Specific Process Knowledge/Etch/Etching of Silicon Oxide/SiO2 etch using RIE1 or RIE2

Feedback to this page: click here
This page is written by Berit Herstrøm @ DTU Nanolab (BGHE) if nothing else is stated

Both RIE's (RIE1 and RIE2) for silicon based etching has been decommissioned

RIE (Reactive Ion Etch) can be used for etching silicon oxide. The etch is anisotropic with vertical or angled sidewalls depending on the process recipe and the masking material and geometry. The SiO₂ is etched by flour radicals assisted by ion bombardment.

The substrate:

Must be a wafer of 4" or below or small pieces, not higher than about 2mm. The system can be set up for 6", please ask 2 weeks in advance.

The silicon can be masked by these materials:

• Photoresist
• E-beam resist
• Silicon
• Silicon Nitride
• Aluminium
• Other metals if they cover less than 5% of the wafer area.

Most used recipes:

QC results

Quality Controle (QC) for RIE2 - oxide etch

The QC procedure for RIE2 The newest QC data for RIE2 QC Recipe: QCOXIDE CF4 flow 14 sccm CHF3 flow 26 sccm Pressure 100 mTorr RF-power 60 W Etch Load 100% QC limits RIE2 Etch rate in Si 6 - 17 nm/min Non-uniformity <±13.5%

Resist as masking material:

Recipe:	1SiO2mre
Masking material	Photoresist
CF4 flow	14 sccm
CHF3 flow	26 sccm
Pressure	100 mTorr
RF-power	60 W

	Expected results in RIE1 (has been decommissioned)	Expected results in RIE2
Etch rate in SiO2	~20-30 nm/min	~20-30 nm/min
Etch rate in Si3N4	?	?
Etch rate in P-Si (when used as mask)	~30 nm/min	~30 nm/min
Etch rate in Si (etching down to Si)	Expected <2 nm/min	Expected <2 nm/min

mre:mask resist

PolySilicon as masking material

Recipe:	1SiO2msi
Masking material	Silicon
CF4 flow	8 sccm
CHF3 flow	40 sccm
Pressure	150 mTorr
RF-power	150 W

	Expected results in RIE1 (HAS BEEN DECOMMISSIONED)	Expected results in RIE2
Etch rate in SiO2	~70-80 nm/min	~70-80 nm/min
Etch rate in Si3N4	?	?
Etch rate in P-Si (when used as mask)	~20-30 nm/min	~20-30 nm/min
Etch rate in Si (etching down to Si)	?	?
Images	Etches with p-Si as masking material	Etches with p-Si as masking material

msi:mask silicon

The recipe: 1SiO2bre (se below) can also be used with p-Si as masking material.

Burned resist as masking material (developed for waveguides by Haiyan Ou @DTU Photonics):

Recipe:	1SiO2mbr
Masking material	Burned resist (250oC 1h)
CHF3 flow	84 sccm
Pressure	60 mTorr
RF-power	200 W

	Expected results in RIE1 (HAS BEEN DECOMMISSIONED)	Expected results in RIE2
Etch rate in SiO2	~120 nm/min	~120 nm/min
Etch rate burned resist	?	?
Etch rate in Si3N4	?	?
Etch rate in P-Si (when used as mask)	? nm/min	? nm/min
Etch rate in Si (etching down to Si)	?	?
Images	Etches with p-Si as masking material Etches with burned resist as masking material	Etches with burned resist as masking material

mbr:mask burned resist

A survey of the parameter space using photoresist as masking material

A DOE has been performed. A short summery is given here: (in progress)

RIE1: RIE1 has been decommissioned but the results are probably similar on RIE2

A Design Of Experiment (DOE) has been performed. A screening experiment: full factorial with to levels: 9 different runs and 3 repetitions. Results from a screening experiment should be taken "with a gain of salt" meaning use it only as a rough guide line and test your recipe before use.

The Design of the experiment Recipe: Total flow rate (CHF3 flow + CF4 flow) 40 sccm CHF3 flow rate/Total flow rate 65-75% Pressure 100-120 mTorr RF-power 45-65 W	Etch rate Etch rate as a function of power and pressure at 70% CHF3. The etch rate decreases a little as a function of %CHF3.

Selectivity to photoresist

Selectivity as a function of power and pressure at 65% CHF3. Selectivity as a function of power and pressure at 70% CHF3. Selectivity as a function of power and pressure at 75% CHF3.

Etch profile angle <br\> The etch profile was also examined but there was not found any significant correlation between the profile angle and the process parameters.