Linemake: An Atomic and Molecular Line List Generator

Vinicius M. Placco; Christopher Sneden; Ian U. Roederer; James E. Lawler; Elizabeth A. Den Hartog; Neda Hejazi; Zachary Maas; Peter Bernath

doi:10.3847/2515-5172/abf651

1. Introduction

Stellar and Galactic chemical evolution can be illuminated by abundance analyses, but such studies must strive for abundance accuracy. This effort depends on many factors, but none is more crucial than access to trustworthy basic atomic and molecular transition data. Several excellent large transition databases exist, the most well-known being VALD.⁹ Here we introduce the utility code linemake, created to generate synthetic spectrum input line lists by merging Kurucz¹⁰ (Kurucz 2011) atomic/molecular line compendium information with updated and very accurate transition probabilities, hyperfine structure (HFS), and isotopic substructure data. The atomic data included in linemake are primarily those that have been published by the University of Wisconsin atomic physics group (e.g., Lawler et al. 2009, among others) and the molecular data are from the Old Dominion University molecular physics group (Bernath 2020). The Wisconsin data now include nearly all Fe- and lanthanide-group elements, and recently a study of Ca i initiates work on lighter elements.

With increased interest in low-temperature stars (spectral types mid-K and cooler) it has become attractive to include molecules such as TiO and H₂O in linemake. These molecules have substantial numbers of transitions available in large databases but are not easy to incorporate in synthetic spectrum line lists. They are included in linemake with choices that keep the number of transitions to a manageable size. linemake does not try to compete with the VALD resource or other more comprehensive line compendia. Our effort is to produce a carefully controlled line list of easily referenced sources from laboratory physics. In this document, we outline linemake and point to the sources of its data.

2. About `linemake`

linemake produces synthesis line lists compatible with those needed by the line analysis code MOOG ¹¹ (Sneden 1973). It starts with the Kurucz compendium and then substitutes or supplements these lists with atomic data from the Wisconsin group and molecular data from the Old Dominion group. linemake is written in standard FORTRAN, and it can be compiled and executed on all UNIX/Linux-based operating systems.

linemake is primarily aimed for spectroscopic studies of stars cooler than B spectral type. Some limitations in the available line lists have been imposed in order to keep the resulting synthesis lists to sensible sizes. The ionization states available are neutral and first ion. The maximum lower excitation energy of any line is 7.5 eV, except for some transitions of the light elements H, C, N, O, Mg, Al, Si, P, and S; higher excitation transitions are available for these elements. Additionally, for Fe ii, 8.5 eV is the maximum lower energy. Obvious warnings should be given about the output line lists: we believe that they are correct, but there is no substitute for users having a close look to assure themselves of the quality of these lists. If a user is uncertain about an output line list, the individual files for different species can be examined easily. These species files are also useful in simply making laboratory-based reliable transition lists for individual (mostly atomic) species.

A few "local rules" apply when using the code. First and most importantly, in almost all cases in which an atomic (and more often molecular) species is represented by multiple isotopes in its transitions, linemake presents the lists with isotopic identification but no assumptions about isotopic ratios. For a given transition, the total gf value for each isotope is the same. When applying spectrum synthesis codes like MOOG to the line lists generated by linemake, users must set the desired isotopic fractions in parameter files. second, linemake catalogs some Fe-group hyperfine substructure patterns from the Kurucz database when no recent laboratory studies are available. These can be included in output files at the user's option, but they will probably add many extra transitions into the synthesis line lists, and the pedigrees of these are not guaranteed. Users are cautioned to be careful about including them and are encouraged to test them with stellar syntheses.

3. Brief Description of the Database

Here we briefly discuss the atomic and molecular data sources. The detailed description for each category, including literature references and decisions made to maximize the utility of line lists for high-resolution spectroscopic studies, are given in the README.md file of the GitHub repository. Figure 1 summarizes the transition availability in the linemake database, and its most up-to-date version can also be found in the README.md file.

**Figure 1.** Elements with curated transitions currently available in `linemake`. The most up-to-date version of this periodic table is available at the GitHub repository.
Start interaction Open in new tab
Download figure:
Standard image High-resolution image Figure data file

3.1. Atomic Species: Fe-group Elements

The Fe group is defined here as those elements with $21\leqslant Z\leqslant 30$ . Significant HFS usually is a feature of transitions of odd-Z elements. If the README.md description indicates that a species includes HFS, then line lists have full HFS substructure patterns when they are available from laboratory studies. Typically, these are known for many of the lines with laboratory transition probabilities. linemake manages these data internally, but users should examine the output synthesis lists to understand whether HFS patterns have been included in any transition of interest. Information about isotopic substructure is scarce. This is largely an issue for even-Z elements, as odd-Z elements have few (often just one) naturally occurring isotopes. Moreover, for Fe-group elements, usually one isotope dominates the solar system abundance, and other isotopes can safely be ignored. Exceptions are Ti and Ni, and users should account for isotopic effects on transitions in the red and infrared spectral regions.

3.2. Atomic Species: Neutron-capture Elements

Neutron-capture elements are defined as those with $Z\gt 30$ . For lanthanide elements and Hf ( $57\leqslant Z\leqslant 72$ ), the Wisconsin group results dominate, but for lighter and heavier elements other sources are included if they have been subject to recent assessment for reliability. In this case, NIST refers to the National Institute of Standards and Technology's Atomic Spectra Database.¹²

3.3. Atomic Species: Other Elements

Light elements ( $Z\leqslant 20$ ) have often not had recent extensive laboratory transition probability studies. Some progress, both in laboratory and theoretical, has been made, and a few selected elements are included in the linemake database.

3.4. Molecular (Mostly Diatomic) Species

Line data for molecular species have improved considerably recently. Motivated by an interest in M-type stars, which are often involved in exoplanet radial velocity or atmospheric transmission studies, we have included molecules that only appear in stellar spectra of very cool stars. The data sources for many molecules are laboratory studies from Peter Bernath's group (MoLLIST ¹³ ). For others we have translated into MOOG format the line information from two prominent molecular physics groups: HITRAN ¹⁴ and EXOMOL.¹⁵

4. Conclusions

In this note, we have briefly described the main features of the linemake code. This discussion is not intended to be exhaustive, but we hope it may serve as a helpful guideline for potential users. The GitHub database will continue to be updated with the most recent laboratory and theoretical work, and input from the user community is certainly welcome.

Linemake: An Atomic and Molecular Line List Generator

Article metrics

Author affiliations

ORCID iDs

Dates

Abstract

1. Introduction

2. About `linemake`