The steamer also eliminates the need for a carrier
gas by delivering ultra-pure steam at a constant positive pressure. This enables delivery of 100% pure water
vapor, ensuring that maximum theoretical oxide growth
rate is achieved.
Data from foundries growing thick oxides indicated
that these steamers could increase oxide growth rate,
improve wafer-to-wafer and within-wafer uniformity,
improve film quality, and reduce operating cost when
compared against other steam technologies1.
The MEMS manufacturer evaluated our steamer in a
new project for one of its primary MEMS customers. The
resulting oxide films were judged by the manufacturer
and its MEMS customer to be equal to oxide films grown
with pyrolytic steam. This gave the manufacturer the
justification to put the steamers into production and
sideline their torches.
Figure 1: Growth rate comparison. Faster oxide growth with the steamer
due to higher partial pressure of steam species (H2O –Steamer vs. H2O +
O2-Torch).
Challenges of Pyrolytic Steam
Wet thermal oxidation is typically used when oxide
thicknesses of greater than 1000 Angstrom (Å) are
needed. Historically, pyrolytic steam (generated from the combustion of hydrogen and oxygen) has been used in this process.
This choice was based on requirements for steam purity and film
repeatability, which are not achievable with water-based steam
delivery systems such as direct liquid injection or water bubblers.
However, there are some serious drawbacks to pyrolytic steam:
Material cost of the hydrogen and oxygen ■
Results and Discussion
The following results were collected by the MEMS
manufacturer over a 24 month period to support the
