Meeting Report

A Technical Meeting on Standards and Software Tools for Atomic and Molecular Databases, was held in room M0E-100 at IAEA Headquarters in Vienna from 27 – 29 November 2019. The Scientific Secretary was Christian HILL.

Overview

The goals of this meeting are:

• To expand, update and clarify the classification of collisional processes in plasmas given in the 2003 report written by Humbert et al. [pdf] outlining the recommendations of the IAEA Data Centres Network. Of particular interest is the inclusion of a more complete set of plasma-surface interaction processes.

• To codify a set of standards for unambiguously identifying:

  • Atoms and ions and their isotopes; small molecules and molecular ions and their isotopologues; photons and subatomic particles of relevance to plasma processes;
  • The rotational, vibrational, and electronic states of atoms, ions and (small) molecules.

Some progress has been made at the Atomic and Molecular Data Unit to implement a set of text-based conventions on its databases: a demonstration of the PyValem application for describing, validating and transforming reactions, species and states of relevance to atomic and molecular physics is available online.

• To recommend relational database structures for storing and software for parsing and querying information relating to atomic and molecular data.

• To recommend a nomenclature for interconversion between unit systems considered as equivalent in the fields of spectroscopy and collisional physics (e.g. frequency ≌ wavenumber ≌ wavelength ≌ photon energy).

Examples and Motivation

Collisional processes

Some of the processes missing from the existing recommended classification are:

• Collisions involving a change of excitation, particularly by heavy-particle collisions, e.g. Ar^* + O → Ar + O^*

• Various kinds of molecular dissociation and reaction, e.g. N₃⁺ → N₂ + N⁺, O + O₃^- → e^- + O₂ + O₂;

• Reactions occurring on surfaces with or without desorption of the product;

• Distinctions between changes of excitation to a specific state vs. a generic or unknown state (e.g. Ar + e^- → Ar^* + e- vs. Ar + e^- → Ar 4s[3/2]₂);

• Rotational and vibrational excitation of molecules;

• Processes for which the products may be partially unknown or a mixture of stoichiometric possibilities.

Identification of particles, atoms and molecules and their states

• Journals, in particular, have incompatible and inconsistently applied conventions for referring to these species: e.g. “W LXIV” vs. “Na-like W” vs. “W+63” vs. “W63+” vs. “W^{+63}”, “W⁺⁶³” and many more combinations, making searching difficult and error-prone.

A recommendation for identifying such species in a database and tools for transforming the output according to different conventions would help here. We might think of hosting a database to resolve (as far as possible) different conventions into their canonical form;

• The same applies (more so, in fact), to atomic and molecular states. Where MathML or images are used, there is rarely any semantic metadata attached to these entities to make them meaningful to parsers. For example, the title “Dipole-allowed transitions between the X3Σ−, A′3∆, A′′3Σ+, A3Π, B3Σ−, and C3Π states of the sulfur monoxide radical”, [JQSRT 234, 98 (2019)] refers to several electronic states of a diatomic molecule, but cannot be reliably transcribed into a bibliography and the state labels are not easily searched. One doesn't have to look far to find examples of awkward implementations of incompatible conventions here:

• No agreed convention exists to specify the extent to which an atomic or molecular state is known: it would be helpful to have a distinct notation for an “unspecified” state (a species identified only by its stoichiometry and (in the case of molecules) connectivity), an “unknown” state, a partially-specified state (e.g. “all rotational states within a specified vibrational state”, “all rovibrational states belonging to a given electronic state”, etc.)

• A great many standards already exist for the description of individual species (InChI[Key], SMILES, SYBYL, and so on), and the Virtual Atomic and Molecular Data Centre (VAMDC) consortium has developed the XSAMS standard for data exchange in atomic and molecular physics. The goal of this meeting is not to recommend yet another alternative standard, but rather to suggest a practical, unambiguous, text-based internal representation of relevant quantities for online databases, and to motivate the development of software tools for the interconversion between existing standards.

Unit conventions in spectroscopy and collisional physics

It would helpful to specify precisely what is meant by various unit conversions in the context of particular fields, especially those which are not dimensionally correct: for example, the transformation from a photon’s (vacuum) wavelength to its frequency and energy; conversion between the wavelength and wavenumber of a spectroscopic transition, etc. Of relevance to this is the long-standing and unresolved debate about dimensionless units in the SI sysem, such as the radian [1, 2]: the status quo of distinguishing the implied units of periodic and cyclic frequencies (related by a factor of 2π) through the symbol used to denote a physical quantity (typically ν or ω) is unsatisfactory.

[1] P. R. Bunker et al., "The Planck Constant and its Units", Journal of Quantitative Spectroscopy and Radiative Transfer 237, 106594 (2019).

[2] P. J. Mohr and W. D. Phillips, "Dimensionless units in the SI", Metrologia 52, 40 (2015).