Try what? Add some pointless solder to a disk and then what? Wait years to see how long it takes to die? How many do I need for a sample size? Do I need to test the same model? What about workloads the drives should be under?
This is pure untestable unverifiable snake oil.
Gold does not oxidize in atmosphere. You are spouting nonsense you do not understand.