sputter yield

Sputter yield is defined as the average number of target atoms ejected per incident ion during the sputtering process, serving as the fundamental efficiency metric for PVD deposition. It is a dimensionless quantity typically ranging from 0.5 to 3.0 for most semiconductor-relevant materials sputtered with argon ions at standard operating energies (300-600 eV). Sputter yield depends on several interrelated factors: incident ion energy, ion mass, target material properties (atomic mass, surface binding energy, crystal structure), and the angle of ion incidence. The yield increases with ion energy above a threshold energy (typically 20-50 eV) following a roughly linear relationship at low energies before rolling over at very high energies (>1 keV) where ion implantation begins to dominate. Heavier incident ions (krypton, xenon) generally produce higher yields than lighter ions (neon, argon) due to more efficient momentum transfer, though argon remains the universal choice as a balance of cost, availability, yield, and compatibility. The angular dependence of sputter yield shows a maximum at approximately 50-70° from surface normal, which is relevant to profile control in etch processes and explains phenomena like sidewall bowing and faceting. Representative sputter yields for 500 eV Ar⁺ bombardment include: aluminum ~1.0, copper ~2.3, silver ~3.1, titanium ~0.5, tantalum ~0.6, tungsten ~0.6, silicon ~0.5, and SiO2 ~0.13. The lower yields of refractory metals (Ta, W, Ti) and compounds compared to softer metals reflect their higher surface binding energies and atomic masses. Sputter yield data is essential for calculating deposition rates (deposition rate is proportional to yield × ion current / target-to-substrate geometry), predicting target lifetime, and designing etch processes where physical sputtering is a component. Yield measurements are performed through weight loss techniques, Rutherford backscattering spectrometry (RBS), or quartz crystal microbalance methods. Computer simulation codes like SRIM/TRIM calculate theoretical yields based on binary collision approximation models of ion-solid interactions.

Go deeper with CFSGPT

Get AI-powered deep-dives, save terms, and run advanced simulations — free account.

Create Free Account