Looking for Companionship: Radial Velocity Follow-Up of Lithium-Rich Giants with ESPRESSO

arXiv:2510.17966 · astro-ph.SR, astro-ph.EP · Submitted 2025-10-20 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.

Vera: I'm Vera.

Jocelyn: And I'm Jocelyn, here with Subrahmanyan.

Vera: We're looking at the paper: "Looking for Companionship".

Jocelyn: Here's the gist of the paper. Lithium-rich red giants have been a long-standing mystery in stellar astrophysics because interactions with a close companion are a leading theory to explain these chemically peculiar and rare objects The overall companion rate found is 27% (9/33) Secondary masses reveal one planetary companion (M sin i ≈ 7 MJup), three brown dwarfs (M sin i = 30 − 33 MJup), and five stellar-mass companions (M sin i = 0.2 − 0.8 M⊙) Our findings suggest that Li-rich red giants with lower lithium abundance (A(Li) ≈ 1.5 dex) tend to host companions as compared to those with higher lithium abundance, and Li-rich red giants with log g = 2−3 dex have a higher companion rate than those outside of this range We offer two potential formation mechanisms of our Li-rich sample: (i) the progenitor mass of stellar mass companions suggest that these objects were potentially lithium-producing, intermediate-mass AGB stars; (ii) the sub-stellar companions were initially in multi-planet systems, but dynamical instability caused the tidal dissipation of close-in planets thereby enhancing the red giant in lithium Extended baselines and dedicated follow-up with Gaia DR4 astrometry are required to confirm the orbital parameters of our systems and distinguish between mechanisms Sample Selection and Data Acquisition The research focused on collecting high-resolution spectra of 33 Li-rich red giants using ESPRESSO and constraining their orbital parameters with The Joker Our ESPRESSO targets were chosen from the parent sample of Li-rich giants in Sayeed et al.

Vera: (2024) The ESPRESSO target sample covers the overall distribution of the reference Li-rich sample along the red giant branch, with effective temperature Teff =

3000, 5730: K, surface gravity log g =

−1, 3.2: dex, and A(Li) ≥ 1.5 dex The ESPRESSO data was obtained using programs P112-X or P113-X over several observation periods Radial velocity (RV) measurements were obtained from ESPRESSO’s own Data Reduction Software The average SNR of the spectrum at 550 nm ranged from 12–94 per star for P112 (µ = 31) and 5–106 per star for P113 (µ = 25) Orbital Parameter Derivation The Joker was used to sample possible orbital solutions consistent with the data The Joker describes a radial velocity curve by six parameters: the period (P), eccentricity (e), pericenter phase (ϕ0) and argument (ω), velocity semi-amplitude (K), and barycenter velocity (v0), respectively The Joker incorporates and infers an unknown radial velocity scatter as an overall “jitter” parameter, s squared, which we also vary as a free parameter We used The Joker’s default prior pdf for all six parameters as detailed below Binary Candidate Identification A selection criterion was implemented using the framework developed by Hekker et al.

Jocelyn: (2008) to select targets that are more likely to be binaries For each of our targets with a given surface gravity from GALAH, we calculate the expected semi-amplitude We classify a red giant as a binary candidate if its 5th percentile semi-amplitude is above 3.5σ of the expected semi-amplitude (Equation 12) Based on this criterion, nine of our 33 targets satisfied this condition, including the two uni-modal targets Interpretation of Results We find Li-rich giants with a lower Li abundance (A(Li) = 1.5 − 2.0 dex) have a higher binary rate (78%) than Li-rich red giants with A(Li) > 2.0 dex This suggests that Li-rich giants above this threshold most likely become Li-enhanced via other mechanisms than binarity By deriving companion masses, we find one potential planetary companion, three brown dwarf companions, and five stellar mass companions Two of our nine systems are uni-modal with orbital periods of 3.5 and 4.3 days, and eccentricities of 0.6 and 0.25 All our binary candidates have log g = 2 − 3 dex, with a binary rate of 43% in this log g range Formation Mechanisms and Future Work We offer two potential formation mechanisms based on our nine binary candidates Firstly, progenitor masses of five stellar companions reveal these systems could have been intermediate-mass asymptotic giant branch stars that produced lithium in their interiors, and enriched the red giant companion in lithium via mass transfer Secondly, the four sub-stellar mass companions (three brown dwarfs and one planetary mass) could have enriched the red giant in lithium via planet engulfment if the system was initially a multi-body system Extended baselines and additional RV data for larger samples of Li-rich giants are crucial in interpreting the significance of our companion binary fraction The upcoming Gaia DR4 survey will be pivotal in uncovering hidden companions around Lirich giants While we selected our targets to span relevant parameter spaces (e.g., lithium abundance, evolutionary state, RUWE), we were still only limited to 33 targets Extended baselines and additional RV data for larger samples of Li-rich giants are crucial in interpreting the significance of our companion binary fraction The similarity in companion fraction between a sub-sample of Lirich giants and a background population of binaries suggests that either the true binarity rate of Li-rich red giants is near 100%, or that the underlying mass or semi-major axis distribution of red giant binaries in APOGEE DR16 is different from that of companions to Li-rich giants Potential Formation Scenarios The analysis suggests that if Li-rich red clump stars become Li-rich due to binarity, their low lithium abundance can be due to low mass transfer rates Corehelium burning red giants typically have log g within 2–3 dex Since almost half of the targeted stars in this range are binary candidates, it is possible Lirich red giants on the red clump become Li-enhanced due to binarity while red giants on the tip of the giant branch or those at the base of the giant branch become Li-enhanced via other mechanisms Tidal Effects and Companion Mass Constraints We use The Joker’s samples of semi-amplitude, orbital period, and eccentricity to calculate ϵ where we use the median for each target as the value of ϵ The median ϵ in our sample of 33 targets is 10−5, with a range from 5 × 10−8 − 2 In comparison, the tidal parameter for Moon’s orbit around the Earth is 6.4 × 10−8 The tidal parameter for our uni-modal systems are therefore orders of magnitude larger than known tidally interacting star–planet systems, suggesting these systems might be undergoing tidal interactions Interesting Systems TIC 405924915 shows a period of 3.47+0.07 −0.00 days, or a separation of 0.10 ± 0.005 AU, and an eccentricity of 0.59+0.6 −- - -8 Given that evolved stars rotate slowly, the fast rotation period is unusual Combined with its tidal parameter of 0.25, orders of magnitude larger than known tidally interacting planet-star systems, this system could be tidally interacting with a close companion, inducing the Cameron–Fowler mechanism, and causing Lienrichment We find Li-rich giants with a lower Li abundance (A(Li) = 1.5 − 2.0 dex) have a higher binary rate (78%) than Li-rich red giants with A(Li) > 2.0 dex By deriving companion masses, we find one potential planetary companion, three brown dwarf companions, and five stellar mass companions Two of our nine systems are uni-modal with orbital periods of 3.5 and 4.3 days, and eccentricities of 0.6 and 0.

Paper summary: Vera: So, wrapping up the whole thing with "Looking for Companionship: Radial Velocity Follow-Up of Lithium-Rich Giants with ESPRESSO", the authors are really emphasizing that while they looked at a limited sample of thirty-three targets, their findings suggest that either the true binary rate of lithium-rich red giants is near one hundred percent, or the underlying mass or semi-major axis distribution for red giant binaries in APOGEE DR16 is different from what they expected.

Jocelyn: The paper really drives home that we need extended baselines and additional radial velocity data for larger samples of lithium-rich giants to properly interpret the significance of their companion binary fraction.

Subrahmanyan: And the upcoming Gaia DR4 survey is mentioned as being pivotal in uncovering hidden companions around these lithium-rich giants, which connects this finding back to broader structure formation models.

Vera: It suggests that whether a red giant gets its lithium from mass transfer or tidal interaction depends heavily on where it sits along the red giant branch and what kind of companion it has.

Jocelyn: The title itself, "Looking for Companionship: Radial Velocity Follow-Up of Lithium-Rich Giants with ESPRESSO", sets up a clear direction for future work in confirming these dynamical interactions.

Conclusion: Vera: So, we've seen how these lithium-rich giants interact with companions through their radial velocity measurements and now we're looking at what this paper actually says about those interactions.

Jocelyn: The title is "Looking for Companionship: Radial Velocity Follow-Up of Lithium-Rich Giants with ESPRESSO", which tells you exactly what they did—they used the ESPRESSO instrument to track companions around these specific stars.

Subrahmanyan: The authors are looking at lithium-rich red giants and trying to figure out if those companions actually exist and how they got there.

Vera: It boils down to this finding that Li-rich giants with lower lithium abundance, A(Li) equal to one point five minus two point zero dex, have a much higher binary rate of seventy eight percent than those with higher lithium abundance.

Jocelyn: That suggests that for the stars with more lithium, they get their enhancement through other processes besides having a close companion.

Subrahmanyan: And when they do find companions, the masses range from one planetary companion up to five stellar mass companions, which is a lot of mass for these objects.

Vera: They also found two uni-modal systems with orbital periods of three point five and four point three days and eccentricities of zero point six and zero point two five.

Jocelyn: Those specific numbers give us a snapshot, but what it really means is that we're seeing a pattern in these systems that we need more data to confirm.

Subrahmanyan: The paper suggests two formation routes, one where the progenitor star was an intermediate-mass AGB star that enriched the companion with lithium during mass transfer.

Vera: And the second route is if sub-stellar companions were initially in multi-planet systems and dynamical instability caused tidal dissipation, which enhanced the red giant's lithium.

Jocelyn: They also pointed out that extended baselines and more radial velocity data for larger samples are crucial to properly interpret what this binary fraction actually tells us.

Subrahmanyan: The next step is waiting for the Gaia DR4 survey, because that survey will be pivotal in uncovering hidden companions around these lithium-rich giants.

Maryum Sayeed, Andrew R. Casey, Benjamin T. Montet, Melissa K. Ness, Adrian M. Price–Whelan, Daniel Huber,  Madeline J. Maldonado Gutierrez

Department of Astronomy, Columbia University · School of Physics & Astronomy, Monash University · Center for Computational Astrophysics, Flatiron Institute · School of Physics, University of New South Wales · Research School of Astronomy & Astrophysics, Australian National University · Institute for Astronomy, University of Hawai’i Department of Physics and Astronomy, Barnard College

astro-ph.SR, astro-ph.EP

Submitted: 2025-10-20

Updated: 2026-10-27

Comments: 26 pages, 12 figures, re-submitted to ApJ. Comments welcome. An electronic version of Tables 1 & 2 will be available after acceptance

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 71/100

The gist: Lithium-rich red giants have been a long-standing mystery in stellar astrophysics because interactions with a close companion are a leading theory to explain these chemically peculiar and rare…

Key concepts

Lithium Abundance (A(Li))
This measures the amount of lithium present in the star compared to a standard reference. The study found that stars with lower lithium levels are more likely to have been affected by companions, suggesting they might have gained their lithium through interaction.
Binary Candidate Selection
A method used to identify stars that are likely in a binary system based on their measured radial velocity. If a star's measured velocity variation is significantly higher than expected for its surface gravity, it is flagged as a potential binary partner.
Companion Mass Derivation
By analyzing the orbital parameters like period and semi-amplitude from radial velocity measurements, researchers calculated the masses of detected companions. The results included one planetary mass, three brown dwarf masses, and five stellar mass companions.
Tidal Parameter ($\epsilon$)
This parameter quantifies the strength of tidal interactions between a star and its companion. The study found that the tidal parameters for their systems are much larger than known planet-star interactions, suggesting these systems might be undergoing significant tidal forces.

Terminology

Summary

Lithium-rich red giants have been a long-standing mystery in stellar astrophysics because interactions with a close companion are a leading theory to explain these chemically peculiar and rare objects The overall companion rate found is 27% (9/33) Secondary masses reveal one planetary companion (M sin i ≈ 7 MJup), three brown dwarfs (M sin i = 30 − 33 MJup), and five stellar-mass companions (M sin i = 0. p. 0

Where the paper says it

Secondary masses reveal one planetary companion (M sin i ≈ 7 MJup), three brown dwarfs (M sin i = 30 − 33 MJup), and five stellar-mass companions (M sin i = 0. Our findings suggest that Li-rich red giants with lower lithium abundance (A(Li) ≈ 1.

2 − 0.8 M⊙) Our findings suggest that Li-rich red giants with lower lithium abundance (A(Li) ≈ 1. p. 0
Where the paper says it

Secondary masses reveal one planetary companion (M sin i ≈ 7 MJup), three brown dwarfs (M sin i = 30 − 33 MJup), and five stellar-mass companions (M sin i = 0. Our findings suggest that Li-rich red giants with lower lithium abundance (A(Li) ≈ 1.

5 dex) tend to host companions as compared to those with higher lithium abundance, and Li-rich red giants with log g = 2−3 dex have a higher companion rate than those outside of this range We offer two potential formation mechanisms of our Li-rich sample: (i) the progenitor mass of stellar mass companions suggest that these objects were potentially lithium-producing, intermediate-mass AGB stars; (ii) the sub-stellar companions were initially in multi-planet systems, but dynamical instability caused the tidal dissipation of close-in planets thereby enhancing the red giant in lithium Extended baselines and dedicated follow-up with Gaia DR4 astrometry are required to confirm the orbital parameters of our systems and distinguish between mechanisms

Sample Selection and Data Acquisition

The research focused on collecting high-resolution spectra of 33 Li-rich red giants using ESPRESSO and constraining their orbital parameters with The Joker Our ESPRESSO targets were chosen from the parent sample of Li-rich giants in Sayeed et al. (2024) The ESPRESSO target sample covers the overall distribution of the reference Li-rich sample along the red giant branch, with effective temperature Teff = [3000, 5730] K, surface gravity log g = [−1, 3. p. 2

Where the paper says it

Our ESPRESSO target sample covers the overall distribution of the reference Li-rich sample along the red giant branch. rich giants as objects with effective temperature Teff = [3000, 5730] K, surface gravity log g = [−1, 3.

2] dex, and A(Li) ≥ 1. p. 0
Where the paper says it

Our findings suggest that Li-rich red giants with lower lithium abundance (A(Li) ≈ 1. 5 dex) tend to host companions as compared to those with higher lithium abundance, and Li-rich red giants with log g = 2−3 dex have a higher companion rate than those outside of this range.

5 dex The ESPRESSO data was obtained using programs P112-X or P113-X over several observation periods Radial velocity (RV) measurements were obtained from ESPRESSO’s own Data Reduction Software The average SNR of the spectrum at 550 nm ranged from 12–94 per star for P112 (µ = 31) and 5–106 per star for P113 (µ = 25)

Orbital Parameter Derivation

The Joker was used to sample possible orbital solutions consistent with the data The Joker describes a radial velocity curve by six parameters: the period (P), eccentricity (e), pericenter phase (ϕ0) and argument (ω), velocity semi-amplitude (K), and barycenter velocity (v0), respectively The Joker incorporates and infers an unknown radial velocity scatter as an overall “jitter” parameter, s squared, which we also vary as a free parameter We used The Joker’s default prior pdf for all six parameters as detailed below

Binary Candidate Identification

A selection criterion was implemented using the framework developed by Hekker et al. (2008) to select targets that are more likely to be binaries For each of our targets with a given surface gravity from GALAH, we calculate the expected semi-amplitude We classify a red giant as a binary candidate if its 5th percentile semi-amplitude is above 3. p. 5

Where the paper says it

For each of our targets with a given surface gravity from GALAH, we calculate the expected semi-amplitude. In this work, we classify a red giant as a binary candidate if its 5th percentile semi-amplitude is above 3.

5σ of the expected semi-amplitude (Equation 12) Based on this criterion, nine of our 33 targets satisfied this condition, including the two uni-modal targets

Interpretation of Results

We find Li-rich giants with a lower Li abundance (A(Li) = 1.5 − 2.0 dex) have a higher binary rate (78%) than Li-rich red giants with A(Li) > 2. p. 0

Where the paper says it

Our findings suggest that Li-rich red giants with lower lithium abundance (A(Li) ≈ 1. 5 dex) tend to host companions as compared to those with higher lithium abundance, and Li-rich red giants with log g = 2−3 dex have a higher companion rate than those outside of this range.

p. 0
Where the paper says it

Our findings suggest that Li-rich red giants with lower lithium abundance (A(Li) ≈ 1. 5 dex) tend to host companions as compared to those with higher lithium abundance, and Li-rich red giants with log g = 2−3 dex have a higher companion rate than those outside of this range.

0 dex This suggests that Li-rich giants above this threshold most likely become Li-enhanced via other mechanisms than binarity By deriving companion masses, we find one potential planetary companion, three brown dwarf companions, and five stellar mass companions Two of our nine systems are uni-modal with orbital periods of 3. p. 0
Where the paper says it

Secondary masses reveal one planetary companion (M sin i ≈ 7 MJup), three brown dwarfs (M sin i = 30 − 33 MJup), and five stellar-mass companions (M sin i = 0. Our findings suggest that Li-rich red giants with lower lithium abundance (A(Li) ≈ 1.

5 and 4.3 days, and eccentricities of 0. p. 9
Where the paper says it

The vertical structure present in non-zero eccentricities are due to time sampling where some orbital solutions are better for detecting systems with specific periods. Given ESPRESSO data, our pipeline cannot detect any companions below 1 M⊙ on circular orbits above 40 days.

p. 9
Where the paper says it

The vertical structure present in non-zero eccentricities are due to time sampling where some orbital solutions are better for detecting systems with specific periods. Given ESPRESSO data, our pipeline cannot detect any companions below 1 M⊙ on circular orbits above 40 days.

6 and 0.25 All our binary candidates have log g = 2 − 3 dex, with a binary rate of 43% in this log g range

Formation Mechanisms and Future Work

We offer two potential formation mechanisms based on our nine binary candidates Firstly, progenitor masses of five stellar companions reveal these systems could have been intermediate-mass asymptotic giant branch stars that produced lithium in their interiors, and enriched the red giant companion in lithium via mass transfer Secondly, the four sub-stellar mass companions (three brown dwarfs and one planetary mass) could have enriched the red giant in lithium via planet engulfment if the system was initially a multi-body system Extended baselines and additional RV data for larger samples of Li-rich giants are crucial in interpreting the significance of our companion binary fraction The upcoming Gaia DR4 survey will be pivotal in uncovering hidden companions around Lirich giants

Conclusion and Implications

While we selected our targets to span relevant parameter spaces (e.g., lithium abundance, evolutionary state, RUWE), we were still only limited to 33 targets Extended baselines and additional RV data for larger samples of Li-rich giants are crucial in interpreting the significance of our companion binary fraction The similarity in companion fraction between a sub-sample of Lirich giants and a background population of binaries suggests that either the true binarity rate of Li-rich red giants is near 100%, or that the underlying mass or semi-major axis distribution of red giant binaries in APOGEE DR16 is different from that of companions to Li-rich giants

Potential Formation Scenarios

The analysis suggests that if Li-rich red clump stars become Li-rich due to binarity, their low lithium abundance can be due to low mass transfer rates Corehelium burning red giants typically have log g within 2–3 dex Since almost half of the targeted stars in this range are binary candidates, it is possible Lirich red giants on the red clump become Li-enhanced due to binarity while red giants on the tip of the giant branch or those at the base of the giant branch become Li-enhanced via other mechanisms

Tidal Effects and Companion Mass Constraints

We use The Joker’s samples of semi-amplitude, orbital period, and eccentricity to calculate ϵ where we use the median for each target as the value of ϵ The median ϵ in our sample of 33 targets is 10−5, with a range from 5 × 10−8 − 2 In comparison, the tidal parameter for Moon’s orbit around the Earth is 6.4 × 10−8 The tidal parameter for our uni-modal systems are therefore orders of magnitude larger than known tidally interacting star–planet systems, suggesting these systems might be undergoing tidal interactions

Interesting Systems

TIC 405924915 shows a period of 3.47+0.07 −0.00 days, or a separation of 0. p. 1

Where the paper says it

2020) to investigate resulting binary architectures of Li-rich giants based on GALAH observations. Under the assumption of a binary scenario where the red giant is enriched by an intermediate-mass AGB star, their simulations found that these systems are found at an average separation of 3.

10 ± 0.005 AU, and an eccentricity of 0. p. 3
Where the paper says it

2017, 2020). The Joker describes a radial velocity curve by six parameters: the period (P), eccentricity (e), pericenter phase (ϕ0) and argument (ω), velocity semi-amplitude (K), and barycenter velocity (v0), respectively.

59+0.6 −- - -8 Given that evolved stars rotate slowly, the fast rotation period is unusual Combined with its tidal parameter of 0. p. 1
Where the paper says it

(2023) combined radial velocity data from Gaia, GALAH, and RAVE to test the binary fraction of Li-rich giants on the first ascent red giant branch as compared to those on the red clump. They conclude red clump stars do not have a preference for being in binary systems, but they emphasize that the accuracy of their analysis is highly dependent on the precision of their radial velocity measurements.

25, orders of magnitude larger than known tidally interacting planet-star systems, this system could be tidally interacting with a close companion, inducing the Cameron–Fowler mechanism, and causing Lienrichment

Summary of Findings

We find Li-rich giants with a lower Li abundance (A(Li) = 1.5 − 2.0 dex) have a higher binary rate (78%) than Li-rich red giants with A(Li) > 2.0 dex By deriving companion masses, we find one potential planetary companion, three brown dwarf companions, and five stellar mass companions Two of our nine systems are uni-modal with orbital periods of 3. p. 0

Where the paper says it

We find an overall companion rate of 27% (9/33). Secondary masses reveal one planetary companion (M sin i ≈ 7 MJup), three brown dwarfs (M sin i = 30 − 33 MJup), and five stellar-mass companions (M sin i = 0.

5 and 4.3 days, and eccentricities of 0.6 and 0.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved companion fraction estimation capability: The AI system can now estimate companion fractions by comparing Li-rich giants against a background population of red giants in APOGEE DR16 data, yielding an expected background binary fraction of 28 percent based on the similarity in companion fractions between sub-samples.

  2. Enhanced formation mechanism inference: The system can infer potential formation mechanisms by analyzing the derived companion masses; specifically, it can suggest that "progenitor masses of five stellar companions reveal these systems could have been intermediate-mass asymptotic giant branch stars that produced lithium in their interiors, and enriched the red giant companion in lithium via mass transfer."

  3. Identification of tidal interaction indicators: The AI system can calculate a dimensionless tidal parameter epsilon for each target and identify systems with orders of magnitude larger values than known tidally interacting planet-star systems, suggesting they might be undergoing tidal interactions.

Abstract

Lithium-rich red giants have been a long-standing mystery in stellar astrophysics. A leading theory to explain these chemically peculiar and rare objects is interactions with a close companion. To investigate their companion fraction, we collected high-resolution spectra of 58 Li-rich red giants using ESPRESSO, and constrained their orbital parameters using a Monte Carlo sampler, The Joker. We recover an overall companion rate of 17% (10/58) consistent with the 16% expected for normal red giants observed with ESPRESSO cadence and precision in APOGEE DR16. All ten binary candidates have A(Li) = 1.5-2.7; dex and g = 2-3; dex; we find no binary candidates with A(Li) > 2.7; dex. Secondary masses reveal three planetary mass companions (M i about 7-11; M Jup), one brown dwarf (M i about 19; M Jup), and six stellar mass companions (M i= 0.08-0.98; M). The significant difference in companion fraction between low and high lithium at A(Li) about 2.7; dex suggests that there are at least two formation mechanisms for Li-rich giants: (i) red giants with low to moderate lithium (A(Li) about 1.5-2.7; dex) become lithium enhanced via planet engulfment, mass transfer, or the Cameron-Fowler mechanism via tidal spin-up; (ii) the dearth of companions around higher lithium giants (A(Li) 2.7; dex) could be due to stellar mergers or prior engulfment of companions. Extended baselines and dedicated follow-up are required to confirm the orbital parameters of these systems and distinguish between mechanisms.

Sources

Related papers