Certified anisotropic meshing of thin, high-aspect-ratio regions in device modeling

Abstract Thin layers embedded in much larger bodies are routine in semiconductor devices and a chronic difficulty for device simulation. A mesher that keeps elements near-isotropic must resolve the layer thickness in the plane as well, so the number of unknowns grows as the square of the layer aspect ratio. We resolve the layer instead with anisotropic structured cells, whose count does not depend on the aspect ratio, and we certify the result. For steady conduction, a conforming P1 potential and an equilibrated Raviart--Thomas flux, computed on the same mesh, enclose the exact resistance between guaranteed bounds without any reference solution. The enclosure width depends on how planar the field is, not on element shape. For layer-parallel flow it is at round-off level, below $10^{-11}$, for cell aspect ratios from 10:1 to 1000:1. For a 10~mm wide, 30~$\upmu$m thick film, 567 nodes then give the exact sheet conductance, where an isotropic mesh would need about $1.6\times10^{7}$ nodes. For a point source on a 30~nm film over a substrate, the anisotropic solution matches the closed-form image series to within 0.14\% once the grounded outer boundary is accounted for. For a contact on a film--substrate structure, the certified enclosures are nested and shrink from 23\% to 6.9\% under refinement. Applied to a three-layer stack, the method shows that a 30~nm sheet, 0.5\% of the total thickness, carries 0.13\% or 13.4\% of the potential drop depending on its doping, and the certified enclosures of the access resistance for the two dopings are disjoint.

Authors

Institutions

Publication Details

Journal
Research Square
Published
2026-11-06
DOI
https://doi.org/10.21203/rs.3.rs-11367763/v1
Primary Topic
Advanced Numerical Methods in Computational Mathematics
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Certified anisotropic meshing of thin, high-aspect-ratio regions in device modeling

Mohamed Ben Chouikha, Hakeim Talleb, Aboubaker Elomari
Research Square
Advanced Numerical Methods in Computational Mathematics
preprint

Certified anisotropic meshing of thin, high-aspect-ratio regions in device modeling

Mohamed Ben Chouikha, Hakeim Talleb, Aboubaker Elomari
preprint en

Abstract

Abstract Thin layers embedded in much larger bodies are routine in semiconductor devices and a chronic difficulty for device simulation. A mesher that keeps elements near-isotropic must resolve the layer thickness in the plane as well, so the number of unknowns grows as the square of the layer aspect ratio. We resolve the layer instead with anisotropic structured cells, whose count does not depend on the aspect ratio, and we certify the result. For steady conduction, a conforming P1 potential and an equilibrated Raviart--Thomas flux, computed on the same mesh, enclose the exact resistance between guaranteed bounds without any reference solution. The enclosure width depends on how planar the field is, not on element shape. For layer-parallel flow it is at round-off level, below $10^{-11}$, for cell aspect ratios from 10:1 to 1000:1. For a 10~mm wide, 30~$\upmu$m thick film, 567 nodes then give the exact sheet conductance, where an isotropic mesh would need about $1.6\times10^{7}$ nodes. For a point source on a 30~nm film over a substrate, the anisotropic solution matches the closed-form image series to within 0.14\% once the grounded outer boundary is accounted for. For a contact on a film--substrate structure, the certified enclosures are nested and shrink from 23\% to 6.9\% under refinement. Applied to a three-layer stack, the method shows that a 30~nm sheet, 0.5\% of the total thickness, carries 0.13\% or 13.4\% of the potential drop depending on its doping, and the certified enclosures of the access resistance for the two dopings are disjoint.

Research Square
Sorbonne Université (FR)
Advanced Numerical Methods in Computational Mathematics
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon