Deposition and Process Flow Review Flashcards
What is unique about the conductivity of semiconductors?
A material or device whose conductivity can be changed by orders of magnitude. This can be accomplished by:
- temperature
- light
- doping
- electric fields
This property allows semiconductors to be made into a switch.
What is the difference between the conductivity of metals and semiconductors?
In metals, the conductivity decreases with temperature (due to increasing electron-phonon collisions). In semiconductors, the conductivity increases with temperature since the number of charge carriers increases.
What are the two parts of semiconductor processing?
- Patterning - making the small, 3-D, interconnected devices made up of semiconductors, insulators, and metallic interconnects
- Doping (selectively)
What is the difference between PVD and CVD?
In PVD, there is no chemical reaction. In CVD, there is a reaction between the substrate and other reactants.
Why was evaporation the preferred method of PVD? Why did that change?
Evaporation was preferred over sputtering since the vacuum is higher (no gas is added) which decreased the possibility of contamination. However, sputtering has better step coverage so it was adopted as the preferred method after the purity of sputtering was improved.
What are the advantages and disadvantages of e-beam evaporation?
Advantages: eliminates possibility of crucible contamination, can heat more materials
Disadvantage: Produces x-rays and secondary electrons that can damage the oxide (especially by creating trapped charges) (this can be solved by annealing)
What method of PVD has the best step coverage? Why?
Sputtering > Evaporation, since the mean free path is lower.
Evaporation has a ‘line of sight’ approach that leads to poor step coverage.
What are the methods of heating the metal for evaporation?
- Current through the crucible - but if the boiling point of the charge is high (eg, W) the crucible may also evaporate
- Inductive heating - reduces but does not eliminate crucible contamination
- E-beam evaporation - no contamination, but x-rays and secondary electrons can damage the wafer
How does sputtering work?
The target (source) is negatively charged/cathode. Plasma from an ionized inert gas (usually Ar) provides positively charged ions that are attracted to the source at high energy, ‘knocking off’ atoms that are deposited on the wafer.
What is the anode and cathode in sputtering?
Target (source) = cathode (negatively charged)
Wafer = can be anode, can be uncharged
What is heated/cooled in sputtering?
The cathode/target/source is cooled since the high-energy collisions produce heat.
The anode/wafer may be heated to spread the deposited material, or cooled.
Cooling the wafer reduces stress between the wafer and the deposited material due to thermal mismatch.
What is the difference between DC and RF sputtering?
DC sputtering is used to deposit metals, RF is used for insulators.
If DC sputtering was used for insulators, the positive charges would build up on the target (capacitative charging) and the process would end.
In RF sputtering, the target attracts positive ions and electrons during the two parts of the cycle. The electrons are more mobile and when they build up, the accumulation of charge is known as ‘self bias’ since it increases the attraction of positive charges during the second half of the cycle.
Compare step coverage in sputtering and evaporation
Evaporation - poor step coverage due to high vacuum, which leads to a line-of-sight approach (long mean free path)
Sputtering - better step coverage. The pressure is low but high enough that there are many collisions so the mean free path is short
Why is a magnetron used in sputtering?
A magnetron is used to improve sputter yield/deposition rate without increasing the pressure (quantity of gas), which would lower the mean free path too much.
By applying a magnetic field, electrons spiral between the target and the wafer, leading to more ionizations. This means that the plasma density is higher - there are more ions in the same quantity of gas.
The magnetron does not affect the step coverage.
The magnetron increases deposition rate 100x and decreases the required pressure 100x.
What is a collimator, and what problem does it solve?
A collimator is a screen with long, thin openings that can be placed between the target and the wafer in sputtering (about 1 cm away from the target).
The purpose of a collimator is to direct the deposition during sputtering in order to ensure that there are not voids in areas with high aspect ratios.
In order to improve via fill, unidirectional deposition is required. This is accomplished by increasing the mean free path to ~1 cm by decreasing the pressure, and then using a collimator.
