Transcript No Slide Title

AoE Project
Nano-Process Modeling:
Lithography modeling and device fabrication
Philip Chan, Mansun Chan
Edmund Lam
Department of ECE, HKUST
Department of EEE, HKU
Processing
The role of Processing Modeling
Lithography
Lateral Dimension (and variation)
Post-Lithography
Processing
Lateral/vertical Dimensions,
dopant/substrate interaction
Device Structure
Structural parameters
Device
Characteristics
Circuit Simulation
Current-voltage, charge, transport,
quantum characteristics
Inter-connected devices characteristics
Lithography
Layout and Lithography
Layout
2D SEM
3D image
Lithography is the driving force for scaling
Lithography also becomes the major source of uncertainties
*Mark E. Mason, DFM EDA Technology: A Lithographic Perspective, 2007 Symposium on VLSI Technology
Lithography
Impact on Device Characteristics
Gate length and parasitics variation
Leakage, noise margin and power consumption variation
currently modeled as many
transistors in parallel
in 22nm technology, need to account for atomic level variations
Lithography models needed for Design for Manufacturing (DFM)
Lithography
Potential Lithography Solutions
*Source:
ITRS Lithography 2008
Lithography
Optical Projection Lithography
*Source: A. Wong, Resolution Enhancement Techniques in Optical Lithography. SPIE Press, 2001. p. 24
Lithography
Mask Design
Forward problem: from mask
design to circuit output
Inverse problem: from circuit
output to mask design
Lithography
Mask Design
Pre-distort the mask!  needs optical system modeling
• Optical proximity correction (OPC)
• Rule-based and model-based
mask
Source adjustment
• Off-axis illumination (OAI)
• Source-mask optimization (SMO)
circuit
Lithography
Design for Manufacturing (I)
Process variations e.g. focus, dose
• Impact circuit behavior
• Require “robust” design
“original”
“robust”
mask
infocus
defocus
Lithography
Design for Manufacturing (II)
Mask manufacturability (cost) need “regularization”
• Few segments
• Rectilinear shapes
• No close shapes (hotspots)
unregularized
regularized
post-Lithography
Beyond Lithography
Controlling vertical dimensions
Further lateral dimension reduction
The most common method to fabricate nanowire MOSFETs:
stress limited oxidation
Y. Tian, 2007 IEDM
CEA/LETI, IEDM 2008
HKUST, EDL May 2009
post-Lithography
Sidewall Etching Profile
♦ Existing Model
♦ Cannot explain the semicircular shape in single sidewall
*
R. Zhou, et. al., “Simulation of the Bosch process with a string-cell hybrid method”, J. of Micromechanics and Microengineering, v. 14 (2004), 851-858
post-Lithography
The Etch-Deposition Model
♦ Etching with deposition
C4F8 – passivation agent
SF6 – etching agent
♦ deposited polymer only removed by the anisotropic
component of the SF6 etchant
♦ final shape depends on the ratio of the etch rate
versus deposition rate
post-Lithography
Oxidation modeling
♦ Can stress really limit the oxidation?
♦ Experimental results show no limit at high temperature
♦ No model can clearly capture these effects
post-Lithography
Existing Model (Deal-Grove)
post-Lithography
Crystal Orientation
♦ Crystal orientation dependent oxidation has not been
accounted for
♦ Oxidation model has to be improved for non-planar
devices due to multiple surfaces used
Devices Task
DFM Objectives
Lithography model
Post-lithography
processing model
Model dependent
intermediate device
parameters
Characteristic
Variations
• Final geometry parameters
• Effects of irregular cross-sectional
geometry on device characteristics
• Non-uniform quantum effects with
non-circular/non-planar devices
Tasks
Lithography Modeling
♦ Develop lithographic model, incorporating particularly
parameters for the light source, and 3D mask effects
♦ Develop mask synthesis algorithms that are
robustness against mask variations
♦ Incorporate regularization methods in solving inverse
problems in imaging to tackle mask complexity
♦ Investigate the resulting parameter variations for input
to the post-lithography processes
Tasks
Etching/Oxidation Model
♦ Develop post-lithography atomic etching model base
on etching with co-deposition
♦ Experimentally verify the effect of stress on nano-scale
oxidation
♦ Develop crystal orientation dependent oxidation model
♦ Incorporate the new models into a process simulator to
predict the final shape of the device
Tasks
Process-Aware Device Modeling
♦ Study the impact of geometrical variation (e.g. nonuniform lateral dimension, non-ideal cross-section,
atomic level dopant fluctuation) on the I-V
characteristics including quantum effects
♦ Develop model based with an orthogonal set of
parameters that correlate the geometry data to electrical
characteristics
♦ Device fabrication and model verification
Tasks
Second Phase of The Project
♦ To Integrate the process models and device models to
take in lithography process parameters and produce the
corresponding device characteristics
♦ From the variations of process to predict the variation of
device characteristics
♦ Develop inverse process algorithm to produce
lithography patterns and process parameters to
minimize the variation of device characteristics