Interatomic potentials for SiO₂ including bond-bending terms

Evidence and attribution¶

Authority of statements

Prose sections below (Summary, Methods, Findings, etc.) are curated summaries of the publication identified by doi, title, and pdf_path in the front matter above. They are not new primary claims by this wiki.

For definitive numerical values, reaction schemes, and interpretations, use the peer-reviewed article (and optional records under normalized/papers/ when present)—not this page alone.

Summary¶

Sanders, Leslie, and Catlow extend classical Born-model silicate potentials by adding explicit O–Si–O bond-bending contributions so that α-quartz and other SiO\(_2\) polymorphs can be described without the well-known failures of pure two-body central-force models. The short-range Si–O and O–O interactions use a Buckingham form, while each O–Si–O angle receives a quadratic penalty \(E_B = K_B(\theta - \theta_0)^2\) with \(\theta_0\) set to the tetrahedral value (109.47°) to represent the directional component of sp\(^3\)-like bonding at silicon. Oxygen shell-model polarizability is retained so dielectric and lattice-dynamic data can be fit simultaneously. The Chem. Commun. communication argues that this extension yields quantitative agreement with experimental elastic constants, dielectric tensors, and phonon-related checks for quartz, whereas a two-body-only shell model fit to the same data remains inferior (including poor phonon dispersion behavior noted in the text).

Methods¶

Potential fitting / static lattice tests (checklist A/C)—this Chem. Commun. note develops a rigid-ion / shell SiO\(_2\) model and validates it with lattice dynamics and property calculations; no finite-temperature MD protocol is reported.

Functional form / lineage¶

Short-range: Buckingham interactions for Si–O and O–O; oxygen treated with the shell model for polarizability (as in prior Catlow/Sanders silicate work referenced in the communication).
Directionality: explicit O–Si–O bond-bending term \(E_B = K_B(\theta-\theta_0)^2\) with \(\theta_0=109.47^\circ\) to capture tetrahedral constraints beyond two-body central forces.

Training / optimization¶

Fitting: least-squares adjustment of \(K_B\), shell parameters, and Buckingham coefficients against experimental α-quartz data (elastic, dielectric, structural) with internal coordinates relaxed in the fit (Sec./tables in the communication).
Lattice dynamics: first and second derivatives constructed for phonon / dispersion evaluation as described in the paper (used to contrast bond-bending vs two-body-only fits).

Transferability tests (static / 0 K)¶

Pressure: α-quartz under hydrostatic pressure—tabulated Si–O–Si angles vs pressure (kbar) in the communication.
Polymorph survey: parameters carried to α-cristobalite, coesite, and tridymite to probe transferability of the bond-bending extension.

MD application (not reported)¶

This Chem. Commun. note develops and tests a lattice-dynamics / static-lattice model; it does not report finite-temperature molecular dynamics production trajectories. N/A — MD engine; N/A — atom count as a dynamical simulation cell; N/A — PBC for MD; N/A — NVE/NVT/NPT MD ensemble; N/A — timestep; N/A — MD duration; N/A — thermostat; N/A — barostat (hydrostatic pressure enters the static α-quartz survey in the communication, not an MD barostat); N/A — MD temperature schedule; N/A — MD stress control; N/A — electric field in MD; N/A — enhanced sampling. Grounding: papers/Others/Sanders_Leslie_Catlow_JChemSoc_SiO2_1984.pdf, normalized/extracts/2003sanders-venue-paper_p1-2.txt.

Force-field training (QM reference)¶

The communication emphasizes experimental α-quartz targets plus lattice-dynamics derivatives. QM / DFT reference: N/A — the indexed summary does not cite a DFT training database for this 1984 shell-model fit (contrast with modern plane-wave fits); parameters are adjusted by least-squares against experimental elastic, dielectric, and structural data together with phonon checks, as described in the primary PDF.

Findings¶

For α-quartz, the bond-bending model reproduces the experimental elastic constants and static/high-frequency dielectric entries in the authors’ Table 2 far better than the two-body-only counterpart fit to the same training set. Phonon dispersion for α-quartz is described as agreeing well with experiment when bond-bending is included, in contrast to the central-force-only parametrization. Pressure-dependent Si–O–Si angles also track measured trends (table of kbar versus angle in the communication). The authors conclude that bond-bending terms can often be grafted onto existing central-force shell models without refitting the repulsive Si–O parameters, simplifying adoption in zeolite and feldspar simulations.

Limitations¶

The model remains fixed-charge / shell ionic mechanics—reactive bond formation, proton transport, and covalent charge redistribution require ReaxFF or QM methods. extraction_quality is partial in metadata because some corpus tooling predates clean full-text extracts; consult papers/Others/Sanders_Leslie_Catlow_JChemSoc_SiO2_1984.pdf for tables.

Relevance to group¶

Historical silica FF lineage relevant to oxide and silicate simulation traditions that intersect ReaxFF Si/O parameterization literature.

Citations and evidence anchors¶

DOI: https://doi.org/10.1039/C39840001271 — J. Chem. Soc., Chem. Commun., 1271 (1984).

reaxff-family