Mono-enriched stars and Galactic chemical evolution — Possible biases in observations and theory

A long sought after goal using chemical abundance patterns derived from metal-poor stars is to understand the chemical evolution of the Galaxy and to pin down the nature of the first stars (Pop III). Metal-poor, old, unevolved stars are excellent tracers as they preserve the abundance pattern of the gas from which they were born, and hence they are frequently targeted in chemical tagging studies. Here, we use a sample of 14 metal-poor stars observed with the high-resolution spectrograph called the Potsdam Echelle Polarimetric and Spectroscopic Instrument (PEPSI) at the Large Binocular Telescope (LBT) to derive abundances of 32 elements (34 including upper limits). We present well-sampled abundance patterns for all stars obtained using local thermodynamic equilibrium (LTE) radiative transfer codes and one-dimensional (1D) hydrostatic model atmospheres. However, it is currently well-known that the assumptions of 1D and LTE may hide several issues, thereby introducing biases in our interpretation as to the nature of the first stars and the chemical evolution of the Galaxy. Hence, we use non-LTE (NLTE) and correct the abundances using three-dimensional (3D) model atmospheres to present a physically more reliable pattern. In order to infer the nature of the first stars, we compare unevolved, cool stars, which have been enriched by a single event (‘mono-enriched’), with a set of yield predictions to pin down the mass and energy of the Pop III progenitor. To date, only few bona fide second generation stars that are mono-enriched are known. A simple χ 2 -fit may bias our inferred mass and energy just as much as the simple 1D LTE abundance pattern, and we therefore carried out our study with an improved fitting technique considering dilution and mixing. Our sample presents Carbon Enhanced Metal-Poor (CEMP) stars, some of which are promising bona fide second generation (mono-enriched) stars. The unevolved, dwarf BD+09_2190 shows a mono-enriched signature which, combined with kinematical data, indicates that it moves in the outer halo and likely has been accreted onto the Milky Way early on. The Pop III progenitor was likely of 25.5 M and 0.6 foe (0.6 1051 erg) in LTE and 19.2 M and 1.5 foe in NLTE, respectively. Finally, we explore the predominant donor and formation site of the rapid and slow neutron-capture elements.

Spectrum synthesis of C, Sm, (Nd), Dy, and Rb in various sample stars. Specifically, molecular and atomic C in TYC5481. We note that [C i/Fe] = 0.6±0.1; [CH/Fe] = 0.1 ± 0.1; in BD-0.1_2439, [Sm/Fe] = 0.3 ± 0.1 and the shown Nd line was not used in our average Nd value due to blends and a poor fit – synthesis shows [Nd/Fe] = 0.3, 0.8, 0.9; [Dy/Fe] = 0.5 ± 0.1; and in HD136343, [Rb/Fe] = 0.4 ± 0.1. In all cases, the green dashed line indicates [X/Fe] = −5.


Read more:  Hansen et al. 2020, A&A, in press

Pepsi Exoplanet Transit Survey Workshop

The Pepsi Exoplanet Transit Survey (PETS@LBT) Workshop is scheduled on

Tuesday Sep 22 – 6am to 9am AZ time (13:00-16:00 UTC)

In this Workshop, we will discuss and wrap-up our community proposal for a large LBT program, to be submitted shortly thereafter. We propose a high-resolution spectroscopic survey of exoplanet transits, secondary eclipses, and host-star characterizations of selected targets; dubbed the “PEPSI/LBT Exoplanet Transit Survey (PETS)”. Immediate goal is characterizing the planetary and stellar atmospheres in detail. During this 3-hr Zoom meeting we shall emphasis again the strengths and uniqueness of this proposal and recall shortcomings. The survey requests 40 nights in total or 12.5 nights per semester for two years in order to observe a representative number of targets starting in semester 2021A.

The zoom link will be provided by e-mail.

Survey Announcement (.pdf)

On the Chemical Abundance of HR 8799 and the Planet c

Comparing chemical abundances of a planet and the host star reveals the origin and formation pathway of the planet. Stellar abundance is measured with high-resolution spectroscopy. Planet abundance, on the other hand, is usually inferred from low-resolution data. For directly imaged exoplanets, the data are available from a slew of high-contrast imaging/spectroscopy instruments. Here, we study the chemical abundance of HR 8799 and its planet c. We measure stellar abundance using LBT/PEPSI (R=120,000) and archival HARPS data: stellar [C/H], [O/H], and C/O are 0.11±0.12, 0.12±0.14, and 0.54+0.09-0.12, all consistent with solar values. We conduct atmospheric retrieval using newly obtained Subaru/CHARIS data together with archival Gemini/GPI and Keck/OSIRIS data. We model the planet spectrum with petitRADTRANS and conduct retrieval using PyMultiNest. Retrieved planetary abundance can vary by ∼0.5 dex, from sub-stellar to stellar C and O abundances. The variation depends on whether strong priors are chosen to ensure a reasonable planet mass. Moreover, comparison with previous works also reveals inconsistency in abundance measurements. We discuss potential issues that can cause the inconsistency, e.g., systematics in individual data sets and different assumptions in the physics and chemistry in retrieval. We conclude that no robust retrieval can be obtained unless the issues are fully resolved.


Read more: Wang et al. 2020, AJ, 160, 150

Benchmark stars, benchmark spectrographs: Detailed spectroscopic comparison of ESPRESSO, PEPSI, and HARPS data for Gaia benchmark stars

Gaia benchmark stars are selected to be calibration stars for different spectroscopic surveys. Very high-quality and homogeneous spectroscopic data for these stars are therefore required. We collected ultrahigh-resolution ESPRESSO spectra for 30 of the 34 Gaia benchmark stars and made them public. We quantify the consistency of the results that are obtained with different high-, and ultrahigh-resolution spectrographs. We also comprehensively studied the effect of using different spectral reduction products of ESPRESSO on the final spectroscopic results. We used ultrahigh- and high-resolution spectra obtained with the ESPRESSO, PEPSI, and HARPS spectrographs to measure spectral line characteristics (line depth; line width; and EW) and determined stellar parameters and abundances for a subset of 11 Gaia benchmark stars. The EW spectral line measurements based on the ESPRESSO, PEPSI, and HARPS spectra agree to within a few percent. However, we note that the lines appear deeper in the ESPRESSO spectra than in PEPSI and HARPS. The stellar parameters derived from each spectrograph by combining the several available spectra agree well overall. We conclude that the ESPRESSO, PEPSI, and HARPS spectrographs can deliver spectroscopic results that are sufficiently consistent for most of the science cases in stellar spectroscopy.


Read more: Adibekyan et al. 2020, A&A, in press

Probing the atmosphere of HD 189733b with the Na I and K I spectral lines

A team around Engin Keles (AIP) compared previously observed high resolution Na I and K I absorption in the atmosphere of HD189733b with synthetic transmission spectra modeled for a variety of temperature and abundance values. The comparison showed that the observed Na I-D-line widths are much larger than the modeled ones. The Na I-D-lines had to be broadened by velocities in the order of 10 km/s to match the observations if only rotational broadening is taken into account. The K I line profi le on the other hand showed only a few km/s broadening comparable with the synthetic line pro files. This hints that either the atmosphere of HD 189733b lacks a significant amount of K I or the alkali lines probe different atmospheric regions with diff erent temperature, which could explain the diff erences in the resolved absorption lines.


Transmission spectrum of the KI-line at 7699 A in [%] subtracted from unity. The black solid line shows the Gaussian fi t. The dashed lines show the expected planetary absorption from the synthetic spectra without any broadening (red) and with the best-matching broadening solution of 3.8 km/s (green). The residual spectrum is shifted for clarity. .

Read more: Keles et al., 2020, MNRAS, in press

Lithium in T Corona Borealis

T Coronae Borealis is a recurrent, symbiotic nova system currently in quiescence between its periodic ≍80 yr cycle of eruptions. Observations during inter-outburst epochs provide an opportunity to study properties of the accretion disk and the M red giant. Here we present new irradiated (blackbody veiling) models, incorporating modern molecular opacities and line lists, of spectra derived from high-resolution (22,000 ≲ R ≲ 120,000) optical echelle observations obtained at two epochs, one prior to and one post the 2015 rebrightening event at similar spectroscopic system phase. We find a lithium abundance in the secondary at both epochs to be comparable. The non-irradiated (classical) model atmospheres yield a lithium abundance, A(Li) = 1.3 ± 0.1. The irradiated model (veiled) atmospheres, which are likely a better representation of the system in which the white dwarf and accretion disk illuminate the red giant, give A(Li) = 2.4 ± 0.1.


PEPSI spectrum around 6708 Å Li I resonance doublet in T CrB (red line), which is highly broadened by macro-turbulence. The green line is  a classical 1D model atmosphere fit and the blue line an irradiated  model fit.

Read more: Woodward et al., 2020, AJ, 159, 231

No Consistent Atmospheric Absorption Detected for the Ultra-hot Jupiter WASP-189b

We observed a partial transit of the ultra-hot Jupiter WASP-189 b with PEPSI on the LBT. We detect a highly variable transit signal in multiple atomic transitions, including H-alpha, Fe I, and Mg I. The signal is not consistent with a transiting planetary atmosphere. We suggest instead that the in-transit signal is due to an inhomogeneous stellar surface. Our observations demonstrate the lack of a highly extended atmosphere in common optical atomic tracers. Although WASP-189 is very bright, atmospheric characterization of the planet will be difficult due to the small transit depth and apparently compact atmosphere.


Average transmission spectra (left panels) and time series absorption (right panels) for the lines with detected absorption signatures. Note that the Mg I transmission spectrum is the average of all three triplet members. The vertical gray lines in the time series panels mark the transit contact points. Although there is some significant in-transit variability, no signal consistent with an extended atmosphere around the planet is detect.


Read more: Wilson et al. 2020, Research Notes of the AAS, Volume 4, Issue 4, id.53

Earth as transiting planet

January 2019 featured a total lunar eclipse. The Moon dimmed by a factor of 20,000 (10.75 mag) during totality. Therefore, the light gathering capability of the 11.8 m Large Binocular Telescope (LBT) in Arizona was needed for the observations. Additionally, the high spectral resolution of the Potsdam Echelle Polarimetric and Spectroscopic Instrument (PEPSI) was necessary to separate the expected tiny spectral-line absorptions of the Earth’s atmosphere from the normal solar spectrum at unprecedented spectral resolution and in polarized light. The radial velocities trace a wavelength dependent Rossiter-McLaughlin effect of the Earth eclipsing the Sun as seen from the Tycho crater confirming earlier observations. No line polarization of any spectral-line feature is detected outside nor inside eclipse. This places an upper limit of ≈0.2% on the degree of line polarization during transmission through Earth’s atmosphere and magnetosphere.

Read more: Strassmeier et al., 2020, A&A, 635, 156

AIP press release
LBT press release

Monitoring solar activity with PEPSI

Synoptic Sun-as-a-star observations are carried out with the Potsdam Echelle Polarimetric and Spectroscopic Instrument (PEPSI), which receives light from the Solar Disk-Integration (SDI) telescope. Daily spectra are produced with a high signal-to-noise ratio, providing access to unprecedented quasi-continuous, long-term, disk-integrated spectra of the Sun with high spectral and temporal resolution. We developed tools to monitor and study solar activity on different time-scales ranging from daily changes, over periods related to solar rotation, to annual and decadal trends. Strong chromospheric absorption lines, such as the Ca ii H & K λ3934 & 3968 Å lines, are powerful diagnostic tools for solar activity studies, since they trace the variations of the solar magnetic field. Other lines, such as Hα λ6563 Å line and the near-infrared (NIR) Ca ii λ8542 Å line, provide additional information on the physical properties in this highly complex and dynamic atmospheric layer. Currently, we work on a data pipeline for extraction, calibration, and analysis of the PEPSI/SDI data. We compare the SDI data with daily spectra from the Integrated Sunlight Spectrometer (ISS), which is part of the Synoptic Long-Term Investigation of the Sun (SOLIS) facility operated by the U.S. National Solar Observatory (NSO). This facilitates cross-calibration and validation of the SDI data.

Read more: Dineva et al. 2020, Astronomy in Focus, IAU XXX General Assembly, Vienna, Austria. Proceedings of the IAU, 2020, pp. 351-353

Download PDF.

A Magnetic Morphology Shift in Old Solar-type Stars

LBT/PEPSI spectropolarimetry tested the hypothesis that the surface magnetic morphology is a crucial component for the spin down of stars. Solar-type stars are born with relatively rapid rotation and strong magnetic fields. Through a process known as magnetic braking, the rotation slows over time as stellar winds gradually remove angular momentum from the system. The rate of angular momentum loss depends sensitively on the magnetic morphology, with the dipole field exerting the largest torque on the star. One hypothesis to explain this reduction in efficiency is a shift in magnetic morphology from predominantly larger to smaller spatial scales. We tested this hypothesis with spectropolarimetric measurements of two stars that sample chromospheric activity levels on opposite sides of the proposed magnetic transition. As predicted, the more active star (HD 100180) exhibits a significant circular polarization signature due to a non-axisymmetric large-scale magnetic field, while the less active star (HD 143761) shows no significant signal.

LSD reconstruction of the Stokes V profiles for HD 100180 (right) and HD 143761 (left).

Read more: Metcalfe et al. 2019, ApJ Letters, 887, 38