Suppression of Statistical Variability in Junctionless FinFET Using Accumulation-Mode and Charge Plasma Structure

Abstract

This paper presents a comprehensive study on the statistical variability in charge plasma junctionless FinFET (CP-JLF) and junctionless accumulation-mode FinFET (JAM) compared with conventional junctionless FinFET (JLF). The variability in key figures of merits induced by line edge roughness (LER), random dopant fluctuation (RDF), and work function variation is analyzed. Results indicate that JAM and CP-JLF are more immune to LER and RDF than conventional JLF. Furthermore, elevating doping concentration in source while remaining relatively lower in drain is beneficial in reducing RDF-induced current variability in JAM. Besides, the impact of gate length on current variability becomes less significant or even negligible as source doping concentration increases, which helps alleviate more serious RDF upon scaling. Moreover, it is essential for JAM and CP-JLF to control geometry and doping variation in the source side that have major influence on LER and RDF. Overall, JAM and CP-JLF with improved current driving and reduced statistical variability mitigate two main obstacles that hinder conventional JLF.