Hidden valley dynamics behind vanishing circular polarization in moiré excitons

Technical Summary: Hidden Valley Dynamics Behind Vanishing Circular Polarization in Moiré Excitons

Problem Statement
Transition-metal dichalcogenide (TMD) heterostructures, particularly those forming moiré superlattices, offer a platform for valleytronics where valley degrees of freedom can be optically addressed. In these systems, interlayer excitons inherit valley-contrasting optical selection rules while gaining long lifetimes and electric dipoles. A common experimental metric for valley polarization is the degree of circular polarization (DCP) measured in steady-state photoluminescence (PL). Conventionally, a vanishing time-integrated DCP is interpreted as evidence of rapid valley depolarization or intervalley relaxation. However, this interpretation assumes a single, monotonically decaying valley population. The authors posit that in moiré systems, where different local atomic registries (A-like and B-like sites) can host exciton states with opposite optical selection rules, a zero net polarization could instead result from the temporal compensation of competing helicity-opposite components. The core problem addressed is the inability of steady-state measurements to distinguish between rapid intrinsic depolarization and the cancellation of distinct dynamical components, potentially leading to false-negative conclusions about valley polarization.

Methodology
The study utilizes a dual-gated WSe₂/WS₂ heterobilayer device encapsulated in hexagonal boron nitride (hBN) with graphite gates. The device is characterized via second harmonic generation (SHG) to confirm a near-0° (R-type) stacking configuration with a twist angle of $\theta \approx 0.14^\circ$. The primary experimental technique involves helicity-resolved time-resolved photoluminescence (TRPL) at 4 K.

• Steady-State Analysis: The authors measure the energy-integrated degree of circular polarization ($P_v$) across various electrostatic gate fields ($V_t, V_b$) to identify regimes where $P_v$ vanishes.
• Time-Domain Dynamics: In regimes where $P_v \approx 0$, they resolve the temporal evolution of co- and cross-circularly polarized emission components ($I_{\sigma+}$ and $I_{\sigma-}$) relative to the excitation helicity.
• Modeling: A minimal two-channel phenomenological model is developed. This model treats the emission as originating from two distinct moiré channels (A-like and B-like) with opposite optical selection rules and distinct effective decay/depolarization rates ($\Gamma_i, \gamma_i$). The model does not assume microscopic hopping between sites but rather pre-established channels with different relaxation kinetics.

Key Results

1. Temporal Crossing in Vanishing Polarization Regimes: In gate-field conditions where the time-integrated circular polarization is nearly zero, the time-resolved data reveals a clear "crossing" behavior. The cross-circularly polarized component dominates at early delay times, while the co-circularly polarized component dominates at later times. This results in a sign reversal of the helicity-difference signal $D(t) = I_{\sigma+}(t) - I_{\sigma-}(t)$.
2. Failure of Single-Channel Models: The observed sign reversal cannot be reproduced by a single-exponential depolarization model ($P(t) \propto e^{-t/\tau_v}$), which would predict a monotonic decay of polarization without sign change. This confirms that the zero steady-state signal is not due to a featureless, rapidly depolarized state.
3. Two-Channel Model Validation: The minimal two-channel model successfully reproduces the helicity crossing. The crossing time ($t_{cross}$) is derived as a function of the relative initial amplitudes ($D_A, D_B$) and the difference in effective valley relaxation rates ($\gamma_A, \gamma_B$) of the two channels. The model demonstrates that the zero net polarization arises from the integration of negative and positive contributions over the emission decay.
4. Electrostatic Tunability: The crossing time $t_{cross}$ and the fitted amplitude magnitudes ($|D_A|, |D_B|$) evolve systematically with the applied gate field. This indicates that the relative weights and effective relaxation rates of the competing helicity channels are electrically controllable. The regime of vanishing steady-state polarization corresponds to a specific balance point in these gate-tunable dynamics, not a loss of valley physics.

Significance and Claims
The paper claims to establish that time-integrated circular polarization can yield a false-negative indication of valley polarization in multichannel valley emitters like moiré excitons. The primary significance lies in demonstrating that a vanishing steady-state DCP does not necessarily imply rapid valley relaxation; rather, it can mask well-defined, competing time-domain dynamics.

By introducing helicity-resolved time-resolved measurements, the authors provide a sensitive probe for "hidden" valley dynamics that are inaccessible via steady-state analysis alone. The work suggests that in complex moiré landscapes, the apparent loss of polarization is often a result of the compensation between A-like and B-like emission channels with opposite selection rules and distinct kinetics. Consequently, the study redefines the interpretation of zero-polarization regimes, positioning them as regions of active, tunable competition between helicity channels rather than depolarized states. This insight is crucial for accurately assessing the viability of moiré excitons for valleytronic and quantum-optical applications where robust valley control is required.