Fig. 2: 2D spectra with polarization control reveal interactions between specific polaron states.

Illustrative 2D spectra identify the expected peak locations arising from interactions between specific pairs of attractive polarons in the co-circular (a) and cross-circular (b) polarization configurations. The detailed pathways that identify the interactions underlying these peaks are described in Supplementary Note 4. The yellow shaded ellipses indicate where peaks are observed in the experiments, and thus which pairs of polarons are interacting. Experimental 1Q rephasing 2D spectra in the vicinity of the attractive polaron peaks for co- (c) and cross-circular (d) polarization schemes, with t₂ = 0. Color scaling represents the normalized absolute signal amplitude (Norm. Amp.). Note: the cross-circular polarized scheme involves two indistinguishable pulse sequences: σ⁺σ⁻σ⁺σ⁻ and σ⁺σ⁺σ⁻σ⁻. Inhomogeneous broadening causes the on-diagonal peaks to merge, while the cross-peaks can be distinguished due to their narrower homogeneous linewidth. Only diagonal peaks are seen for co-circular polarization, and only cross-peaks for cross-circular polarization. There are additional cross-peaks in d at lower ℏω₃ values, which are attributed to pathways involving bipolarons. Simulated 2D spectra for the co- (e) and cross-circular (f) polarization schemes. The simulations reproduce the selection rules for the peaks, confirming that interactions between polarons dressed by the same Fermi sea are dominant and driven by phase-space filling effects.