Scanning Spreading Resistance Microscopy

Scanning spreading resistance microscopy (SSRM) is a contact-mode atomic force microscopy technique that maps dopant concentration across a two-dimensional cross-section of a transistor by physically pressing a hard, conductive diamond tip into the sample and measuring the local electrical resistance beneath it, point by point. It delivers sub-nanometer spatial resolution combined with a dynamic range spanning six or more orders of magnitude in carrier concentration — a combination no other dopant-profiling technique matches, which is exactly why it became essential once transistors moved into fully three-dimensional structures.

SSRM measures spreading resistance, which is the electrical resistance current encounters as it spreads outward from a tiny contact point into the surrounding, much larger volume of semiconductor material. A hard, electrically conductive diamond-coated tip is pressed directly into a cleaved or polished cross-section of the device being studied, applying a small DC voltage and measuring the current that results at that single contact point. Because this spreading resistance depends directly on the local carrier concentration in the semiconductor right beneath the tip, measuring it point by point across a scanned area builds up a genuine map of dopant concentration across that cross-section — not an average or an inference, but a direct electrical measurement at each individual scan location.

The physical contact between a hard diamond tip and the sample is both SSRM's defining feature and the source of its unmatched spatial resolution. Unlike optical or electron-beam techniques, whose resolution is ultimately limited by the wavelength of light or electron beam being used, SSRM's resolution is set by the tip's own physical contact area — a sharp enough tip, pressed with enough force to maintain good electrical contact without damaging the sample, can resolve dopant variations at genuinely sub-nanometer scale. This direct, physical nature is exactly why SSRM remains the gold-standard technique for the finest 2D dopant-profiling resolution available, even as faster, non-contact techniques have become standard for less demanding measurements.

Converting raw measured resistance into an actual carrier concentration number requires calibration against known reference samples, because spreading resistance alone doesn't directly give a concentration value in absolute terms. Staircase calibration samples — specially fabricated structures with precisely known, stepped dopant concentrations at each depth — are measured with the same SSRM setup, establishing the relationship between measured resistance and actual carrier concentration for that specific tip and measurement condition. Every subsequent SSRM measurement on an unknown sample then gets converted through this established calibration curve, turning a raw electrical resistance map into an actual quantitative dopant-concentration map.

SSRM's dynamic range — spanning roughly six decades of carrier concentration in a single measurement — is unusually wide compared to most other characterization techniques, and that range matters because real devices contain exactly this much variation. A single transistor cross-section can contain lightly-doped channel regions, moderately-doped well regions, and extremely heavily-doped source and drain contacts, representing carrier concentrations that differ by many orders of magnitude within a span of just a few hundred nanometers. A technique that could only resolve a narrow slice of that range would miss either the lightly-doped regions or saturate on the heavily-doped ones; SSRM's wide dynamic range lets a single scan capture the full doping landscape of a real, complete device structure in one pass.

SSRM became essential specifically because transistors stopped being flat, which broke simpler one-dimensional dopant-profiling techniques that worked fine for planar devices. Techniques designed around measuring doping as a function of depth alone work well for a planar transistor, where the relevant doping structure really is mostly one-dimensional. FinFETs and gate-all-around transistors have genuinely three-dimensional doping profiles — dopant concentration varies not just with depth but across the fin or nanosheet's width and around its wrapped gate structure — and SSRM's true two-dimensional cross-sectional mapping is one of the few techniques that can actually characterize that full doping landscape rather than collapsing it into a misleadingly simplified depth profile.