Electrical Homo-Junction Delineation Techniques: A Comparative Study

Abstract

In active semiconductor devices, the junction characteristics are critical for the electrical performance. As an alternative of the atomic force microscopy (AFM)-based electrical techniques which provide unique junction characterization, other methods are dedicated for the delineation of the electrical junction such as the wet chemical etching, the electrochemical plating method, the Seebeck effect imaging (SEI) method, the electron-beam induced current (EBIC) technique and the secondary electron potential contrast (SEPC) method. The aim of this paper is in the one hand to compare these five techniques in term of sample preparation, spatial application range, spatial resolution, simplicity and information displayed. In the other hand, this review aims to provide some guidelines for the appropriate delineation method(s) selection. It was confirmed that chemical based techniques are the simplest junction delineation methods but exhibit some drawbacks in term of spatial resolution and reproducibility. Despite of a limited spatial resolution, it was evidenced that EBIC can provide accurate electrical characterization of the junction. Finally, it was demonstrated that SEPC is the most promising technique providing the higher spatial resolution. The effect of the sample preparation method has been described. Even if the comparison was mainly based on homo-micro-Silicon junctions (n-p and n-p-n-p), the results were also discussed for short SiC junctions. The importance of the analysis context was considered in this paper and analysis flow was suggested for specific analysis cases.

Triboelectric junction: a model for dynamic metal–semiconductor contacts

“…For example, in the semiconductor industry, the electrochemical displacement plating of metals like copper (Cu), silver (Ag), platinum (Pt), and gold (Au) on Si is commonly employed for junction depth examination, also known as junction delineation. In the junction delineation process, Si donates electrons, while metal ions in solution accept electrons, 31,32 which is consistent with our experiment results that Si donates electrons and metals (Cu, Ag, Pt, and Au) accept electrons during dynamic contact. The possible reason is attributed to the chemical reactions that occur during tribology, known as mechanochemistry.…”

Selective Staining on Non-Volatile Memory Cells for Data Retrieval

“…The second step of copper deposition highlights the morphology difference between stained and non-stained cells for easy image processing and data extraction. In addition, the two-step staining process does not depend on doping concentration, which has been reported to limit the resolution of one-step staining during silicon homojunction delineation [21]. Instead, the electric field induced selectivity is dependent on charge carrier density in the floating gates, which is decided by the threshold voltage and cell dimensions.…”

“…The current technologies with node size <130 nm pose a more demanding challenge on available data W T-IFS-13513-2021.R1 retrieval techniques and call for new approaches not only for Megabyte-size data extraction but with higher signal resolution. Selective staining has been used to delineate substrate silicon and is considered a well-established technique for effective silicon dopant profiling [20], [21]. Copper staining relies on galvanic displacement reaction between the metal-ion solution and the n-type silicon substrate, with the staining rate controlled by the availability of electrons in silicon [22].…”