The influence of plastic properties of the probe tip/Si contact on spreading resistance analyses

Abstract

Analyses of spreading resistance measurements in Si samples with dopant profiles of the dopant sequence lowly doped on highly doped yield unreliable results if the dopant profiles are extremely steep or ultra-shallow. Reasons for this behavior are seen in effects of penetration of the probe tips into the Si sample and in the presence of a zone of a metallically conducting Si phase beneath the probe tip. Atomic force microscopy has been used to investigate the geometry of probe tip indents and scanning electron microscopy to investigate the geometry of the probe tip itself. Results of loading-unloading experiments by means of nanoindentation with a conventional spreading resistance probe tip as indenter enable us to estimate the depth of the metallic Si phase zone. It ranges from about the fourfold up to the sevenfold of the residual depth after unloading the probe tip.

Properties of probe tip/Si contacts and their connection to spreading resistance analyses

“…18 By comparison of the plastic penetration depth under load d p and the maximum residual penetration depth after unloading d r , we can estimate the depth of the metallic Si͑II͒ phase provided the contact area is known. 17 The AFM investigations exhibit maximum residual penetration depths d r of probe tips in the order of 10 nm using conventional, wellconditioned OsW probe tips and a load of 0.05 N. By nanoindentation we find that d p is about twice the maximum residual depth at that load. Depending on assumptions about the mass density of Si͑IЈ͒, 19 we estimate the bottom of the metallic phase located at a depth of about twice d p .…”

“…In the case of neglecting probe tip penetration, the sr measurement should reveal the thickness of the chosen buried SiO 2 layer by measuring the total resistances at the upper limit of the sr system. With increasing probe tip penetration, the interval in which we measure re- 17 Integrating the z fluctuations, separated for valleys and hills, we get the valley volume v val and the hill volume v hi . Furthermore, the averaged residual valley depth d av that can be obtained is d av ϭ4.7 nm in the case of the contact configuration of Fig.…”

“…During unloading, a reverse phase transition from the metallic Si͑II͒ to an amorphous/polycrystalline phase Si͑IЈ͒ takes place for both indenters, which can be concluded from TEM and AFM investigations. 17 The Si͑IЈ͒ phase has a reduced density causing a characteristic inflection in the unloading branch of the load-displacement curve. 18 By comparison of the plastic penetration depth under load d p and the maximum residual penetration depth after unloading d r , we can estimate the depth of the metallic Si͑II͒ phase provided the contact area is known.…”

“…20, the thickness of the metallic zone varies with the square root of the load. 17 Although there is still some uncertainty with respect to the spatial extent of the metallic phase, the existence of a metallic phase beneath the contact should not be neglected any longer analyzing ld on hd or ultrashallow dopant profiles. Up until now, the influence of this phase on results of sr analyses in case of (z) constant is not well known.…”