Breakdown and reliability of nanoscale dielectric films
Loading...
Authors
Ogden, Sean P.
Issue Date
2017-12
Type
Electronic thesis
Thesis
Thesis
Language
ENG
Keywords
Chemical engineering
Alternative Title
Abstract
The model equations are refined to expand its failure prediction capabilities. The electron temperature is re-defined solely as a function of the electric field, and the defect generation rate is more closely aligned with the thermo-chemical E model, including a thickness dependence. The equations are also solved for in dimensionless form, which shows convection-driven electron transport, slow defect reaction kinetics, and non-uniform local field trends that depend on the thickness and density of mobile electrons in the dielectric. The refined charge transport model is able to retain all of the initial model’s offerings, while also expanding its capabilities to predict dielectric failure as a function of temperature and thickness. The model replicates voltage-dependent activation energy for dielectric failure data, tying the results to the energy barrier for electron injection. Failure distributions are also replicated, based solely on dielectric spacing calculations from voltage ramp tests. The model also provides several predictions for future technology nodes as the dielectric thickness decreases. Based on the model theory, it is expected that thinner dielectrics will yield a higher dielectric strength (for planar dielectrics only) and suffer from degraded variability.
Dielectric breakdown is the formation of a conductive path in an insulating material, and it is observed in nature (lightning strikes) and man-made equipment (spark plugs, microelectronics). In the microelectronics industry, dielectric breakdown is becoming an increasing concern for reliability engineers as device dimensions continue to shrink in order to increase computing power and speed. However, the ability to understand and predict the breakdown behavior of dielectrics in these integrated circuits is becoming more complex as new materials are introduced and the manufacturing process required to fabricate these insulators and their surrounding metal layers is constantly changing.
Due to time constraints, engineers employ a strategy to test a sample population at several high voltages in order to wear out the dielectric over time and cause failure (also called time-dependent-dielectric-breakdown). In order to extrapolate their results to low voltages and longer lifetimes, empirical models are used to fit the experimental data and then extrapolate to operating conditions. There is dispute over which empirical model provides the most accurate prediction for device lifetime, and it is most likely that a single model cannot be used to predict failure for every type of dielectric and process involved in an integrated circuit. As a result, a charge transport model was developed to provide a more comprehensive understanding of dielectric failure, in order to tie together the physical mechanism of failure to the dielectric material and the processing conditions.
The charge transport model is 1-D and incorporates current-driven and field-driven failure mechanisms. Conduction into the dielectric occurs when electrons overcome an energy barrier, and their transport throughout the dielectric depends on the electric field and the donor-type defects located in the matrix. The energetic electrons collide with the matrix, breaking bonds to increase the defect density. Breakdown occurs when a critical defect concentration accumulates, resulting in electron tunneling and the emptying of positively charged traps. The enhanced local electric field lowers the barrier for electron injection into the dielectric, causing a positive feedforward failure. The model uses a minimal number of adjustable parameters, none of which are directly used to fit the slope of the time-to-failure versus applied electric field curve. In addition, all the parameters have some theoretical basis or have been measured experimentally.
The charge transport model is able to replicate leakage current and failure behavior for several types of dielectric materials, including low-κ SiCOH and high-κ SiN dielectrics, which are both commonly found in integrated circuits. The model can reproduce I-V and I-t curves, capturing the current decay at early stress times and the rapid current increase observed at failure for constant voltage testing. Most importantly, the model is able to predict the time-to-failure as a function of the applied electric field, without using any parameters to fit this slope. The model is able to directly compare SiCOH and SiN failure trends, and finds silicon nitride films have superior reliability due to their higher dielectric constant and deeper defect energy levels.
Dielectric breakdown is the formation of a conductive path in an insulating material, and it is observed in nature (lightning strikes) and man-made equipment (spark plugs, microelectronics). In the microelectronics industry, dielectric breakdown is becoming an increasing concern for reliability engineers as device dimensions continue to shrink in order to increase computing power and speed. However, the ability to understand and predict the breakdown behavior of dielectrics in these integrated circuits is becoming more complex as new materials are introduced and the manufacturing process required to fabricate these insulators and their surrounding metal layers is constantly changing.
Due to time constraints, engineers employ a strategy to test a sample population at several high voltages in order to wear out the dielectric over time and cause failure (also called time-dependent-dielectric-breakdown). In order to extrapolate their results to low voltages and longer lifetimes, empirical models are used to fit the experimental data and then extrapolate to operating conditions. There is dispute over which empirical model provides the most accurate prediction for device lifetime, and it is most likely that a single model cannot be used to predict failure for every type of dielectric and process involved in an integrated circuit. As a result, a charge transport model was developed to provide a more comprehensive understanding of dielectric failure, in order to tie together the physical mechanism of failure to the dielectric material and the processing conditions.
The charge transport model is 1-D and incorporates current-driven and field-driven failure mechanisms. Conduction into the dielectric occurs when electrons overcome an energy barrier, and their transport throughout the dielectric depends on the electric field and the donor-type defects located in the matrix. The energetic electrons collide with the matrix, breaking bonds to increase the defect density. Breakdown occurs when a critical defect concentration accumulates, resulting in electron tunneling and the emptying of positively charged traps. The enhanced local electric field lowers the barrier for electron injection into the dielectric, causing a positive feedforward failure. The model uses a minimal number of adjustable parameters, none of which are directly used to fit the slope of the time-to-failure versus applied electric field curve. In addition, all the parameters have some theoretical basis or have been measured experimentally.
The charge transport model is able to replicate leakage current and failure behavior for several types of dielectric materials, including low-κ SiCOH and high-κ SiN dielectrics, which are both commonly found in integrated circuits. The model can reproduce I-V and I-t curves, capturing the current decay at early stress times and the rapid current increase observed at failure for constant voltage testing. Most importantly, the model is able to predict the time-to-failure as a function of the applied electric field, without using any parameters to fit this slope. The model is able to directly compare SiCOH and SiN failure trends, and finds silicon nitride films have superior reliability due to their higher dielectric constant and deeper defect energy levels.
Description
December 2017
School of Engineering
School of Engineering
Full Citation
Publisher
Rensselaer Polytechnic Institute, Troy, NY