Chemical Composition of Bright Stars in the Continuous Viewing Zone of the TESS Space Mission

Having our observational possibilities and methods of analysis developed in mind, we decided to observe high-resolution spectra of all bright (V < 8 mag) F5 and cooler stars (T_eff < 6500 K) in the whole TESS northern CVZ, which is a region around the northern ecliptic pole with a diameter of around 24°.

We examined a color–magnitude diagram of all V  <  8 mag and (B − V) > 0.39 mag stars in the selected field (Figure 1). In M18, we showed that T_eff < 6500 K corresponds to approximately (B − V) > 0.39 mag. In this way, we found 302 stars in the selected field that met these criteria (see Figure 2) and we have observed all of them during the period of 2018–2019. Of the observed stars, 53 fall within the 5° radius CVZ of the upcoming James Webb Space Telescope (JWST) mission (Gardner et al. 2006).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

We used the 1.65 m telescope at the Moletai Astronomical Observatory of Vilnius University in Lithuania that is equipped with the high-resolution Vilnius University Echelle Spectrograph (VUES; Jurgenson et al. 2016). This spectrograph has a wavelength coverage from 400 to 900 nm in R ∼ 36,000, ∼51,000, and ∼68,000 resolution modes. For our work, we used the ∼68,000 mode for the M spectral type stars and the ∼36,000 mode for other objects. Exposure times varied between 900 and 2400 s and S/Ns varied between 75 and 200 with the median value at 96, depending on stellar magnitudes. The VUES data reduction was accomplished on the site using an automated pipeline described by Jurgenson et al. (2016).

For an initial spectral analysis, we used the standard cross-correlation function (CCF) method to obtain spectroscopic radial velocity values. The CCF revealed 11 double-line binaries and 10 fast-rotating stars, which we postponed for a further analysis. Figure 3 shows the CCF examples. It was not possible to measure equivalent widths of lines for those fast-rotating stars with a satisfactory quality because of the broad and blended lines. The classical equivalent width method that we used to determine the main stellar atmospheric parameters could not be applied to the fast-rotating stars with strongly broadened and diminished lines. From the subsequent analysis, we also excluded the four coolest stars (see Section 2.3) with severe line-blending. In this way, 277 stars filled our final sample of the present analysis.

Zoom In Zoom Out Reset image size

In order to determine the main stellar atmospheric parameters (effective temperature, T_eff; surface gravity, log g; microturbulence velocity, v_t; and metallicity, $\langle [\mathrm{Fe}/{\rm{H}}]\rangle$), we adopted a classical method of equivalent widths of atomic neutral and ionized iron lines. We used a combination of the DAOSPEC (Stetson & Pancino 2008) and MOOG (Sneden 1973) codes the same way as the Vilnius node was using in the Gaia-European Southern Observatory (ESO) Survey (see Smiljanic et al. 2014; M18).

Detailed abundances of 24 chemical species were determined applying a spectral synthesis method with the TURBOSPECTRUM code (Alvarez & Plez 1998). The spectral analysis was done using a grid of Model Atmospheres in Radiative and Convective Scheme stellar atmosphere models (Gustafsson et al. 2008) and the Solar abundances by Grevesse et al. (2007). Atomic lines were selected from the Gaia-ESO line-list by Heiter et al. (2015). We have also used the molecular line lists: C₂ (Brooke et al. 2013; Ram et al. 2014); CN (Sneden et al. 2014); CH (Masseron et al. 2014); SiH (Kurucz 1993),; FeH (Dulick et al. 2003); CaH (B. Plez 2020, private communication); and OH, MgH, and NH (T. Masseron 2020, private communication). For the carbon abundance determination, we used two regions: the C₂ Swan (1, 0) band head at 5135 Å and the C₂ Swan (0, 1) band head at 5635 Å. For the nitrogen abundance determination, we used ${}^{12}{{\rm{C}}}^{14}{\rm{N}}$ molecular lines in the region 7980–8005 Å. The oxygen abundance was determined from the forbidden [O i] line at 6300 Å. These elements require a more detailed analysis, as they are bound by the molecular equilibrium. First, we performed a couple of iterations until the determinations of carbon and oxygen abundances converged. After that, we used both carbon and oxygen values to determine the abundance of nitrogen.

For more details of the chemical composition analysis, we refer to our recent studies (M19; Stonkutė et al. 2020).

The applied method did not work for four M3 and cooler stars: V* HW Dra, V* GP Dra, and V* TX Dra are M3–M5 type stars with Gaia DR2 T_eff ≈ 3300 K; a specific analysis is also needed for V* UX Dra, which is classified as a C-N5 carbon star with T_eff ≈ 2817 K (Lambert et al. 1986).

To calculate the stellar ages and their errors, we used the code UniDAM (the unified tool to estimate distances, ages, and masses) by Mints & Hekker (2017). The code uses a Bayesian approach and the PAdova and TRieste Stellar Evolution Code isochrones (Bressan et al. 2012). As an input, we used the stellar atmospheric parameters determined in this work together with the J, H, and K magnitudes from the Two Micron All-Sky Survey (2MASS) survey (Skrutskie et al. 2006) and the W1 and W2 magnitudes from AllWISE (Cutri et al. 2013).

Crossmatching of our sample stars with the 2MASS and All Wide-field Infrared Survey Explorer (AllWISE) program catalogs provided us data for 273 stars that had entries in at least one infrared photometry study: 198 stars had magnitudes from both, and 75 had magnitudes just from 2MASS. After calculating the ages, we discarded 59 stars that had flags reported after calculation that meant either unreliable photometry, the result being off the model grid, or just an unreliable determination (see Section 6.1 in Mints & Hekker 2017 for more interpretations). Finally, we were left with 214 stars for which we report the derived ages in this work.

The main kinematic parameters (R_mean, z_max, e, U, V, and W) for the stars were calculated using the python-based package for galactic-dynamics calculations galpy³ by Bovy (2015). The parallaxes, proper motions, and coordinates required for galpy were taken from the Gaia data release 2 (DR2) catalog (Luri et al. 2018; Katz et al. 2019).

The galpy was set to integrate orbits for 5 Gyr. Observational errors were estimated using 1000 Monte Carlo calculations according to the errors in the input parameters. The position and movement of the Sun are those from Bovy et al. (2012; ${R}_{\mathrm{gc}\odot }=8\,\mathrm{kpc}$ and V_⊙ = 220 km s⁻¹), the distance from the Galactic plane z_⊙ = 0.02 kpc (Joshi 2007), and the local standard of rest (LSR) from Schönrich et al. (2010) (U, V, W = 11.1, 12.24, 7.25 km s⁻¹).

The errors on the atmospheric parameters were estimated as follows:

• 1.  
  We tested the sensitivity of computed atmospheric parameters to the quality of the spectra. As a representative star for this test, we choose TYC 3910-1710-1—a giant star with high signal-to-noise spectra. We artificially degraded the spectra of this star to the S/Ns of 25, 50, and 75 per pixel and determined its atmospheric parameters. Of the randomly degraded spectra for each S/N, 100 were used and the generated statistics are presented in Table 1. The same test for dwarf stars was shown in M18.
• 2.  
  The uncertainties in the atomic parameters of the used lines create a scatter of measured iron abundances and also an error in a linear regression fit that can be directly propagated to follow the uncertainties of atmospheric parameters. Therefore, the uncertainties for each of the main atmospheric parameters are provided for every star in machine-readable Table A1 and they are computed the same way as published in the description of the Gaia-ESO Vilnius node by Smiljanic et al. (2014). The median errors measured by the algorithm in the full stellar sample are ${\sigma }_{{T}_{\mathrm{eff}}}$ = 60 K, σ_{log g} = 0.21 dex, ${\sigma }_{[\mathrm{Fe}/{\rm{H}}]}$ = 0.11 dex, and ${\sigma }_{{v}_{{\rm{t}}}}$ = 0.23 $\mathrm{km}\,{{\rm{s}}}^{-1}$.

We have made several tests to understand the error budget in our chemical abundance measurements:

• 1.  
  We used the same 300 generated spectra of the giant star TYC 3910-1710-1 for three S/N values to measure abundances in order to estimate their sensitivity to quality of the spectrum. These results are provided in Table 1. The same test for dwarf stars was accomplished in M19.
• 2.  
  Evaluation of the line-to-line scatter is a way to estimate random errors if the number of lines is large enough. A median of the standard deviation for a given element, ${\sigma }_{\mathrm{scat}}^{* }$, is presented in the sixth column of Table 2.
• 3.  
  The uncertainties of the main atmospheric parameters were propagated into the errors of chemical abundances. The median errors of this type over the stellar sample are provided in Table 2.

The final error for every element for every star that is given in machine-readable Table A1 is a quadratic sum of effects due to uncertainty in four atmospheric parameters and the abundance scatter given by the lines.

Table 1.  Errors Due to the Uncertain Continuum Placement and Equivalent width Measurement, Based on the Monte Carlo Simulations

	S/N = 25	S/N = 50	S/N = 75
TYC 3910-1710-1
Teff = 4458 K, log g = 2.72, $[\mathrm{Fe}/{\rm{H}}]=-0.14$
${\sigma }_{{T}_{\mathrm{eff}}}$	49	45	32
σlog g	0.09	0.09	0.09
${\sigma }_{[\mathrm{Fe}/{\rm{H}}]}$	0.03	0.03	0.01
${\sigma }_{{v}_{{\rm{t}}}}$	0.08	0.08	0.04
Na i	0.08	0.08	0.05
Mg i	0.10	0.06	0.03
Al i	0.09	0.07	0.04
Si i	0.05	0.04	0.03
Si ii	0.08	0.08	0.05
Ca i	0.10	0.07	0.05
Ca ii	0.09	0.08	0.07
Sc i	0.12	0.08	0.06
Sc ii	0.11	0.11	0.06
Ti i	0.10	0.08	0.06
Ti ii	0.11	0.09	0.03
V i	0.09	0.08	0.06
Cr i	0.09	0.09	0.05
Cr ii	0.12	0.09	0.05
Mn i	0.10	0.08	0.06
Fe i	0.09	0.06	0.04
Fe ii	0.13	0.09	0.07
Co i	0.09	0.07	0.01
Ni i	0.06	0.05	0.02
Cu i	0.07	0.06	0.06
Zn i	0.12	0.09	0.06

Download table as:  ASCIITypeset image

Table 2.  Median Effects on the Derived Abundances Resulting from the Atmospheric Parameter Uncertainties for the Sample Stars

El	ΔTeff	${\rm{\Delta }}\mathrm{log}g$	${\rm{\Delta }}[\mathrm{Fe}/{\rm{H}}]$	${\rm{\Delta }}{v}_{{\rm{t}}}$	${\sigma }_{\mathrm{scat}}$ a	Nmaxb	${\sigma }_{\mathrm{total}\left[\tfrac{\mathrm{El}}{{\rm{H}}}\right]}$ c	${\sigma }_{\mathrm{all}\left[\tfrac{\mathrm{El}}{{\rm{H}}}\right]}$ d
	(K)			(km s−1)
C(C2)	0.00	0.03	0.01	0.01	0.02	2	0.03	0.04
N(CN)	0.06	0.05	0.01	0.01	0.05	7	0.08	0.08
O([O i])	0.01	0.08	0.07	0.00	0.06	1	0.11	0.13
Na i	0.00	0.03	0.02	0.02	0.03	4	0.05	0.05
Mg i	0.01	0.07	0.02	0.03	0.06	5	0.08	0.12
Al i	0.01	0.02	0.03	0.03	0.04	5	0.03	0.07
Si i	0.01	0.02	0.02	0.02	0.03	14	0.02	0.03
Si ii	0.01	0.05	0.01	0.02	0.07	7	0.06	0.10
Ca i	0.02	0.07	0.02	0.02	0.06	31	0.08	0.10
Ca ii	0.03	0.06	0.02	0.04	0.05	7	0.07	0.09
Sc i	0.04	0.03	0.02	0.03	0.08	7	0.05	0.11
Sc ii	0.02	0.08	0.02	0.04	0.03	12	0.10	0.10
Ti i	0.04	0.04	0.01	0.02	0.03	81	0.07	0.08
Ti ii	0.01	0.08	0.04	0.04	0.04	19	0.09	0.09
V i	0.03	0.01	0.02	0.04	0.05	8	0.05	0.07
Cr i	0.02	0.03	0.02	0.03	0.05	21	0.05	0.06
Cr ii	0.02	0.08	0.02	0.04	0.03	2	0.10	0.11
Mn i	0.03	0.03	0.02	0.04	0.05	14	0.05	0.07
Fe i	0.02	0.03	0.02	0.03	0.05	138	0.05	0.06
Fe ii	0.01	0.08	0.03	0.04	0.04	11	0.09	0.12
Co i	0.01	0.01	0.01	0.02	0.05	7	0.02	0.07
Ni i	0.01	0.02	0.01	0.04	0.03	30	0.03	0.07
Cu i	0.03	0.02	0.02	0.02	0.05	6	0.04	0.06
Zn i	0.01	0.02	0.02	0.02	0.09	3	0.04	0.11

Notes.

^aσ_scat stands for the median line-to-line scatter. ^bN_max presents the number of lines investigated. ^c ${\sigma }_{\mathrm{total}([\mathrm{El}/{\rm{H}}])}$ stands for the median of the quadratic sum of all four effects on [El/H]. ^d ${\sigma }_{\mathrm{all}([\mathrm{El}/{\rm{H}}])}$ is a median of the combined effect of ${\sigma }_{\mathrm{total}([\mathrm{El}/{\rm{H}}])}$ and the line-to-line scatter σ_scat.

Download table as:  ASCIITypeset image

In Section 3.6 of M19, we have discussed that it should be safe to use the classical local thermodynamic equilibrium (LTE) approach to compute elemental abundances in the metallicity regime of our sample stars. The NLTE effects for the majority of elemental abundances in our sample are negligible; the non-LTE (NLTE) corrections were adopted just for the potassium abundances as they were determined from the large 7698.9 Å line. For manganese and copper, we have accounted for a hyperfine splitting as described in M19.

Since abundances of C, N, and O are bound together by the molecular equilibrium in the stellar atmospheres, we investigated how an error in one of them typically influences the abundance determination of another. We determined that Δ[O/H] = 0.10 causes Δ[C/H] = 0.02 and Δ[N/H] = 0.04, and Δ[C/H] = 0.10 causes Δ[N/H] = −0.11 and Δ[O/H] = 0.02, while Δ[N/H] = 0.10 has no effect on either the carbon or the oxygen abundances.

Our sample of 277 slowly rotating stars have T_eff between 3700 and 6600 K with a peak at 4700 K, [Fe/H] are from −1.5 dex to 0.25 dex with a peak at −0.25 dex, and log g are from 0.8 to 3.5 with a peak at 2.7 for giants and from 3.6 to 4.5 with a peak at 4.3 for dwarfs (more detailed distributions are presented in Figure 4). A log g versus T_eff diagram of the stars colored according to their metallicity is presented in Figure 5.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

From the fe_h catalog of the Set of Identifications, Measurements and Bibliography for Astronomical Data (SIMBAD; Wenger et al. 2000) database and the PASTEL catalog (Soubiran et al. 2010), we found that there are only few studies that have derived stellar parameters for some stars of our sample. We have only 47 stars that were observed before with atmospheric parameters delivered using high-resolution spectrographs. This spectroscopic comparison sample is collected from McWilliam (1990), Hekker & Meléndez (2007), Luck & Heiter (2007), da Silva et al. (2011), Lee et al. (2011), Pompéia et al. (2011), Britavskiy et al. (2012), Afşar et al. (2012), Roederer et al. (2014), Liu et al. (2014), Adamczak & Lambert (2014), da Silva et al. (2015), Maldonado & Villaver (2016), Brewer et al. (2016), Jönsson et al. (2017), Luck (2017), Aguilera-Gómez et al. (2018), and Deka-Szymankiewicz et al. (2018). The effective temperature and surface gravity consistency between our sample and the spectroscopic comparison sample of 47 stars are quite good: $\langle {\rm{\Delta }}{T}_{\mathrm{eff}}\rangle =-11\pm 58$ K and $\langle {\rm{\Delta }}\mathrm{log}\,g\rangle =-0.02\pm 0.18$.

We found 18 stars in common with the near-infrared, large-scale, stellar spectroscopic survey Apache Point Observatory Galactic Evolution Experiment's (APOGEE) 16th data release (DR16) (Ahumada et al. 2019). For the comparison, we were using calibrated parameters and abundances determined with the APOGEE Stellar Parameters and Chemical Abundance Pipeline (García Pérez et al. 2016). The biases for the main stellar atmospheric parameters from our sample are $\langle {\rm{\Delta }}{T}_{\mathrm{eff}}\rangle =-48\pm 47$ K and $\langle {\rm{\Delta }}\mathrm{log}\,g\rangle =0.06\pm 0.18$ dex. In Figure 6, we show a comparison of [Element/H] abundances for up to 18 stars in common with the APOGEE DR16. The average differences for all stars and standard deviations are calculated as our values minus the comparison values. The sample of common stars is not large; however, as one can see, the agreement for the majority of elements is very good. Regarding the C, N, and O elements, the APOGEE survey uses the infrared lines of CH, CN, and OH molecules, respectively, and the agreement for the carbon and oxygen abundances is quite good. The differences of 0.2 dex are between nitrogen and also of manganese abundances. As was shown in Smith et al. (2013, see their Figure 4), the larger Mn abundances in the APOGEE survey could be caused by a blending of the CN bands. The nitrogen abundance values are accumulating uncertainties of carbon and oxygen abundance determinations apart of others, thus they can be larger than for other chemical elements.

Zoom In Zoom Out Reset image size

We find a similar T_eff bias if we compare our sample with the Gaia DR2 data. $\langle {\rm{\Delta }}{T}_{\mathrm{eff}}\rangle =-62\pm 99$ K for all 277 stars. The APOGEE and Gaia DR2 T_eff bias is similar probably because the Gaia DR2 T_eff data was taken from the Astrophysical Parameters Inference System (Apsis; Andrae et al. 2018), which was trained on a number of samples where the APOGEE giant star input was one of the dominant ones.

In Table A1, columns 29–76 contain relative to Solar abundances and uncertainties of C(C₂), N(CN), [O i], Na i, Mg i, Al i, Si i, Si ii, Ca i, Ca ii, Sc i, Sc ii, Ti i, Ti ii, V i, Cr i, Cr ii, Mn i, Fe i, Fe ii, Co i, Ni i, Cu i, and Zn i for the 277 stars investigated in the present study. The abundances are presented in [Element/H] form. The [Fe i/H] and [Fe ii/H] values occupy columns 77–80 of Table A1. All elemental abundance ratios with respect to [Fe i/H] are show in Figures 7 and 8 where the stars are colored according to their attribution to the Galactic subcomponents. The stellar attribution to subcomponents is presented in column 83 of Table A1 and is described in the next section.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The majority of the stars have radial velocities between −40 and +20 km s⁻¹. We have compared our radial velocities with the Gaia DR2 catalog data. The mean and standard deviation of differences between the two sets is $\langle {\rm{\Delta }}{V}_{\mathrm{rad}}\rangle \,=0.05\pm 0.53$ km s⁻¹. Overall radial velocities and corresponding errors are presented in columns 11–12 of Table A1.

The ages of our sample stars are from about 1.0 to 11 Gyr; the majority are close to Solar, about 4 Gyr. The age values and uncertainties are presented in columns 13–14 of Table A1.

The U, V, and W velocities; distances; R_mean; z_max; and orbital eccentricities, e, with corresponding errors are presented in columns 15–27 of Table A1 and are exhibited in Figure 9.

Zoom In Zoom Out Reset image size

It is widely accepted that Galactic subcomponents like thin and thick disks differ in a number of parameters. There are two widely used methods to separate them: kinematical (e.g., Bensby et al. 2003, 2005, 2014) and chemical (e.g., Adibekyan et al. 2012; Recio-Blanco et al. 2014).

The method introduced by Bensby et al. (2003, 2014) employs the thick-to-thin disk probability ratios. Stars with TD/D > 2 are potential thick-disk stars, stars with TD/D <0.5 potentially belong to the thin disk, and stars with 0.5 < TD/D < 2.0 are called "in-between stars." Column 29 in Table A1 presents the thick-to-thin disk probability ratios (TD/D). The panel (d) in Figure 9 displays the stellar TD/D distribution with age. According to the Toomre diagram (panel (d) in Figure 9), almost all our stars kinematically belong to the thin or thick disks, except HD 175305, which, according to Nissen & Schuster (2010), has a clear halo kinematics (${\left({U}_{\mathrm{LSR}}^{2}+{{\boldsymbol{W}}}_{\mathrm{LSR}}^{2}\right)}^{1/2}$ = 290 km s⁻¹, V_LSR = −72 km s⁻¹, R_mean = 16 kpc, and z_max = 20.6 kpc). Thus, using the kinematical method, we found that our sample consists of 262 thin-disk stars, nine thick-disk stars, and the remaining five stars are "in-between stars" and one halo star.

The chemical separation method can employ [Mg i/Fe i] (Adibekyan et al. 2012; Mikolaitis et al. 2014), [Ti i/Fe i] (Bensby et al. 2014), or [α/Fe i] (Recio-Blanco et al. 2014) abundance ratios. As in our previous studies, we used [Mg i/Fe i] versus [Fe i/H] to separate the low-α from high-α stars that potentially belong to the thin or thick disks, respectively. The most metal-poor star HD 175305 with halo kinematics should belong to the high-α halo population according to criteria by Nissen & Schuster (2010). Thus from chemical signatures, we revealed in our sample that there are 219 thin-disk, 57 thick-disk, and one high-α halo stars (column 83 in Table A1 presents information to which of the Galactic subcomponents the stars were attributed).

In Figure 10, we show [α/Fe i] is an average of Mg i, Si i, Si ii, Ca i, Ca ii, Ti i, and Ti ii (the [α/Fe i] values and standard errors of the mean are presented in columns 81 and 82 of Table A1). The two disks are separated quite well. Exoplanets found near thin- and thick-disk stars will be of different chemical content.

Zoom In Zoom Out Reset image size

The Mg/Si number ratio plays an important role in determining the plate tectonics and habitability of extra-solar planets. Theoretical studies have shown that if the Mg/Si number ratio is less than one, the extra-solar planet composition will be mostly made of pyroxene with other silicate based minerals, such as feldspar and a fraction of olivine (Bond et al. 2010). Furthermore, if the Mg/Si number ratio is between 1.0 and 2.0, as in our own planet Earth where Mg/Si_⊙ = 1.05, the extra-solar planet composition will be formed by pyroxene and olivine evenly. And lastly, if the Mg/Si number ratio is more than two, most of the planets will form from material with an olivine structure and the remaining magnesium will form oxides (Bond et al. 2010; Thiabaud et al. 2015). Thus, the planets originating from these three divisions would be different in term of plate tectonics, surface mineral chemistry, and inner geology.

Furthermore, Dorn et al. (2015) in their work concluded that in order to constrain the models for interior structure of rocky planets, the stellar chemical abundances, such as iron, silicon, and magnesium, are the key ingredients to reduce degeneracy in interior structure models and to constrain the mantle composition.

As suggested by, e.g., Bond et al. (2010) and Suárez-Andrés et al. (2018), the number ratio of Mg/Si = 1.0 marks a division between stars that could form potentially gaseous or rocky planets. About 83% of our stars have Mg/Si values in the range between 1.0 and 2.0 with the mean value of Mg/Si = 1.18 ± 0.13, which could suggest that they may have planets with a composition close to that of our planet Earth. The remaining ∼17% of the stars have Mg/Si ≤ 1.0, where planets may have a magnesium-depleted mineralogy. If we add to the sample of 277 stars investigated in this work and 249 stars investigated by M19 using the same instrument and method of analysis, the percentage of stars in the interval of $1.0\lt \mathrm{Mg}/\mathrm{Si}\lt 2.0$ increases by 1.4%. Suárez-Andrés et al. (2018) also found that ∼83% of their sample of 499 Solar-like stars to have a Mg/Si number ratio between 1.0 and 2.0, while Brewer & Fischer (2016) estimated a somewhat lower percentage among 847 investigated stars: ∼64%. The percentage of stars with the particular Mg/Si ratio may vary depending on how many stars belonging to the different Galactic subcomponents are in a sample. In Figure 11, we show a distribution of thin- and thick-disk stars according to their Mg/Si number ratios for the sample of 526 stars from our and M19 studies.

Zoom In Zoom Out Reset image size

In Figure 12, we also plotted the stars in the C/O versus Mg/Si number ratio diagrams. The left panel shows where stars are located according their dependence to the Galactic subcomponents, while in the right one, we can see how the C/O ratio differs in dwarfs and giants. Together with the data from this work, we plot 249 stars previously investigated with the same instrument and method of analysis (M19; Stonkutė et al. 2020). It can be seen that in giants, due to evolutionary changes of carbon, the C/O ratio, on average, is lower by about 0.1, thus we have to keep in mind that planets that we find around evolved stars were formed when their host stars were young and had larger C/O ratios.

Zoom In Zoom Out Reset image size