Skip to content

Orientation-dependent etching of silicon by fluorine molecules: a quantum chemistry computational study

Summary

This corpus entry registers an archive-style PDF (papers/Others/Si_F2_etching_archive_Kaganovic.pdf) summarizing quantum chemistry modeling of F\(_2\) reactions on fluorine-terminated silicon surfaces, motivated by anisotropic etch rates in plasma-less or otherwise facet-selective Si processing. The abstract-level description states that dissociative chemisorption barriers for F\(_2\) differ strongly by facet, with Si(111) exhibiting comparatively slow Si–Si bond breaking relative to (100) and (110) surfaces, while (100) and (110) accommodate more fluorine, polarize Si–Si bonds, and lower chemisorption barriers in the QM modeling. The authors also report classical MD with a REBO-class potential that underestimates the QM barrier lowering on (100)/(110), motivating reparameterization for low-temperature anisotropic etch simulations.

Methods

Static quantum chemistry (C)

  • Problem / structures / pathways: F\(_2\) dissociative chemisorption on fluorine-terminated Si (111), (100), and (110) surfaces; barrier and energy trends support a facet-dependent etch picture in the QM work.
  • Functional, dispersion, basis, k-sampling, properties: DFT program(s), exchange–correlation functional, dispersion correction, basis/cutoff, k-mesh, and reported barrier/energy metrics are N/A on this stub—read the ingested archive PDF (no DOI in front matter yet).

Classical molecular dynamics with REBO (B)

  • Model & comparison: REBO-class reactive bond-order Si/F chemistry; MD checks facet barrier ordering against QM and reports underestimation of QM barrier lowering on (100)/(110) without reparameterization.

Corpus / bibliographic note

DOI is unset in front matter—treat papers/Others/Si_F2_etching_archive_Kaganovic.pdf as archive provenance until a VOR is registered.

MD application (REBO) — integrated

Engine / code (typical in this line of work; confirm in the PDF): REBO-class reactive potential in a molecular dynamics code such as LAMMPS for F/Si surface dynamics. System & composition F\(_2\) + fluorine-terminated Si slabs; N/A — exact atom counts not restated in corpus metadata. PBC for slab models; N/A — field / strain not the focus. Ensemble: NVT-style constant-volume slab MD is typical for this REBO barrier sampling (read the archive for NVE/NPH if used). Duration / stages: N/A — equilibration and production lengths in ps/ns are not restated in this corpus note—see PDF. Timestep, thermostat, barostat, target temperature (K), target pressure, electric field, enhanced sampling (metadynamics/REX): N/A on this stub—see VOR when a DOI exists.

Findings

Facet-dependent QM picture

QM predicts strong differences in dissociative chemisorption barriers and fluorine uptake by facet, with Si(111) described as having comparatively slow Si–Si bond breaking vs (100) and (110) in the abstract-level summary.

Classical vs QM gap

REBO MD does not match the QM facet ordering for barrier lowering without reparameterization, motivating improved classical models for low-temperature anisotropic etch simulations. Corpus honesty: the present note follows the ingested archive PDF; VOR-level comparisons to plasma/experiment (where claimed) are out of scope until a DOI and full text are registered.

Limitations

Partial extraction; not ReaxFF; DOI absent—confidence stays low until a VOR is registered. Do not invent publisher metadata. The candidate_tags entry REBO-potential signals a classical reactive model distinct from ReaxFF workflows elsewhere in the corpus; keep taxonomy separation when indexing.

Reader notes (navigation)

When a peer-reviewed DOI is identified, update doi, venue, and confidence in the same edit and regenerate chunks so Phase 5 metadata stays coherent. Until then, prefer this page for archive provenance rather than classroom teaching.

Relevance to group

Adjacent reference for reactive silicon halogen surface chemistry and classical vs QM potential benchmarking.

Citations and evidence anchors

Curator note (2026-04-22)

This page was revised during Stage A · wave 004 · batch 02 paper curation to meet vanDuinWiki minimum body-length guidance for Phase 4 narrative notes, while keeping scientific statements scoped to the ingested PDFs and verified bibliographic anchors available in this repository. Maintenance should preserve the Summary / Methods / Findings spine, keep updated timestamps accurate when prose changes, and align confidence with extract coverage. For archive PDFs without a DOI, prefer expanding this page only when new primary sources are ingested rather than inferring journal metadata.