Here is a figure show us the 2D distribution between age and metallicity of stars in 3 TNG100 galaxies as well as the marginalized distributions. I get age of stars by transforming GFM_StellarFormationTime to age by astropy. And metallicity is calculated by GFM_Metallicity/0.0127 as indicated in data specification.
I wonder is it consistent with observation or our understanding? These very old stars are belong to halo. But halo stars should have smaller metallicity. These young stars are belong to disk. But disk stars should have higher metallicity. In addition, there are some stars have extremely high metallicity (about 3~5 Z_sun).
I remember that chemistry of TNG has some problems. I don't know if this trend is because of these problems or what I think is wrong.
What kinds of galaxies are these? E.g., are they Milky Way-like galaxies, or dwarfs, or more massive systems?
I would say the plots you include look reasonable to me. In general, older stars have lower metallicities? I would suggest to plot log(Z/Z_solar) instead Z/Z_solar and to change the axis range to 1e-2 to 2.0 for example, the primary range of interest, unless you are looking at very metal-poor stars and want to go to lower metallicities.
You can find the stellar age vs M_star, and stellar metallicity vs M_star, relations, at redshift zero, in comparison to data, in Nelson+2018 - see Figure 2.
Thank you for your suggestions. These galaxies are massive galaxies (M_*>1e11 M_sun)