We h a ve applied a simultaneous combination of scanning Kelvin probe microscopy and scanning atomic force microscopy to the problem of pro ling dopant concentrations in two dimensions in silicon microstructures. By measuring the electrochemical potential di erence which minimizes the electrostatic force between probe tip and sample surface, we estimate the work function di erence between the tip and surface. To the extent that this work function di erence is a consequence of the dopant concentration at, or near, the sample surface, we infer doping pro les from our measurement. Structures examined and presented here include contact holes, and the technologically signi cant lightly-doped drain of a metal-oxide-silicon eld-e ect transistor. Using this methodology, w e are able to distinguish relative c hanges in dopant concentration with lateral resolution less than 100 nm. Sample preparation is minimal, and measurement time is fast compared to other techniques. Our measurements have been compared to predictions based on two-and three-dimensional process and device simulation tools. The comparisons show our technique is sensitive t o c hanges in dopant concentration, from 10 15 cm ,3 to 10 20 cm ,3 , of less than ten percent at these size scales. Suggestions to resolve absolute dopant concentration are made.
Measurement of dopant density in silicon with lateral resolution on the 200 nm scale has been demonstrated with a near-field capacitance technique. The technique is based upon the measurement of local capacitance between a 100 nm tip and a semiconducting surface. Lateral dopant imaging is achieved by the measurement of the voltage-dependent capacitance between tip and sample due to the depletion of carriers in the semiconductor, as the tip is scanned laterally over the surface. Measurements of dopant density have been demonstrated over a dopant range of 1015–1020 cm−3. Capacitance-voltage measurements have been made on a submicrometer scale.
A force-based scanning Kelvin probe microscope has been applied to the problem of dopant profiling in silicon. Initial data analysis assumed the detected electrostatic force couples the sample and only the tip at the end of a force sensing cantilever. Attempts to compare measurements quantitatively against device structures with this simple model failed. A significant contribution arises from the electrostatic force between the sample and the entire cantilever, which depends strongly upon the relative size of the tip, cantilever, and lateral inhomogeneities in the surface topography and material composition of the sample. Actual and simulated measurements which demonstrate the characteristic signature of this effect are presented.
Quantitative dopant profile measurements are performed on a nanometer scale by scanning capacitance microscopy (SCM). An atomic force microscope is used to position a nanometer scale tip at a semiconductor surface, and local capacitance change is measured as a function of sample bias. A new feedback method has been demonstrated in which the magnitude of the ac bias voltage applied to the sample is adjusted to maintain a constant capacitance change as the tip is scanned across the sample surface. A quasi-1D model is used to extract dopant density profiles from the SCM measurements. The inverted SCM dopant profiles are compared with profiles obtained by process simulation and secondary ion mass spectroscopy measurement. Good agreement was found between the SCM measured profile and the lateral profile predicted by SUPREM 4 over the concentration range from 1017 to 1020 cm−3.
A point-defect-based model for the stress effects on dopant diffusion in silicon is presented. Variations in binding energies and diffusivities of dopant-defect pairs under hydrostatic pressure are modeled, and a pressure-dependent dopant diffusion equation is derived. New experimental work was performed on boron pileup near dislocation loops, and compared to the model. Qualitative agreement is possible, which suggests that stress might be a significant effect in scaled modern device structures.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.