Fig. 3

Valley polarization as a function of on energy and pump fluence. Valley polarization \(P_v^n\) = \(\left( {N_n^K - N_n^{K{\prime}}} \right){\mathrm{/}}\left( {N_n^K + N_n^{K{\prime}}} \right)\) as a function of excitation energy in WS₂ for a A (n = A) and b B excitons (n = B). Experimental values extracted from pump-probe measurements are shown for both cases as dots. For a better comparison of the qualitative trends, we have normalized the experimental and theoretical values to the maximum measured in experiment and calculated in theory. In the experiment, the A exciton contribution has been fitted by a Voigt function and subtracted from the spectra prior to integration. The limited spectral bandwidth of the measurement leads to uncertainties and determines the width of the error bars of the extracted absolute valley polarization degrees. Due to the Dexter coupling, the A and B valley polarizations are drastically reduced and change sign at excitation energies close to the resonance of the B_1s and A_1s exciton, respectively. c, d Pump fluence dependence of Dexter coupling: Here we show the enhanced excitation of the B exciton as the pump fluence of the excitation pulse (resonant to the A exciton) increases. Due to the larger excitation density we observe in both experiment and in theory an increase of the excitation of B excitons in both valleys. The purple line shows the off-resonantly excited B excitons in the pumped valley, the blue line that B excitons induced via Dexter coupling in the other valley. The latter becomes more efficient for stronger excitation. As a result, we find a higher increase of the population of B excitons in the unpumped valley than in the pumped valley. Note that this is in full analogy to the findings of ref. ¹⁹, where an increased signal of the B excitons was observed for larger pump fluences