Monitored chaotic scattering

The gist The authors extend random-matrix theory to quantum dots whose dynamics are monitored by time-resolved measurements by constructing an ensemble of Kraus operators from circular-ensemble scattering matrices and fixed measurement protocols, which allows for a comparison with closed-form random-matrix predictions.

Construction of the Kraus Operator Ensemble

The core idea is to move from the scattering matrix to a quantum channel represented by Kraus operators, as the dynamics are monitored. The construction involves drawing an ensemble of Kraus operators induced by circular-ensemble scattering matrices and a fixed time-resolved measurement protocol, referred to as the “circular Kraus ensemble”. This construction is designed to capture the interplay between unitary evolution and projective measurements.

The process involves several steps:

1. Starting from a scattering matrix drawn from a circular ensemble, one constructs the corresponding ensemble of Kraus operators for the monitored evolution of the many-body density matrix.

2. In the single-particle sector, this leads to a discrete-time quantum master equation for the transferred charge.

3. The goal is to construct an analogue of the circular unitary ensemble for this monitored setting, though no invariance under arbitrary unitary rotations of the Kraus representation is implied.

Electrical Conduction and Transferred Charge

The study focuses on the transferred charge in response to a voltage bias, which is represented via a density matrix map. The total transferred charge T is calculated as the limit of the trace of the charge operator Qˆ times the density matrix, which follows from summing over all modes and times.

Key aspects of electrical conduction include:

(3.5) Eq. (3.5) gives the transferred charge T for V1 = V, V2 = 0.

Monitoring breaks reciprocity, leading to a difference between the case of an asymmetric bias and a symmetric bias. The differential conductance G = dI/dV is found to be proportional to the transferred charge T, i.e., G = (e2/h)T.

Quantum Master Equation and Numerical Solution

The single-particle sector of Fock space allows the operator map to be reduced to a vector-space recursion relation, or a “quantum master equation”. This equation can be solved efficiently using numerical linear algebra methods.

The single-particle density matrix evolves according to the map ρ(t + 1) = L[ρ(t)]. The master equation (4.5) is derived by summing over measurement outcomes algebraically, reducing the number of terms from 2N to N. The solution for the transmission probability Tn'n involves the operator (1 - L)−1 [ρn(0)].

Comparison with Random-Matrix Theory (RMT)

The numerical solution of the master equation can be accurately described by an RMT calculation based on a conjectured “equipartition rule”. This conjecture states that monitored electrons are distributed equally over the outgoing modes.

The comparison between the master equation and RMT calculations tests this equipartition rule, as the latter provides a closed-form solution that should describe what the master equation converges to in the limit of many monitored modes M → ∞. For single-mode monitoring, this comparison yields probability density functions P(T) that are in good agreement with the numerics of the master equation.

Reciprocity Breaking and Shot Noise

Monitoring introduces a breaking of reciprocity, as shown by comparing the transferred charge T and T. The mean squared difference ∆2β quantifies this breaking, which is maximal near γ = 2.

Furthermore, higher moments of the charge operator are encoded in the generating function F(ξ). For asymmetric bias, the mean and variance of the transferred charge are given by Q = NV T and Var Q = NV T (1 − T). This relates to the shot noise power Pnoise = (e3/h)T (1 − T).

Comparison with Voltage-Probe Model

In the multi-mode weak-measurement regime, the RMT description reduces the problem to Poisson-kernel averages, allowing a direct comparison with B¨uttiker’s voltage-probe model. The voltage probe transmission TVP is given by Eq. (7.5). Although both models are charge-conserving, monitoring exchanges energy with the environment, whereas the voltage probe does not. The difference appears because of the time scale of dephasing process: monitoring is a time-resolved projective measurement on a much shorter time scale δt < τdwell < τV < 1/eV.

Conclusion

The study successfully generalizes the chaotic scattering ensemble to include projective measurements, constructing a quantum master equation that can be solved efficiently and providing insight through RMT connections based on an equipartition rule conjecture. The construction leads to concrete consequences regarding the probability distribution of differential conductance for both broken and preserved timereversal symmetry. The monitoring-induced breaking of reciprocity distinguishes asymmetric and symmetric voltage biasing, while in the multi-mode weak-measurement regime, RMT reduces the problem to Poisson-kernel averages for comparison with the voltage-probe model. The natural mathematical problem left open is proving the equipartition rule as a limiting theorem for the circular ensemble.

How it works

The construction of the Kraus operator ensemble involves alternating unitary evolution with projective measurements, where measurements are described by POVMs. The Kraus operators Kˆs are labeled by the binary string s = (s1, s2,... sM). The evolution is given by ρˆ(t + 1) = X sum s Kˆsρˆ(t)Kˆ †s.

The single-particle sector reduces the dynamics to a vector-space recursion relation, or a “quantum master equation”. This equation can be solved efficiently because the sum over measurement outcomes is carried out algebraically.

The observable transferred charge T is calculated using the geometric series expansion of the density matrix evolution, leading to T = (1 - L)−1 [ρ(0)] where L is the recursion operator.

Key Findings

(5.6) P(T) = (4T if 0 < T < 1/2, 4(1 − T) if 1/2 < T < 1, in good agreement with the numerics of Fig. 3.

The RMT calculation based on the equipartition rule predicts specific probability density functions P(T) that are shown to be in good agreement with the master equation numerics.

(5.8) P(T) = (2 ln 2 if 0 < T < 1/2, −2 ln T if 1/2 < T < 1, in good agreement with the numerics of Fig. 4.

For the case where time-reversal symmetry is preserved, the RMT calculation yields probability density functions P(T) that are in good agreement with the master equation numerics.

**(6.

1. textbfReal-time Non-Equilibrium Transport Simulation: The improved AI can simulate quantum transport through chaotic dots under continuous, time-resolved projective measurements by solving a discrete-time quantum master equation for the density matrix, as described by Eq. (4.5). This allows the system to model monitored quantum dynamics which includes alternating coherent unitary evolution and projective measurements, capturing effects like current flow even in the absence of a voltage bias (a current may flow even in the absence of a voltage bias).

2. textbfModel Validation via RMT Equipartition Rule Testing: The improved AI can rigorously test the equipartition rule by comparing numerical solutions of the master equation (4.5) with closed-form Random Matrix Theory (RMT) predictions, as shown in Secs. V and VI. This capability allows for a direct assessment of whether the conjecture—monitored electrons are distributed equally over the outgoing modes—holds under different conditions, providing crucial insight into the limitations of RMT approximations.

3. textbfReciprocity Breaking Quantification: The improved AI can quantify the breaking of reciprocity in transport by calculating metrics like the mean squared difference between T and T' (Eq. (6.3)) for multi-mode monitoring, as well as deriving its dependence on measurement strength gamma." This enables the system to distinguish statistically between asymmetric and symmetric voltage biasing regimes, which is critical for accurately predicting conductance measurements in real devices.

4. textbfEnhanced Shot Noise Prediction: The improved AI can predict higher moments of the charge operator Qˆ to calculate shot noise power, including the variance for symmetric bias (Eq. (8.6)). This allows for the calculation of shot noise power Pnoise = (e 3/h)T(1-T) and its reduction due to reciprocity breaking, offering a more complete statistical picture than just the differential conductance.

5. textbfModel Comparison with Voltage-Probe Dephasing: The improved AI can compare the results of the monitored quantum transport model directly against established models like B¨uttiker’s voltage-probe model (Eq. (7.5)). This comparison highlights a key distinction: Monitoring is a time-resolved projective measurement, on a time scale δt = L/vF much less than the mean time τdwell, allowing for the development of more physically accurate transport descriptions in the low-voltage, linear response regime.