You are here

Semiconductor Device Modeling, Simulation, and Failure Prediction for Electrostatic Discharge Conditions

Download pdf | Full Screen View

Date Issued:
2019
Abstract/Description:
Electrostatic Discharge (ESD) caused failures are major reliability issues in IC industry. Device modeling for ESD conditions is necessary to evaluate ESD robustness in simulation. Although SPICE model is accurate and efficient for circuit simulations in most cases, devices under ESD conditions operate in abnormal status. SPICE model cannot cover the device operating region beyond normal operation.Thermal failure is one of the main reasons to cause device failure under ESD conditions. A compact model is developed to predict thermal failure with circuit simulators. Instead of considering the detailed failure mechanisms, a failure temperature is introduced to indicate device failure. The developed model is implemented by a multiple-stage thermal network.P-N junction is the fundamental structure for ESD protection devices. An enhanced diode model is proposed and is used to simulate the device behaviors for ESD events. The model includes all physical effects for ESD conditions, which are voltage overshoot, self-heating effect, velocity saturation and thermal failure. The proposed model not only can fit the I-V and transient characteristics, but also can predict failure for different pulses.Safe Operating Area (SOA) is an important factor to evaluate the LDMOS performance. The transient SOA boundary is considered as power-defined. By placing the failure monitor under certain conditions, the developed modeling methodology can predict the boundary of transient SOA for any short pulse stress conditions. No matter failure happens before or after snapback phenomenon.Weibull distribution is popular to evaluate the dielectric lifetime for CVS. By using the transformative version of power law, the pulsing stresses are converted into CVS, and TDDB under ESD conditions for SiN MIMCAPs is analyzed. The thickness dependency and area independency of capacitor breakdown voltage is observed, which can be explained by the constant ?E model instead of conventional percolation model.
Title: Semiconductor Device Modeling, Simulation, and Failure Prediction for Electrostatic Discharge Conditions.
17 views
5 downloads
Name(s): Li, Hang, Author
Sundaram, Kalpathy, Committee Chair
Batarseh, Issa, Committee Member
Fan, Deliang, Committee Member
Gong, Xun, Committee Member
Salcedo, Javier, Committee Member
University of Central Florida, Degree Grantor
Type of Resource: text
Date Issued: 2019
Publisher: University of Central Florida
Language(s): English
Abstract/Description: Electrostatic Discharge (ESD) caused failures are major reliability issues in IC industry. Device modeling for ESD conditions is necessary to evaluate ESD robustness in simulation. Although SPICE model is accurate and efficient for circuit simulations in most cases, devices under ESD conditions operate in abnormal status. SPICE model cannot cover the device operating region beyond normal operation.Thermal failure is one of the main reasons to cause device failure under ESD conditions. A compact model is developed to predict thermal failure with circuit simulators. Instead of considering the detailed failure mechanisms, a failure temperature is introduced to indicate device failure. The developed model is implemented by a multiple-stage thermal network.P-N junction is the fundamental structure for ESD protection devices. An enhanced diode model is proposed and is used to simulate the device behaviors for ESD events. The model includes all physical effects for ESD conditions, which are voltage overshoot, self-heating effect, velocity saturation and thermal failure. The proposed model not only can fit the I-V and transient characteristics, but also can predict failure for different pulses.Safe Operating Area (SOA) is an important factor to evaluate the LDMOS performance. The transient SOA boundary is considered as power-defined. By placing the failure monitor under certain conditions, the developed modeling methodology can predict the boundary of transient SOA for any short pulse stress conditions. No matter failure happens before or after snapback phenomenon.Weibull distribution is popular to evaluate the dielectric lifetime for CVS. By using the transformative version of power law, the pulsing stresses are converted into CVS, and TDDB under ESD conditions for SiN MIMCAPs is analyzed. The thickness dependency and area independency of capacitor breakdown voltage is observed, which can be explained by the constant ?E model instead of conventional percolation model.
Identifier: CFE0007670 (IID), ucf:52512 (fedora)
Note(s): 2019-08-01
Ph.D.
Engineering and Computer Science, Electrical and Computer Engineering
Doctoral
This record was generated from author submitted information.
Subject(s): ESD -- modeling -- thermal failure -- diode -- safe operating area -- LDMOS -- TDDB
Persistent Link to This Record: http://purl.flvc.org/ucf/fd/CFE0007670
Restrictions on Access: public 2019-08-15
Host Institution: UCF

In Collections