We present analysis of eclipsing binary star LL Aqr which contains a solar twin component.
A solar twin in the eclipsing binary LL Aquarii
D. Graczyk, R. Smolec, K. Pavlovski, J. Southworth, G. Pietrzyński, P.F.L. Maxted, P. Konorski, W. Gieren, B. Pilecki, M. Taormina, K. Suchomska, P. Karczmarek, M. Górski, P. Wielgórski, R.I. Anderson
We present analysis of eclipsing binary star LL Aqr which contains a solar twin component. This is the first known star with properties of a solar twin existing in a non-interacting eclipsing binary, offering an excellent opportunity to fully characterise its physical properties with very high precision.
We used extensive light curve from the Super-WASP project augmented by multi-band, archival photometry to derive photometric solutions and colors of components. Using high-resolution and high-precision spectrographs, HARPS and CORALIE, we determined radial velocities and an orbital solution with the Wilson-Devinney code. Spectra of both components were decomposed and we performed a detailed atmospheric LTE abundance analysis. The atmospheric chemical composition and metallicity of LL Aqr is very close to solar one.
LL Aqr consists of two main-sequence stars (F9 V + G3 V) with masses of M1 = 1.1949 ± 0.0007 and M2 = 1.0337 ± 0.0007 M☉, radii R1 = 1.321 ± 0.006 and R2 = 1.002 ± 0.005 R☉, temperatures T1 = 6080 ± 45 and T2 = 5703 ± 50 K and solar chemical composition [M/H] = 0.02 ± 0.05 dex. The absolute dimensions, radiative and photometric properties, and atmospheric abundances of the secondary are all fully consistent with being a solar twin. The system is significantly younger that the Sun, with an age between 2.3 Gyr and 2.7 Gyr, which agrees well with the relatively high lithium abundance of the secondary, A(Li) = 1.65 ± 0.10 dex.
The resulting highly precise stellar parameters were used for a detailed comparison with PARSEC, MESA, and GARSTEC stellar evolution models. Both stars are cooler by about 3.5σ or less metal abundant by 5σ than predicted by standard sets of stellar evolution models. When advanced modelling was performed, we found that full agreement with observations can only be obtained for values of the mixing length and envelope overshooting parameters that are hard to accept. The most reasonable and physically justified model fits found with MESA and GARSTEC codes still have discrepancies with observations but only at the level of 1σ.