Atomic Force Microscope

AFM, STM, and SAM all probe a sample without using light or a transmitted electron beam at all — AFM feels the surface mechanically with a physical tip, STM reads a quantum tunneling current, and SAM bounces high-frequency sound waves through the material — each reaching information the optical and electron-beam techniques elsewhere in the fab's analytical toolkit simply don't provide.

SAM is the only one of the three that can see beneath the surface at all, which is exactly why it's the standard tool for package-level defect detection. AFM and STM are both fundamentally surface techniques — they characterize whatever's exposed at the top, nothing beneath it; SAM instead sends high-frequency ultrasonic pulses into the material and listens for reflections from internal interfaces, catching subsurface voids, delamination between package layers, and internal cracks that no surface-probing or surface-imaging technique could ever detect, since those defects are, by definition, hidden below whatever surface the other methods can reach.

STM's requirement for electrical conductivity is a hard limit that AFM simply doesn't share, which is why AFM has become the far more generally applicable tool despite STM reaching comparable or finer resolution on the surfaces it can image. Scanning tunneling microscopy depends entirely on a tunneling current flowing between tip and sample, which means it only works on conductive or semiconductive surfaces — it's blind to insulating materials entirely. AFM senses physical force between tip and surface instead, which works on virtually any material regardless of conductivity, making it the default atomic-force technique across far more of the materials a fab actually needs to characterize.

AFM, STM, and SAM each access a dimension of information the fab's optical and electron-beam tools can't reach on their own — true surface force topography, atomic-scale conductive structure, and hidden subsurface defects respectively — and a materials or failure-analysis lab typically keeps all three available because no single one of them substitutes for what the other two see.