Fig. 3

Sub-gap states in the partly screened regime. a Stability diagram showing linear N-DQD-S conductance vs. plunger gates at T = 30 mK. b DQD-S zero-bandwidth model reproducing the experimental behavior in a for intermediate coupling t_S to S. Orange regions marking ground state transitions have singlet-doublet splitting less than 0.015 meV. c–e Bias spectroscopy of sub-gap states vs. QD_N occupation, sweeping \(V_{{\mathrm{g}}_{\mathrm{N}}}\) as indicated by red arrows in a. f–h Sub-gap spectroscopy vs. QD_S occupation sweeping along green arrows in a. All plots show clear sub-gap resonances consistent with the ground states (doublet \({\cal D}\) or singlet \({\cal S}\)) for different gate voltages indicated below. i–n Zero-bandwidth model calculation of sub-gap excitations corresponding to experimental plots c–h. Parameters used in b, i–n are (meV): U_N = 2.5, U_S = 0.8, U_D = 0.1, t_d = 0.27, t_S = 0.22, and Δ = 0.14. Additional data related to the partly screened regime and how the above parameters are extracted from experimental data can be found in the Supplementary Note 2, Device A