The issue of how to condition or “Break In” a new sputtering target to prepare it for continuous use during plasma deposition has come up numerous times over the years, so I thought I should probably address it here once and for all.
In general, a new sputtering target should be preconditioned or “Broken In” prior to the actual deposition onto a given substrate. This is due to the method(s) of fabrication in the normal course of the target fabrication. This “Break In” procedure reduces any internal stresses that may have been introduced within the target, due to mechanical working in the fabrication process, as well as removing any surface contamination that may have been initiated during processing. Such surface contamination may exist in the form of a thin layer of oxide or nitride formation (as a result of being in contact with air), a hydride formation (from contact with the moisture in the air) or from cleaning solvents (used in the target preparation and cleansing procedure).
Usually the target conditioning consists of igniting the initial plasma at the lowest possible power density, holding or “soaking” the plasma at this level for a given length of time (typically several minutes per increment) then slowly raising the power density by some given increment (usually 25-50 watts or so) and then soaking it again at the increased power level. These are referred to “rise” and “soak” steps respectively. This procedure should be repeated as many times as necessary until the anticipated or specified power density for operation/production is reached and the plasma is stable. After soaking the target at this anticipated operating level for several minutes, one more increment should be employed, so as to guarantee that the target will be operating in an absolute equilibrium during normal process operation.
For subsequent runs (after the initial conditioning has been completed) it will not be necessary to repeat this procedure for each start up, although it is never advisable to ramp up the power level on any given target too rapidly for fear of over stressing the target as a result of thermal variances caused within the target from heat being generated at the target surface (top) exothermically from the atomic bond breaking due to ion bombardment in the plasma and then being dissipated through the thickness of the target to the backside (bottom) where the water cooling takes place in the magnetically enhanced cathode assembly.
