Qualification of spreading resistance probe operations. I

Abstract
In this article we discuss the final results obtained from an extensive spreading resistance probe (SRP) experimental intercomparison conducted among 21 international laboratories with different levels of expertise both in the U.S. and in eight European countries. Twenty different structures representative of current SRP needs (complementary metal–oxide–semiconductor, bipolar, epi) were used. Specialized structures such as buried oxides with different thicknesses (10–40 nm) and identical implants junction isolated and nonisolated (carrier diffusion) were included. The SRP accuracy was determined according to four point probe sheet resistance mappings and secondary ion mass spectrometry depth profiles. In order for SRP to gain wider credibility an error of less than 10% (std dev) should be routinely achievable on the typical SRP results (e.g., dose, sheet, shape carrier profile) irrespective of where, when, or by whom the measurement is performed. The round robin that was conducted showed that present American Society for Testing and Measurement standards are not detailed enough to obtain this target accuracy. In order to achieve this accuracy, precise quantitative specifications on as many as 14 primary qualification parameters must be satisfied simultaneously. Of these, some of the more important ones are low surface roughness, sharp bevel edge, low probe penetration, stable probe contact, small contact size, accurate bevel angle, accurate starting point, sufficient data point density, low probe separation, good probe alignment, and a standardized contact model. The required specifications will be illustrated with results from the round robin. Also, other topics investigated in the round robin will be discussed briefly such as the on-bevel stepsize reproducibility, the microcontact distribution, the different contact radii involved, the amount of barrier resistance, rectification, and noise.

This publication has 10 references indexed in Scilit: