Chapter 8 - Surface Characterization For VLSI

Publisher Summary The revolution in surface analysis techniques was stimulated in part by early demands of the semiconductor industry for improved understanding and control of the surface and interface behavior of solid-state devices. With commercial availability of the scanning electron microscope (SEM), high-resolution images supplied a wealth of information about surface topography, which was used to discover and remedy problems such as pinholes in dielectric layers, undercutting in patterned metallization, and particulate contaminants on photo masks. The concept of a surface became influenced largely by topography visible in the SEM and by the electron microprobe, which made possible elemental analyses of surface material as thin as 1 μ m. The concept of a surface now requires distinction between the outermost monolayer, foreign atoms adsorbed and absorbed onto it, a near-surface transition region, and the bulk substrate region. The surface analyst becomes involved in characterizing the design change and the original failure. It is clear that over the next decade, line widths in very large-scale integration (VLSI) circuits will continue to shrink toward and beyond 0.5 μ m to achieve higher computing power and speed per chip. The lithography required to fabricate these patterns will at first push optical systems to their diffraction limit and then use direct-write electron-beam machines, which are already extensively used for generation of optical VLSI reticles.