Plasma Etch: Polysilicon Etch Overview

Polysilicon Plasma Etching

The following information is provided to SNF users to aid in the determination of appropriate equipment for their etching needs. Standard or typical programs and results are given. The user may want to tailor the programs to fit their specific process. Etch rates and selectivities are given as a starting point only and should not be considered to be current. It is highly recommended that users establish the etch rates for their work by the use of test wafers as close in material and masking pattern as possible to the device wafers to be etched.

Drytek 2

Chemistry:

Fluorine based etching only using Freons and SF6

Undercut:

The undercut during the C2ClF5/SF6 poly-Si and Si etch processes depend on the by-products from resist erosion. If there is no resist present on the wafer surface, the undercut can be as much as 50% of the etch depth. For 50% resist coverage, the uncut is in range of 0.1 um for a 0.5 um etch depth. With this resist coverage the increase in uncut with etch depth is less than linear.
See log book for CF4/O2 process. This is a nearly pure isotropic or conformal etch process with the undercut being equal to the etch depth.

Endpoint Detection:

There is a laser interferometer and chart recorder set up for monitoring endpoint. The etch rate can only be monitored in a 1 cm diameter region in the center of the wafer.

Pre-treatments:

None needed.

Post-treatments:

No special cleaning is usually needed. The sidewall inhibitor polymer is very thin (< few hundred A) and appears to come off during normal cleaning processes, such as a pre-diffusion clean.

Etch Process (typical program):

Pressure	150mTorr
SF6	50 sccm
C2ClF5	50 sccm
RF Power	375 Watts

Etch Rates and selectivities:

Material	Etch Rate
Poly (undoped)	2100A/min
Resist	450A/min
Oxide (thermal)	230A/min
Oxide (LTO,undensified)	200A/min
4% Oxide (LTO, undensified)	245A/min
8% Oxide (LTO, undensified)	340A/min
LTO Densified	215A/min
4% Oxide (LTO, densified)	200A/min
8% Oxide (LTO, densified)	240A/min
Stressed Nitride	650A/min
Low Stress Nitride	670A/min
Si	2600A/min
W	0 A/min
Ti	0 A/min

Drytek 1

Chemistry:

Fluorine based etching only using Freons and SF6.

Pre-treatments:

None needed.
Gold contaminated equipment.

Undercut, Endpoint Detection and Post Treatments:

See section on Drytek 1

Etch Process (typical program):

Pressure	150 mTorr
SF6	50 sccm
C2ClF5	50 sccm
RF Power	60 Watts

Etch Rates and selectivities:

Material	Etch Rate
Poly (undoped)	1800A/min
Resist	270A/min
Oxide (thermal)	155A/min
Oxide (LTO,undensified)	170A/min
4% Oxide (LTO, undensified)	245A/min
8% Oxide (LTO, undensified)	330A/min
LTO densified	175A/min
4% Oxide (LTO, densified)	190 A/min
8% Oxide (LTO, densified)	250 A/min
Stressed Nitride	100 A/min
Low Stressed Nitride	125 A/min
Si	1900 A/min
Al	0 A/min
W	700 A/min
Ti	130 A/min
TiW	230 A/min

Lampoly (Lam Polysilicon Etcher)

Features:

Cassette to cassette loading
Low pressure, separate bias control, high density plasma for sub-half micron etching.
Mechanical clamp chuck
Helium backside cooling
No metals
Polysilicon main etch at 320 nm/min
Oxide main etch at 15 nm/min.

Mask Layer:

Photoresist is normally used as the etch mask. The resist selectivity is not great. For G-line resist the selectivity is around 3 for the std poly-Si etch process.

Endpoint Detection:

The LAM has two optical sensors used in endpoint detection. Channel A detects Si (405nm) and Channel B (520 nm) detects CO. Recipe 2 is set up for endpoint detection. There are five parameters which must be set for endpoint detection. See example recipe or manual for more details parameter values.

Post-treatments:

10 second 50:1 HF dip followed by 30 min of hot Piranha. With a HBr/Cl2 etch process the sidewall inhibitor layer is oxide-like and not polymeric. This means that HF is normally used to remove the sidewall layer this process chemistry.

Pre-treatments:

None needed.

Etch Process (typical program):

Recipe 1 (timed) or

Recipe 2 (endpoint)

Operation	Breakthrough	Main Etch	Over etch
Pressure	13 mT	10 mT	15 mT
Cl2	--- sccm	40 sccm	--- sccm
HBr	--- sccm	100 sccm	50 sccm
O2 (20%)	--- sccm	5 sccm	5 sccm
CF4
He
C2F6	100 sccm	--- sccm	--- sccm
RF Top	250 W	250 W	250 W
RF Bottom	40 W	60 W	45 W

Etch Rates and Selectivities

Operation	Etch Rate
Break Through	Poly 207A/min Ox 185A/min
Main Etch	Poly 3627A/min Ox 108A/min
Over Etch	Poly 485A/min Ox 10A/min

Post-treatments:

10 second 50:1 HF dip to remove polymer before resist strip.

P5000etch

Mask Layer:

Photoresist is normally used as the etch mask. The resist selectivity is not great. For G-line resist the selectivity is around 3 for the std poly-Si etch process.
Oxide or nitride (not resist) mask needed for trench etch process.

Endpoint Detection:

Chamber C has a dedicated optical fiber and spectrometer for endpoint detection. For poly-Si etching the 470.5 nm line is monitored. The end point algorithm program has about 15 parameter that can be set. Users normally use one of the already set-up programs, such as su_poly.alg.

Post-treatments:

10 second 50:1 HF dip followed by 30 min of hot Piranha. With a HBr/Cl2 etch process the sidewall inhibitor layer is oxide like and not polymeric. This means that HF is normally used to remove the sidewall layer.

Etch Process (typical program):

CH.C POLY ETCH

Operation	Breakthrough	Main Etch	Over Etch
Pressure	100 mT	100 mT	100 mT
CF4	35 sccm	--- sccm	--- sccm
Cl2	--- sccm	20 sccm	10 sccm
HBr	--- sccm	20 sccm	30 sccm
HeO2	--- sccm	--- sccm	6 sccm
RF power	250 W	200 W	90 W

Endpoint program: su_poly.alg

Etch Rates and Selectivities:

Operation	Etch Rate	Selectivity
Main Etch	Poly 3000A/min	Poly:PR >3:1 Poly:Tox 10:1
Over Etch	Poly 1500A/min	Poly:Tox 50:1

MRC

Chemistry:

Fluorine based Si etching only using Freons and SF6. Ar sputtering can also be used.

Endpoint Detection:

None

Pre-treatments:

None needed.

Gold contaminated equipment.

Etch Process (typical program):

Pressure	3
SF6	5
F115	5
RF Forward	100
RF Vpeak	550

Post-treatments:

None needed.

_______________________________________________________________
Stanford Nanofabrication Facility
webmaster@snf.stanford.edu
Last Modified 08/29/2003