Something green beneath the surface: The dynamical nature of fossil structures in IllustrisTNG-100

Fossil structures (FSs) have traditionally been considered dynamically relaxed end products of group evolution, characterized by a large magnitude gap (Δ m_ 1,2 ≥ 2). However, recent observations and simulations suggested that this picture is incomplete, motivating a reassessment of their nature using cosmological simulations. We investigated the dynamical relaxation state of FSs, the differences in their galaxy populations relative to non-fossil systems (non-FSs), and the assembly history of the two populations. We focused on the system dynamics and the evolution of galaxies within them. Using IllustrisTNG-100, a state-of-the-art cosmological simulation, we selected 182 structures (M_ 200 > 10^ 13 ,M_⊙) at z = 0, classified them as FSs or non-FSs based on Δ 1,2 in the r band. We tracked the Δ m_ 1,2 evolution over 9,Gyr and analyzed (1) the emergence of Δ m_ 1,2 and its relation to the last significant accretion event onto the system, (2) the fraction of quenched galaxies (specific star formation rate $< 10^ -11 $,yr -1 ), (3) the distribution of galaxies in color--stellar mass space, and (4) the centroid shift in the gas--Brightest Structure Galaxy (BSG) as a dynamical proxy. The magnitude gap in FSs is primarily driven by the absence of massive recent accretion: FSs exhibit significantly lower BSG-to-satellite stellar mass ratios (μ^ FS _⋆=0.17 vs. μ^ NFS _⋆=0.39) for the most massive satellite accreted within the last 6,Gyr. FSs also host a more prominent red sequence and marginally higher quenched fractions than non-FSs. Furthermore, non-FSs show more prominent green-valley populations, which are associated with the secondary massive galaxy behind the small-magnitude gap. When we compared the gas--BSG centroid offsets, we found that both populations occupy an intermediate dynamical state ̊m BSG-CM ≈ 0.14--0.16 R/R_200) that is broadly consistent with unperturbed systems, but neither population has reached a relaxed state. Our findings indicate that while the magnitude gap effectively identifies systems that have ceased major mergers in the last 3-6 Gyr, it is a poor proxy for their current global dynamical state, as reflected in the weak anticorrelation between Δ m_ 1,2 and D_̊m BSG-CM. The FS and non-FS populations exhibit intermediate gas-BSG offsets (D_ BSG-CM ≈ 0.15 R/R_ 200 ) and fail to reach relaxation. This decoupling suggests that the magnitude gap traces the assembly history of massive components and not the overall stability of the intracluster medium.