An important quote from Eliminating Bond Stresses of Sputtering Targets at Operating Temperatures is: “This suggests that a target bonded at one-half of the temperature difference between room temperature and operational temperature would have the most overall desirable characteristics, since it is impossible to rule out thermal expansion mismatches in most materials.”
Yes, we have to accept the fact that heating assemblies (like bonded sputtering targets) forces the materials to expand. Let’s take a look at how we can minimize the stresses in the bond, which correlate to the differential strain in the materials.
ε is strain
α is the coefficient of expansion
ΔT is the change in temperature
When ε1 does not equal ε2 and the materials are bound together, shear forces will create stresses in the bond area.
As you can clearly see, there are not too many variables for us to manipulate. Keeping strain ε1 and ε2 as similar as possible is key. We can either match the coefficient as closely as possible or we can make sure the ΔT stays low. The first step you should take is to try to match the CTE (α) of your backing plate and target material. When this isn’t possible, it’s time to think of ways to keep ΔT as low as possible.