[Paper Review] Re-grouping stars based on the chemical tagging technique: A case study of M67 and IC4651

S. Blanco-Cuaresma, C. Soubiran|arXiv (Cornell University)|Sep 29, 2016

Stellar, planetary, and galactic studies3 citations

TL;DR

This study tests the chemical tagging technique by analyzing 28-element chemical abundances in stars from the open clusters M67 and IC4651 using high-resolution spectroscopy and an improved iSpec pipeline. Despite high-precision measurements (better than 0.05 dex), the chemical signatures of both clusters remain indistinguishable, suggesting that even with expanded elemental coverage, chemical tagging cannot reliably separate stars from these two clusters, challenging the method's current limits.

ABSTRACT

The chemical tagging technique proposed by Freeman & Bland-Hawthorn (2002) is based on the idea that stars formed from the same molecular cloud should share the same chemical signature. Thus, using only the chemical composition of stars we should be able to re-group the ones that once belonged to the same stellar aggregate. In Blanco-Cuaresma et al. (2015), we tested the technique on open cluster stars using iSpec (Blanco-Cuaresma et al. 2014a), we demonstrated their chemical homogeneity but we found that the 14 studied elements lead to chemical signatures too similar to reliably distinguish stars from different clusters. This represents a challenge to the technique and a new question was open: Could the inclusion of other elements help to better distinguish stars from different aggregates? With an updated and improved version of iSpec, we derived abundances for 28 elements using spectra from HARPS, UVES and NARVAL archives for the open clusters M67 and IC4651, and we found that the chemical signatures of both clusters are very similar.

Motivation & Objective

• To assess whether increasing the number of elements from 14 to 28 improves the ability to chemically distinguish stars from different open clusters.
• To evaluate if the chemical tagging technique can successfully re-group stars from M67 and IC4651 based solely on their chemical compositions.
• To test a novel differential abundance analysis method using cluster-specific reference stars (M67 No164 and M67 No1194) to minimize systematic biases between dwarfs and giants.
• To investigate whether the observed chemical similarity between M67 and IC4651 implies a shared origin or common enrichment history.

Proposed method

• High-resolution spectroscopic data from HARPS, UVES, and NARVAL archives were used, with radial velocity filtering and co-addition of spectra from the same instrument to improve signal-to-noise (S/N > 100).
• An automated spectroscopic pipeline based on iSpec was employed, using SPECTRUM as the radiative transfer code and MARCS model atmospheres with Grevesse et al. (2007) solar abundances.
• Atmospheric parameters (Teff, log g, [Fe/H]) were derived using selected lines in the 480–680 nm range, with iterative refinement and line-by-line abundance analysis.
• Differential abundances were computed using M67 No164 (giant) and M67 No1194 (dwarf) as reference stars to minimize systematic differences between evolutionary stages.
• Lines were rejected if they showed deviations >0.10 dex under metallicity perturbation, S/N variation, or blend-free synthesis, ensuring robustness.
• Abundance uncertainties were estimated using median absolute deviation (MAD) to maintain conservative error estimates.

Experimental results

Research questions

• RQ1 Can the inclusion of 28 elements instead of 14 improve the ability to chemically distinguish stars from M67 and IC4651?
• RQ2 Is the chemical tagging technique capable of reliably re-grouping stars from two distinct open clusters based solely on their chemical compositions?
• RQ3 Do systematic differences between dwarfs and giants in abundance analysis invalidate the use of a single solar reference, and can cluster-specific references mitigate this?
• RQ4 Why do M67 and IC4651 exhibit such similar chemical signatures despite being separated by over 100° in galactic longitude and potentially different ages?

Key findings

• The chemical signatures of M67 and IC4651 are indistinguishable across 28 elements, with most elements showing dispersions below 0.05 dex, indicating high measurement precision.
• Even the most distinct elements—barium, praseodymium, and sulfur—exhibited large dispersions (>0.05 dex), preventing clear separation of the two clusters.
• Individual star-by-star abundance plots showed extensive overlap between M67 and IC4651 stars in the 28-element chemical space, making cluster membership assignment visually impossible.
• The use of cluster-specific reference stars (M67 No164 and M67 No1194) successfully minimized systematic biases between dwarfs and giants, enabling combined analysis.
• Despite high-precision measurements, the chemical homogeneity between M67 and IC4651 suggests either a shared origin from a single, long-lived molecular cloud or independent enrichment by similar events.
• The results challenge the current limits of the chemical tagging technique, indicating that chemical composition alone may not suffice to distinguish stars from otherwise similar clusters.