Fig. 3: Electrical tuning of exciton valley dynamics.

a Gate-dependent transient differential reflectance spectra for excitons with ΔE of 7.8 meV for Device #3 at three representative voltages: −15 V, −10 V, and 1 V, corresponding to the cases of positively charged, charge neutrality, and negatively charged, respectively. The black and red dots (solid lines) represent the measured data (exponential fit) of the Co and Cross (Cr) circularly polarized signal. The signals are normalized with respect to the intensity of the Co-polarized signal. The blue dots depict the dynamics of extracted DVP. b The gate-dependence of fitted intravalley decay time \(1/{\Gamma }_{{{{{{\rm{X}}}}}}}^{{{{{{\rm{ia}}}}}}}\) (black dots), and intervalley decay times \(1/{\Gamma }_{{{{{{\rm{X}}}}}}}^{{{{{{\rm{ir}}}}}}}\) (blue dots). The solid lines are polynomial fits. Error bars represent the uncertainty of lifetimes by exponential fitting. The decay variation of DVP at NCR is very small, leading to a large uncertainty of the exponential fitting of polarization decay. c The colored contour of DVP mapping in the plane of gate voltages and time delays, and the calculated DVP (magenta dots) for excitons using measured decay time in b by Eq. (2a) assuming P₀ of 100%.