Specific Process Knowledge/Thin film deposition/Deposition of Silicon Oxide

Feedback to this page: click here

All contents by DTU Nanolab staff.

Deposition of Silicon Oxide can be done with either LPCVD, PECVD, by sputtering, by e-beam evaporation, or by ALD. You can also make a silicon oxide layer by growing a thermal oxide in a hot furnace but that requires a silicon surface as a starting point.

Deposition of Silicon Oxide using LPCVD

The LPCVD oxide you can deposit at DTU Nanolab is called TEOS oxide. It can be made in the LPCVD TEOS furnace. It is a batch process meaning you can run a batch of 13 wafers at a time. The deposition takes place at temperatures of 725 degrees Celsius. The TEOS oxide has good step coverage and hole filing/covering properties and the film thickness is very uniform over the wafer. We have two standard TEOS processes: One for depositing standard layers ~(0-1.5 µm) and one for deposition thick layers ~(1.5µm-4µm). The TEOS oxide has a dielectric constant very close to the one for thermal oxide (3.65 for TEOS).

Deposition of Silicon Oxide using LPCVD TEOS

Deposition of Silicon Oxide using PECVD

PECVD oxide can be deposited in one of the PECVD systems. You can run 1-3 wafers at a time depending on which one of the PECVD's you use. The deposition takes place at 300 degrees Celcius. This can be of importance for some applications but it gives a less dense film and the oxide is expected to have some hydrogen incorporated. The step coverage and thickness uniformity of the film is not as good as for the LPCVD TEOS oxide. PECVD oxide has excellent floating properties when doped with boron and/or phosphorus. Then it can be used eg. as top cladding for waveguides or encapsulation of various structures/components. In one of our PECVD systems (PECVD3) we allow small amounts of metal on the wafers entering the system, this is not allowed in the LPCVD and in PECVD4.

Deposition of Silicon Oxide using PECVD

Deposition of Silicon Oxide using sputter deposition

At DTU Nanolab you can also deposit silicon oxide using the Sputter-System (Lesker) or the Sputter-System Metal Oxide(PC1). An advantage here is that you can deposit on many kinds of material.

Formerly we also had the possibility sputter silicon oxide using the IBE Ionfab300. You can read more about that here.

Deposition of Silicon Oxide using ALD

Thin films of silicon oxide up to 50 nm can also be deposited in the ALD2. The ALD2 uses the plasma source and can therefore only deposit on one wafer at a time. The deposition takes place at 300 ^oC, where the growth rate is 0.1222nm on flat samples. It is also possible to deposit uniform layers on high aspect ratio structures with a growth rate of 0.1629 nm/cycle.

Deposition of Silicon Oxide using ALD2

Deposition of Silicon Oxide using e-beam evaporation

It is possible to e-beam evaporate silicon dioxide at Nanolab using the E-beam evaporator (10-pockets). You can use silicon dioxide pellets as a starting point or silicon with an oxygen flow - in the latter case we expect the resultant films to be oxygen poor. As with sputtering you can deposit on almost any material. In e-beam evaporation the deposition is line-of-sight and will be suitable for lift-off. However for 8" wafers the system is not optimized for lift-off on the full diameter of the wafer.

Wet SiO2 growth

Link to risk assessment and procedure in Labmanager (password needed)

Wet SiO2 growth using hot HNO3.

Done in fume hood 1 or 2 in D-3. Growth rate is 1,5 - 2 nm on 10 min

Training and risk assessment always needed

Comparison of the methods for deposition of Silicon Oxide

	LPCVD(TEOS)	PECVD	Sputter System Lesker	Sputter-system Metal-Oxide(PC1)	E-beam evaporation (E-beam evaporator (10-pockets))	Wet SiO2 growth in hot HNO3
General description	Low Presure Chemical Vapor Deposition TEOS gives a good quality SiO2 and is a batch process.	Plasma Enhanced Chemical Vapor Deposition has the advantage that a silicon oxide and be deposited with a quit high deposition rate at a rather low temperature	Sputter deposition: can be done on top of a large range of materials	Sputter deposition: can be done on top of a large range of materials.	E-beam evaporation: line-of-sight deposition on top of a large range of materials.	Wet SiO2 growth using hot HNO3. Done in fume hood 1 or 2 in D-3. Training and risk assessment always needed
Stochiometry	SiO₂	Si_xO_yH_z Can be doped with boron, phosphorus or germanium		Slightly O-poor as deposited non-reactively (O:Si=64:36), may be tunable if reactively sputtered with O₂ (see acceptance test results here)	to be verified; may depend on O₂ flow	SiO2
Film thickness range	~300 nm - 4 µm	~40 nm - 30 µm	Thin layers (up to 200-300 nm)	Thin layers (up to 200-300 nm)	Thin layers (up to 100 nm)*	1,5-2 nm after 10 min.
Process Temperature	725 °C	300 °C	Room temp	Room temp to 600 °C	Room temp to 250 °C	80°C
Step Coverage	Excellent. Very high surface mobility. Deposition on both sides of the substrate.	Less good When doped with phosphorus and/or Boron the oxide can float at about 1000 °C in a wet oxidation. Deposition on one side of the substrate	Not Known Deposition on one side of the substrate	Medium. Perhaps use of HIPIMS can improve step coverage (requires significant process development) Deposition on one side of the substrate	No step coverage expected unless using tilt holder, in which case step coverage can be very good and can be tuned with the tilt angle. Deposition on one side of the substrate	Not known
Film Quality	Less than thermal oxide. Annealing makes it more dense. Few defects	Less dense film Incorporation of hydrogen in the film
Substrate size / Batch size	1-13 100 mm wafers	Several small samples 1-7 50 mm wafers 1-3 100 mm wafers 1 150 mm wafer	Pieces or 1x4" wafer or 1x6" wafer	Many smaller pieces or up to 10x 4" or 6" wafer	Beaker depended
Allowed materials	Silicon wafers with layers of silicon oxide or silicon (oxy)nitride Quartz wafers	Silicon with layers of silicon oxide or silicon (oxy)nitride Quartz IIIV materials (in PECVD2) Small amount of metal (in PECVD3)	Silicon Silicon oxide Silicon nitride Silicon (oxy)nitride Photoresist PMMA Mylar SU-8 Metals Carbon	Almost any that will not degas and is not very poisonous See cross-contamination sheet	Acids react with a number of metals to produce hydrogen which, in contact with the air, may cause explosions

* If you wish to deposit more than 100 nm, please talk to responsible staff or write to thinfilm@nanolab.dtu.dk