I am working on refrigeration system valves. I am currently developing a new valve housing machined out of uncoated 7075-T651 material. This housing is a tube about 2 inches long and 3/8" diameter containing a valve seat with a 50 degree conical shape into which an 1/8" ball is seated. I am finding high leakage on the seat when compared to a similar seat out of anodized 6061. After the seat is honed with lapping compound the leakage is acceptable. Additionally, after sitting on the bench a few days, a thin film of what I assume is aluminum oxide forms. This coating is white and has a fine powdery appearance and is fixed to the aluminum but forms only on the inside of the part. The outside of the part appears shiny and freshly machined.

Questions:
What is the physical mechanism by which aluminum oxide forms?
Is this coating hard enough to prevent a steel ball from forming a seal?
Inexpensive ways to retard the formation of this layer?

I appreciate any information on these topics. Thank you.