We have a RIE system that use 90%CF4 mixed with 10%O2 for SiO2 etching (gas is premixed). My question is how to setup a recipe that will give a straight(anisotropic) etching profile. I have four possible recipes here, the power is 100W for all of them
Pressure, gas flow: 50 mtorr, 50 sccm
Pressure, gas flow:50 mtorr, 100 sccm
Pressure, gas flow:100 mtorr, 50 sccm
Pressure, gas flow:100 mtorr, 100 sccm
Which of this will be a best to give a anisotropic etching on SiO2, and why?
Thanks in advance!
Normally to get a straight profile (anisotropic etch), one uses low pressure. Low pressure means a high mean free path, which means that the ion will not change direction due to collision with other ions, radicals, or neutrals. Normally shorter mean free path leads to isotropic etch which happens at high pressure. In nutshell, you should be using low pressure for the straight profile. I think the flow rate does not affect the recipe too much, maybe it helps in achieving the chamber pressure quickly.
I would choose 50mTorr (I would even try 10 mTorr) and 50 SCCM.
Shelender Kumar, I second your answer. I have tried recipes with 10-50 mTorr for anisotropic profiles and obtained a better structure with the lowest of pressures. Can you comment on the effect of the variation in the gas flow rate at differently set pressures?
Thanks to Jerry Liu for starting the thread.
Abishek Kumar , My understanding is that flow rate is related to the amount of gas in the chamber but ultimately you are limited by the pressure in the chamber. This only means that the valve in front of the turbo will adjust to maintain the required pressure. I can give a scenario when this parameter will be useful.
- Let us say you want 100mT of pressure in the chamber but then the flow rate is 5 sccm, this will take a lot of time to reach the 100mT. In this case, you may want to increase the flow. This is the scenario, you may not want it, so flow is adjusted so that the turbo can maintain the pressure without struggling.
I hope it helps
It really helped in understanding the operation. Thanks and good luck to you! Shelender Kumar
To get a more anisotropic profile you need to ensure that the ions are reaching the substrate with a narrow angular distribution (with angles close to 90 deg), and also that they have high kinetic energy. To ensure the first point, you need to decrease chamber pressure and keep gas flow rate at values where there are sufficient number of ions still available for etching to occur. This basically means the mean free path of the ions will be high and they do not undergo many collision events which would deviate them from their straight path. As you increase the flow rate at same pressure, the residence time of the ions inside the chamber will increase thereby increasing the probability to suffer collision events. To optimize this you can try out 3-4 different values of flow rate while keeping the chamber pressure at the lowest value permitted by your reactor's vacuum circuit, and then in SEM for permissible levels of undercutting, etch rate etc. For the second point, you need to increase the bias voltage, which can be done by increasing the power supplied to the lower electrode.
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts