A test circuit based on a ring oscillator array for statistical characterization of Plasma-Induced Damage

Abstract

We propose a test circuit for characterizing PlasmaInduced Damage (PID) based on a ring oscillator array for collecting high-quality BTI statistics. Two types of ring oscillators, PID protected and PID damaged, with built-in antenna structures were designed to separate PID from other effects. A beat frequency (BF) detection scheme was adopted to achieve high frequency measurement precision (>0.01%) in a short measurement time (>1μs) to prevent unwanted BTI recovery. The proposed circuit enables accurate PID-induced BTI lifetime prediction with different Antenna Ratios (ARs) in any type of device with any topology of antenna structure under any fabrication process. Measured frequency statistics from a 65nm test chip shows a 1.15% shift in the average frequency as a result of PID.

Evaluation of plasma-induced damage and bias temperature instability depending on type of antenna layer using current-starved ring oscillators

“…In our previous measurement circuits, antennas are directly connected to wires inside ROs. 1,28) This measurement circuit cannot separate PID effects between PMOS and NMOS. Moreover, it is difficult to convert frequency fluctuation to V th because each MOSFET suffers from degradations independently.…”

“…In previous studies, differences between SiO 2 and HK gate dielectrics 25) in terms of antenna shape (e.g., square and comb) 26) and thickness 27) were evaluated. BTI caused by PID was also evaluated in the study of Choi et al 28) Our research focused on PID depending on the type of antenna layer, particularly the second to fifth metal layers (M2-M5) in SSDM 2017. 29) These metal layers are used for local routing and have the same thickness.…”

Defect generation in electronic devices under plasma exposure: Plasma-induced damage

“…Various approaches have been conducted individually in the field of plasma process developments, circuit designs, and material designs. Recently, some collaborative studies have been carried out 20,21) because the suppression of PID is no longer only an issue in each of the development fields but one of the most critical interdisciplinary concerns in the whole microelectronic industry.…”