Scanning spreading resistance microscopy (SSRM) is a powerful technique for quantitative two-and three-dimensional carrier profiling of semiconductor devices with sub-nm spatial resolution. However, considering the sub-10nm dimensions of advanced devices and the introduction of three-dimensional architectures like FinFETs and nanowires, the measured spreading resistance is easily impacted by parasitic series resistances present in the system. The limited amount of material, presence of multiple interfaces and confined current paths may increase the total resistance measured by SSRM beyond the expected spreading resistance, which can ultimately lead to an inaccurate carrier quantification. Here, we report a TCAD assisted experimental study to identify the different parameters affecting the SSRM measurements in confined volumes. Experimentally, the two-dimensional current confinement is obtained by progressively thinning down uniformly doped blanket SOI wafers using scalpel SSRM. The concomitant SSRM provides detailed electrical information as a function of depth up to oxide interface. We show that the resistance is most affected by the interface traps in case of a heterogeneous sample, followed by the intrinsic resistance of the current carrying paths. Further, we show that accurate carrier quantification is ensured for typical back-contact distances of 1 µm if the region of interest is at least 9 times larger than the probe radius.

Scanning Spreading Resistance Microscopy is a well-established technique for obtaining quantitative two- and three-dimensional carrier profiles in semiconductor devices with sub-nm spatial resolution. However, for sub-100 nm devices, the use of focused ion beam becomes inevitable for exposing the region of interest on a sample cross section. In this work, we investigate the impact of the focused ion beam milling on spreading resistance analysis and we show that the electrical effect of the focused ion beam extends far beyond the amorphous region and depends on the dopant concentration, ion beam energy, impact angle, and current density. For example, for dopant concentrations between 1.0 × 1020 and 1.5 × 1016 cm−3 we observe dopant deactivation at least between 23 and 175 nm for a glancing 30 keV ion beam. Further, we show that dopant deactivation is caused by defect diffusion during milling and is not directly impacted by the presence of Gallium in the sample. Later, we also discuss potential ways to mitigate these effects.