball) for at least 10 minutes. This
requires that the O2 tank is turned on, the regulator is set to 10 psi (MAX), the post-
regulator valve is on, the valve on the flow meter is opened to read 15 cm+, and the
O2 valve at tube 3 is opened. *** This step is critical; if there is not sufficient
oxygen in the tube when the TCA starts flowing, the TCA will not oxidize (burn)
completely and will form soot inside the tube.***
9. Start the TCA bubbling by closing the N2 valve at tube 3, opening the TCA valve at
tube 3, opening the valve to the TCA bubbler, and opening the N2/TCA valve on the
flow controller. Keep this to a low flow rate, say 1-2 on the scale. It is important to
open valves starting at the tube end. Flowing gas into the bubbler vessel with the
tube 3 TCA valve off will pressurize the bubbler vessel, causing it to explode. This
would be very bad. Leaving the tube 3 N2 valve on will cause N2 gas to circumvent
the bubbler vessel. If bright blue flames are visible at the tube inlet, decrease the N2
flow rate. The flames are a result of too much free carbon in the tube, i.e. not enough
O2 to react with the incoming TCA. Too much TCA will cause a small explosion.
(This has happened before. The other end of the tube was blown off.) If no bubbles
are seen, check that the TCA level is above the frosted zone of the N2 tube in the
bubbler vessel.
10. TCA clean for 1 hour at 1100 °C. TCA removes heavy-metal contaminants by
supplying the chlorine to form volatile chlorides: N2 + 2 O2 + C2H3Cl3 -> 2 CO2 + 3
HCl + N2.
11. Turn off N2/TCA flow, and close TCA bubbler valve, and the tube 3 TCA valve. It is
important to close valves starting at the flow meter, and working towards tube 3,
thereby not pressurizing the TCA bubbler.
12. Wait 10 minutes after turning off TCA.
13. Switch gases to N2 = 4 cm and O2 = 1.5 cm for wafer loading. This is 90% N2 and
10% O2.
Week 4
7
15. Standard Clean process wafers and an implanted control wafer from Week 1. Label the
control wafer "OCR DATE", were OCR stands for Oxide Control Wafer, and DATE is
the current date. Make sure not to HF dip for more than 20 seconds.
1. Mix piranha solution as follows: (Piranha is used to remove metals and organics.)
1. Measure out 5 parts of H2SO4 in Pyrex beaker
2. Very slowly add 1 part of H2O2. Note: This mixture is self heating. When cool it
may be refreshed by very slowly adding 1 part of H2O2.
2. Immerse wafers in piranha solution for 10 minutes. Piranha removes organic
contaminants by oxidizing (burning) them, and metals by forming soluble complexes.
3. Pour 500 ml of 10:1 BHF into a plastic beaker.
4. Rinse the wafers of in DI water for 10 seconds in each of 3 dedicated rinse beakers
successively, then spray rinse.
5. Dip in 10:1 BHF for 20 seconds. This "HF dip" removes the native oxide.
6. Rinse in DI water for 10 seconds in each of 3 dedicated rinse beakers successively, spray
rinse
7. Place in Polymetrics rinser until resistivity meter indicates 10-15 M or for 10 minutes.
Note: Clean, bare silicon is hydrophobic and metallic in appearance.
8. Spin dry.
16. Wafer Loading and Oxidation
1. Confirm that the gases are set to N2 = 4 cm and O2 = 1.5 cm. This is 90% N2 and 10%
O2.
2. Remove the end cap from tube (holding it with an insulating glove) and place it on
the fire brick on the counter.
3. Attach the cylindrical carrier (elephant) to the end of tube.
4. Pull the boat into the carrier slowly (1" every 10 seconds to avoid thermal shock
which could break the boat).
5. Detach the cylindrical carrier and carefully push the boat onto the half-shell boat
carrier and let the boat cool.
6. Load wafers into the boat very carefully. Make sure the pair of grooves you intend to
use to hold the wafer slant in the proper direction (facing each other). This step has
broken many wafers.
7. Pull the boat into cylindrical carrier.
8. Attach the cylindrical carrier to the end of the tube and push the boat into the tube
very slowly (1" every 10 seconds to avoid thermal shock which could break the
boat). Push the boat in a total of 36 inches which should take 6 minutes. Time
yourself to be sure.
9. Remove the cylindrical carrier and place the end cap on the tube, making sure gas
outlet is pointing towards the exhaust (scavenger) door.
Week 4
8
10. Change the gas flows, maintaining the following sequence: O2 to 12 cm and N2 to 0
cm.
11. Time for 34 minutes oxidation time. Dry (rather than wet) oxidation is done for a
higher quality oxide with a more controlled thickness and a better oxide-silicon
interface.
12. Turn off O2 valve at the flowmeter, then the O2 valve at tube 3. Set N2 flow to 4 cm.
13. Anneal in N2 for 10 minutes to allow the silicon atoms to diffuse and heal some of
the damage done during oxidation. Pull wafers out in N2 as described in step #8
above.
14. Remove the end cap, attach the cylindrical carrier to the end of the tube and pull the
boat out slowly (1" every 10 seconds to avoid thermal shock which could break the
boat) into the carrier.
15. Place carrier with boat on table and let it cool.
16. Remove the wafers very carefully! This step has broken the MOST wafers!
17. Pull the boat into the half-shell carrier and unload the wafers.
18. Put the boat back into the end of tube (not all the way into the center) and replace end
cap.
19. Return the push rods to the appropriate storage tubes.
20. Turn off all gases.
21. Never mix quartz ware between tubes, to avoid cross contamination.
17. Thickness and resistivity measurement
1. Check oxide thickness on control wafer with the NanoSpec.
2. Etch off oxide completely in buffered HF.
3. Measure resistivity with the Four-Point Probe and use this to estimate channel doping
concentration.
Week 4
9
Week 5 Poly-Si Deposition. (Performed by TA in Microlab.)
18. Standard clean wafers.
1. Mix piranha solution as follows: (Piranha is used to remove metals and organics.)
1. Measure out 5 parts of H2SO4 in Pyrex beaker
2. Very slowly add 1 part of H2O2. Note: This mixture is self heating. When cool it
may be refreshed by very slowly adding 1 part of H2O2.
2. Immerse wafers in piranha solution for 10 minutes. Piranha removes organic
contaminants by oxidizing (burning) them, and metals by forming soluble complexes.
3. Pour 500 ml of 10:1 BHF into a plastic beaker.
4. Rinse the wafers off in DI water for 10 seconds in each of 3 dedicated rinse beakers
successively, then spray rinse.
5. Dip in 10:1 BHF for 20 seconds. This "HF dip" removes the native oxide.
6. Rinse in DI water for 10 seconds in each of 3 dedicated rinse beakers successively, spray
rinse
7. Place in Polymetrics rinser until resistivity meter indicates 10-15 M or for 10 minutes.
Note: Clean, bare silicon is hydrophobic and metallic in appearance.
8. Spin dry.
19. Deposit 3500 Å Phosphorus-doped polysilicon. SiH4 flow = 100 SCCM; PH3 flow = 1
SCCM. The silane supplies the silicon, while the phosphine provides phosphorus for n-
type doping. Wafers should be loaded in the center boat for tylan11. Try to avoid the last
work position in the center boat and you can also use the first 5 work positions in the rear
boat. Do not remove the 4" dummies and load the 3" wafers in the same slots, with their
backs against the 4"dummies. The 3"wafers stay nicely and they keep the vertical
position during the deposition. Be sure to include a 3" process monitor on either side of
the 3" work wafers. These control wafers should have a 1000 Å SiO2 on them.
20. Measure poly-Si thickness with the NanoSpec.
Week 5
10
Week 6: Gate Photolithography
21. Apply standard resist coating
1. If practical, spin wafers immediately after high temperature treatment. Otherwise dehydrate
the wafers in an oven for 20 minutes at 120 °C.
2. Place wafers in HMDS vapor for 2 minutes or more minutes. HMDS promotes adhesion
between oxides (e.g., native silicon dioxide) and photoresist.
3. Lift the black lever directly under the front of the spinner.
4. Make sure the vacuum on the hot plate is not being used. If it is then wait until the vacuum
is turned off. Using the spinner when the hot plate vacuum is being used can cause the
wafer to fly off the spinner and break!
5. Verify the spinner is set to the right spin speed and time by loading and spinning a dummy
wafer. The wafer should spin at 3000 RPM for 30 seconds.
6. To start the spinner use the large button on the foot pedal. To stop the spinner use the
small button on the foot pedal.
7. Get a bottle of OCG 825 photoresist from the refrigerator beneath the spinner.
8. Dispense one eye dropper of resist on the wafer and start a spin cycle.
9. Return the bottle of OCG 825 photoresist to the refrigerator beneath the spinner.
10. Soft bake at 90 °C for 1 minute on the hot plate. Soft baking evaporates most of the solvent
in the PR so that it is not sticky.
1. Turn on the hot plate vacuum (black lever) if the spinner is not being used.
2. Place the wafer on the hot plate.
3. After the wafer is sucked down to the hot plate, turn OFF the hot-plate vacuum so that
the spinner may be used.
4. Center the wafer on the hot plate with a metal tweezer. (Placing the wafer on the hot
plate without the vacuum on will cause the wafer to slide all over the hot plate. If the
wafer slides off the wafer may break.)
Week 6
11
22. Standard photomasking: Mask #2 (POLY)
1. Following Kasper Mask Aligner Instructions, align wafers to mask and expose. Typical
exposure times are 30 to 60 seconds. Shorter times are needed when the lamp is newer
and for more-reflective substrates such as aluminum. Cover wafer box with Al foil when
carrying exposed wafer into etching room.
2. Pour 500 ml of premixed KTI 934 1:2 developer solution into a beaker.
3. Turn off the white lights in the etching room, develop and etch only in yellow light. The
yellow filters filter out shorter, more energetic wavelengths of light (e.g. green, blue,
violet, and UV), which could cause exposure of the entire wafer.
4. Dip the exposed wafer into the developer for typically 40-60 seconds (ask the TA for the
current optimal exposure time; it takes more time if the developer has been used more)
and agitate slowly. Observe the reddish liquid forming near the wafer. It is the dissolved
photoresist.
5. Rinse in DI water for 10 seconds in each of 3 dedicated rinse beakers successively, then
spray rinse.
NOTE: DO NOT use the same set of 3 rinse beakers for more than one kind of rinse. For
example, do not use the developer rinse beakers for rinsing after an HF etch. Instead
prepare another set of 3 dedicated rinsing beakers.
6. Spin dry. The spin dryer does NOT have a vacuum interlock so turning off the vacuum
before the spinner has come to a rest will cause the wafer to fly off and break. This event
is a major source of wafer breakage in the 143 lab. DO NOT LET THIS HAPPEN TO
YOU! BE PATIENT!
7. Inspect under microscope with yellow filter. Measure line widths of test pattern [1] with
filar eye piece. Record data.
8. If not completely developed repeat Step 4 for 15 seconds, then Steps 5-7.
9. Hard bake the photoresist in an oven for 20 minutes at 120 °C. Hard baking cross-
polymerizes the PR polymer, making the the photoresist physically hard, more adhesive,
and less permeable to chemicals.
Week 6
12
23. Etch poly-Si
1. Pour 500 ml of BHF into a plastic beaker.
2. Pour 500 ml of a premixed silicon etchant, obtained from the Microlab, into a pyrex
beaker. Etch rate: ~3100 Å/min. Composition: 64% HNO3 / 33% H2O / 3% NH4F.
3. Determine necessary etch time based on polysilicon thickness and add a 10% overetch.
4. Dip wafers in water (Rinse tank 1 is acceptable) to wet the surface. Because of the surface
tension of BHF, air bubbles can sometimes get trapped on the wafer surface if the surface
is completely dry, leading to localized areas where the oxide is not removed.
5. Dip the wafer into the BHF solution for 20 seconds to remove any native oxide, which is
etched more slowly than polysilicon. Rinse the wafers for 10 seconds in DI water in each
of 3 dedicated rinse beakers successively.
6. Immerse wafers in the silicon etchant for the determined etch time. Watch the color
changes as the silicon is etched.
7. Rinse the wafers for 10 seconds in DI water in each of 3 dedicated rinse beakers
successively.
8. Spin dry or continue to etch.
9. Inspect etching for completion under microscope.
24. Etch oxide in active area until clear (~1 minute) in BHF.
1. Pour 500 ml of buffered 5:1 HF (BHF) into a plastic beaker. Buffered HF is a mixture
of NH4F (ammonium fluoride) and HF 5:1. Its etch rate of thermal SiO2 is ~1000
Å/min. at 25 °C. BHF is used rather than plain diluted HF because the buffer keeps the
strength and thus the etch rate closer to constant.
2. Determine the etching time according to the oxide thickness to be etched plus a 10 %
overetch. The overetch is performed for process latitude (i.e., the oxide thickness and the
etch rate both may vary across a wafer and among wafers).
3. Dip wafers in water (Rinse tank 1 is acceptable) to wet the surface. Because of the
surface tension of BHF, air bubbles can sometimes get trapped on the wafer surface if
the surface is completely dry, leading to localized areas where the oxide is not removed.
4. Dip wafer in buffered 5:1 HF for the length of time determined in Step 2. Etching is
complete when the etchant "beads" on bare Si; (i.e., a hydrophobic surface is detected.)
5. Rinse in DI water for 10 seconds in each of 3 dedicated rinse beakers successively, then
spray rinse only (no Polymetrics).
6. Spin dry.
7. Inspect etching for completion under microscope.
Week 6
13
25. Do the standard resist strip
1. Dip wafer in acetone for 2 minutes.
2. Rinse in DI water for 10 seconds in each of 3 dedicated rinse beakers successively, then
spray rinse.
3. Spin dry.
Week 6
14
Week 7a: Source-Drain Deposition (N+)
26. Standard clean wafers without HF dip.
1. Mix piranha solution as follows: (Piranha is used to remove metals and organics.)
1. Measure out 5 parts of H2SO4 in Pyrex beaker
2. Very slowly add 1 part of H2O2. Note: This mixture is self heating. When cool it
may be refreshed by very slowly adding 1 part of H2O2.
2. Immerse wafers in piranha solution for 10 minutes. Piranha removes organic
contaminants by oxidizing (burning) them, and metals by forming soluble complexes.
3. Rinse the wafers off in DI water for 10 seconds in each of 3 dedicated rinse beakers
successively, then spray rinse.
4. Spin dry.
27. Turn on bubbler heater to 98 °C. Turn on heater tape. Check water level.
28. Spin Filmtronics Phosphorosilica spin-on-glass at 3000 RPM for 20 seconds including p-
type test wafer from week 4, step 4. Less than half of a pipette-full of SOG is needed.
29. Bake at 200 °C for 15 minutes in bake oven.
NOTE: Do Not use a Teflon holder for the wafers in the 200 °C oven! The Teflon boats
will flow at this temperature. Place the wafers on glass slides in the furnace.
30. N+ pre-diffusion is done at 1050 °C in furnace tube #2 (center). Set O2 flowmeter to 1.5
cm and N2 to 4 cm. (This corresponds to 10% O2 and 90% N2.) The setting is constant
during push, deposition and pull.
31. Push in wafers at 1 inch/10 seconds.
32. Hold 5 minutes in hot zone.
33. Pull out at 1 inch/10 seconds.
34. Phosphorus glass removal: dip wafers in 10:1 BHF for amount of time determined from
annealed PSG rate (4700 Å/min). Note that wafer may not be hydrophobic due to film
left by Spin-On Glass. Piranha cleaning will remove this film.
35. Rinse in D.I. water for 10 seconds in each of 3 beakers successively, then spray rinse.
36. Spin dry.
37. Measure resistivity on the control wafer with the Four-Point Probe.
Week 7a
15
Week 7b: Source-Drain (N+) Drive and Intermediate Oxidation
38. This step is done at 1050 °C in tube #2. Check gas connection to allow for steam
oxidation, dry O2 and N2 flow. Set dry O2 to 6.5 cm on flowmeter. Set N2 to 0 cm (OFF).
39. Fill bubbler ~80% with D.I. water. Use a clean, well rinsed beaker to fill bubbler.
40. Standard clean wafers, including control wafer used in step 7a..
1. Mix piranha solution as follows: (Piranha is used to remove metals and organics.)
1. Measure out 5 parts of H2SO4 in Pyrex beaker
2. Very slowly add 1 part of H2O2. Note: This mixture is self heating. When cool it may
be refreshed by very slowly adding 1 part of H2O2.
2. Immerse wafers in piranha solution for 10 minutes. Piranha removes organic
contaminants by oxidizing (burning) them, and metals by forming soluble complexes.
3. Rinse the wafers off in DI water for 10 seconds in each of 3 dedicated rinse beakers
successively, then spray rinse.
4. Spin dry.
41. When the water is boiling in the bubbler (approximately 20 minutes) push wafers into
furnace in dry O2 (O2 should already be on to 6.5 cm). N2 should be OFF.
42. Set O2 flowmeter to 2.5 cm and switch valves to wet O2. Wet oxidation time is 12
minutes. Wet oxidation (versus dry) is done because it grows oxide faster. Some of the
single-crystal silicon of the source and drain and the polysilicon forming the gate are
consumed during this oxidation.
43. Turn off wet O2 and turn on N2 to 4 cm on scale.
44. Anneal for 25 minutes in N2.
45. Pull wafers out in N2. Pull out at 1 inch/10 seconds.
46. Check intermediate oxide thickness on control wafer, then etch oxide completely off in
buffered HF and measure resistivity with the Four-Point Probe.
Week 7b
16
Week 8: Contact Cut
47. Apply standard resist coating
1. If practical, spin wafers immediately after high temperature treatment. Otherwise dehydrate
the wafers in an oven for 20 minutes at 120 °C.
2. Place wafers in HMDS vapor for 2 minutes or more minutes. HMDS promotes adhesion
between oxides (e.g., native silicon dioxide) and photoresist.
3. Lift the black lever directly under the front of the spinner.
4. Make sure the vacuum on the hot plate is not being used. If it is then wait until the vacuum
is turned off. Using the spinner when the hot plate vacuum is being used can cause the
wafer to fly off the spinner and break!
5. Verify the spinner is set to the right spin speed and time by loading and spinning a dummy
wafer. The wafer should spin at 3000 RPM for 30 seconds.
6. To start the spinner use the large button on the foot pedal. To stop the spinner use the
small button on the foot pedal to stop the spinner.
7. Dispense one eye dropper of resist on the wafer and start a spin cycle.
8. Soft bake at 90 °C for 1 minute on the hot plate. Soft baking evaporates most of the solvent
in the PR so that it is not sticky.
1. Turn on the hot plate vacuum (black lever) if the spinner is not being used.
2. Place the wafer on the hot plate.
3. After the wafer is sucked down to the hot plate, turn OFF the hot-plate vacuum so that
the spinner may be used.
4. Center the wafer on the hot plate with a metal tweezer. (Placing the wafer on the hot
plate without the vacuum on will cause the wafer to slide all over the hot plate. If the
wafer slides off the wafer may break.)
Week 8
17
48. Standard photomasking: Mask #3 (CONT).
1. Following Kasper Mask Aligner Instructions, align wafers to mask and expose. Typical
exposure times are 30 to 60 seconds. Shorter times are needed when the lamp is newer
and for more-reflective substrates such as aluminum. Cover wafer box with Al foil when
carrying exposed wafer into etching room.
2. Pour 500 ml of premixed KTI 934 1:2 developer solution into a beaker.
3. Turn off the white lights in the etching room, develop and etch only in yellow light. The
yellow filters filter out shorter, more energetic wavelengths of light (e.g. green, blue,
violet, and UV), which could cause exposure of the entire wafer.
4. Dip the exposed wafer into the developer for typically 40-60 seconds (ask the TA for the
current optimal exposure time; it takes more time if the developer has been used more)
and agitate slowly. Observe the reddish liquid forming near the wafer. It is the dissolved
photoresist.
5. Rinse in DI water for 10 seconds in each of 3 dedicated rinse beakers successively, then
spray rinse.
NOTE: DO NOT use the same set of 3 rinse beakers for more than one kind of rinse. For
example, do not use the developer rinse beakers for rinsing after an HF etch. Instead
prepare another set of 3 dedicated rinsing beakers.
6. Spin dry. The spin dryer does NOT have a vacuum interlock so turning off the vacuum
before the spinner has come to a rest will cause the wafer to fly off and break. This event
is a major source of wafer breakage in the 143 lab. DO NOT LET THIS HAPPEN TO
YOU! BE PATIENT!
7. Inspect under microscope with yellow filter. Measure line widths of test pattern [1] with
filar eye piece. Record data.
8. If not completely developed repeat Step 4 for 15 seconds, then Steps 5-7.
9. Hard bake the photoresist in an oven for 20 minutes at 120 °C. Hard baking drives off the
rest of the polymer, making the photoresist physically hard, more adhesive, and less
permeable to chemicals.
Week 8
18
49. Do oxide etch for calculated time. The back side of the wafer should de-wet (metallic in
color) indicating no oxide present. Inspect under microscope.
1. Pour 500 ml of buffered 5:1 HF (BHF) into a plastic beaker. Buffered HF is a mixture
of NH4F (ammonium fluoride) and HF 5:1. Its etch rate of thermal SiO2 is ~1000
Å/min. at 25 °C. BHF is used rather than plain diluted HF because the buffer keeps the
strength and thus the etch rate closer to constant.
2. Determine the etching time according to the oxide thickness to be etched plus a 10 %
overetch. The overetch is performed for process latitude (i.e., the oxide thickness and the
etch rate both may vary across a wafer and among wafers).
3. Dip wafers in water (Rinse tank 1 is acceptable) to wet the surface. Because of the
surface tension of BHF, air bubbles can sometimes get trapped on the wafer surface if
the surface is completely dry, leading to localized areas where the oxide is not removed.
4. Dip wafer in buffered 5:1 HF for the length of time determined in Step 2. Etching is
complete when the etchant "beads" on bare Si; (i.e., a hydrophobic surface is detected.)
5. Rinse in DI water for 10 seconds in each of 3 dedicated rinse beakers successively, then
spray rinse only (no Polymetrics).
6. Spin dry.
7. Inspect etching for completion under microscope.
50. You may find it difficult to tell whether the etch is complete, therefore, remove resist and
inspect wafer again. If the contacts are not clear, etch for an additional 15 seconds and
check again. Contact holes over the silicon substrate (not the poly) appear white when
cleared.
1. Dip wafer in acetone for 2 minutes.
2. Rinse in DI water for 10 seconds in each of 3 dedicated rinse beakers successively, then
spray rinse.
3. Spin dry.
51. Strip oxide from the monitor wafer and measure its sheet resistance using the Four-Point
Probe.
Week 8
19
Week 9: Metallization
52. Standard clean wafers. Do the last dip in 10:1 BHF just before wafer is ready to go into
evaporator to minimize the native oxide at the aluminum-silicon interface.
1. Mix piranha solution as follows: (Piranha is used to remove metals and organics.)
1. Measure out 5 parts of H2SO4 in Pyrex beaker
2. Very slowly add 1 part of H2O2. Note: This mixture is self heating. When cool it
may be refreshed by very slowly adding 1 part of H2O2.
2. Immerse wafers in piranha solution for 10 minutes. Piranha removes organic
contaminants by oxidizing (burning) them, and metals by forming soluble complexes.
3. Pour 500 ml of 10:1 BHF into a plastic beaker.
4. Rinse the wafers off in DI water for 10 seconds in each of 3 dedicated rinse beakers
successively, then spray rinse.
5. Dip in 10:1 BHF for 20 seconds. This "HF dip" removes the native oxide.
6. Rinse in DI water for 10 seconds in each of 3 dedicated rinse beakers successively, spray
rinse
7. Place in Polymetrics rinser until resistivity meter indicates 10-15 M or for 10 minutes.
Note: Clean, bare silicon is hydrophobic and metallic in appearance.
8. Spin dry.
Week 9
20
53. Aluminum Evaporation
1. Turn knob on front of evaporator to "STARTUP". This starts the roughing pump, evacuates
the foreline, and starts the diffusion pump. No valves should be open in this setting. Make
sure to add liquid nitrogen to cold trap after the diffusion pump has warmed up for 20 min.
2. Vent the chamber by turning the knob to "VENT". All the valves should be closed and N2 is
vented into the chamber.
3. After chamber reaches atmospheric pressure, the jar will lift from the steel base. Lift the jar
up and turn the switch back to "STARTUP".
4. Lift stainless steel wafer holder out of the inner glass cylinder ("chimney") and place it on the
table covered with lint-free paper.
5. Loading the sample and supporting accessories.
1. Place a clean glass slide in stand inside chimney for a clear window.
2. Hang 2 clean Al charges (staples) near the middle of the tungsten coil (centered below
the chimney hole). This should result in about 8000 Å of Al (proper placement of the
charges are necessary for good uniformity).
3. Turn shutter knob so that the shutter is covering the charges.
4. Place wafer facing down on top of wafer holder.
6. Place wafer holder back inside the chimney.
7. Wipe stainless steel base and bottom of bell jar with lint free paper soaked with 2-propanol.
8. Lower bell jar and turn switch to "PUMP". Push bell jar cage down and close it all around
until vacuum begins to hold it tight and cannot be lifted. The PUMP setting opens up the
roughing valve to evacuate the chamber, and then switches over


Use: 0.417