Quantum Walk on a Line with Absorbing Boundaries

1. Problem Statement

The paper investigates the dynamics of a two-state discrete-time quantum walk (QW) on a finite one-dimensional line with absorbing boundaries (sinks) located at positions $-N$ and $N$.

• Context: While QWs are fundamental to quantum algorithms (search, universal computation) and transport phenomena, the specific problem of absorption probabilities in finite systems with two sinks has historically relied on implicit formulas involving complex integrals derived from recurrence relations (specifically the work of Konno et al., 2003).
• Gap: Previous results were not in "closed form" suitable for asymptotic analysis. The dependence of absorption probabilities on the initial coin state and the starting position $k$ in the limit of large system size ($N \to \infty$) was not fully explicit.
• Goal: To derive closed-form analytical formulas for the asymptotic absorption probabilities at the left and right boundaries, analyzing two specific regimes:
  1. The starting position $k$ is fixed while $N \to \infty$.
  2. The starting position $k$ is close to a boundary (distance $\delta$ is fixed while $N \to \infty$).

2. Methodology

The authors employ a combination of linear algebra, complex analysis, and advanced convergence theory:

• Model Definition:

  • Hilbert Space: $\mathcal{H} = \mathcal{H}_P \otimes \mathcal{H}_C$, where $\mathcal{H}_P$ is the position space ($-N$ to $N$) and $\mathcal{H}_C$ is the two-dimensional coin space ($|L\rangle, |R\rangle$).
  • Evolution: A single step consists of a coin flip $\hat{C}(\theta)$ followed by a conditional shift $\hat{S}$. The coin is a one-parameter matrix $\hat{C}(\theta)$, generalizing the Hadamard coin ($\theta = \pi/4$).
  • Absorption: Modeled via non-unitary evolution using projection operators $\hat{\Pi}_L$ and $\hat{\Pi}_R$ at the boundaries. The survival probability decays to zero as $t \to \infty$.

• Analytical Derivation:

  • Starting Point: The authors utilize the implicit integral formulas for absorption probabilities $P_L$ and $P_R$ derived by Konno et al. (2003), which involve integrals of functions $p(z)$ and $r(z)$ over the unit circle.
  • Simplification: They first simplify the integrands into trigonometric forms involving an auxiliary variable $\alpha$.
  • Two-Scale Convergence: The core mathematical innovation is the application of two-scale convergence analysis (Nguetseng-Allaire theory). This allows the evaluation of the limit $N \to \infty$ for oscillatory integrands where the frequency scales with $N$.
  • Eigenbasis Decomposition: To simplify the results, the initial coin state is decomposed into the eigenbasis of the coin operator ($|\theta_+\rangle, |\theta_-\rangle$). This reveals hidden symmetries and simplifies the dependence on the initial state parameters.

• Numerical Validation:

  • Extensive numerical simulations were performed for small system sizes ($N$) to verify the analytical asymptotic formulas and study the convergence rate.

3. Key Contributions and Results

A. Closed-Form Asymptotic Formulas

The paper derives explicit formulas for the absorption probability at the left boundary, $P_L$, in the limit $N \to \infty$.

1. Fixed Starting Position ($k$ is constant):

   • When the walker starts at a fixed distance from the center, the absorption probability depends only on the coin parameter $\theta$ and the polar angle $\rho$ of the initial coin state in the coin's eigenbasis.
   • Crucially, it becomes independent of the relative phase (azimuthal angle $\phi$) of the initial state and the specific starting position $k$ (as long as $k$ does not scale with $N$).
   • Formula:
     $$P_L = \frac{1}{2} + \frac{1 - \sin\theta}{2\cos\theta} \cos\rho$$
   • This implies that for a Hadamard walk ($\theta=\pi/4$), the probability ranges between $1/\sqrt{2}$ and $1 - 1/\sqrt{2}$ depending on the initial coin state.

2. Boundary-Proximal Starting Position ($N - k = \delta$ or $N + k = \delta$):

   • When the walker starts at a fixed distance $\delta$ from an absorbing barrier, a correction term appears.
   • This correction depends exponentially on the distance $\delta$ and the relative phase $\phi$.
   • Formula:
     $$P_L(\delta) \approx P_L(\text{fixed}) + q^{\delta-1} \frac{\sin\theta}{2(1+\sin\theta)} (\sin\rho \cos\phi \pm 1)$$
     where $q = \frac{1-\sin\theta}{1+\sin\theta}$.
   • This result highlights that the presence of the opposite boundary (even if far away) influences the absorption probability through wave reflections, provided the walker starts close to one boundary.

B. Parity Symmetry

The authors rigorously demonstrate how parity symmetry relates the absorption probabilities for starting positions $k$ and $-k$, allowing the analysis to be restricted to $0 \le k < N$ without loss of generality.

C. Numerical Convergence and Approximation

• Exponential Convergence: The numerical results confirm that the difference between the finite-$N$ absorption probability and the asymptotic limit decays exponentially with system size $N$. The decay rate is governed by $\log q$.
• Approximation for Finite $N$: Based on the asymptotic results, the authors propose a highly accurate approximation formula for finite $N$ and arbitrary $k$:
  $$P_L(k, N) \approx P_L(N-k) + P_L(-N-k) - P_L$$
  This "superposition" of boundary effects works exceptionally well even for small $N$ (e.g., $N=10$).

D. Physical Insight: The "Far" Barrier Effect

The paper discusses a counter-intuitive phenomenon regarding the order of limits.

• If one considers the limit $t \to \infty$ first (infinite time) for a fixed large $N$, the walker eventually hits the far barrier and reflects.
• If one considers $N \to \infty$ first (infinite line), the far barrier is never reached.
• The authors show that for large but finite $N$, the absorption probability initially stabilizes near the value for a single barrier (e.g., $2/\pi$ for Hadamard), but after a time $t_1 \approx 4N/v$ (time for the wave to travel to the far barrier and back), the probability shifts to the two-barrier asymptotic value. This demonstrates that the "presence" of a distant barrier is felt via wave reflections.

4. Significance

• Theoretical Advancement: The paper moves beyond implicit integral representations to provide exact, closed-form analytical solutions for a fundamental quantum transport problem. This clarifies the role of the coin parameter and initial state geometry in absorption.
• Algorithmic Implications: Understanding absorption probabilities is critical for quantum search algorithms and transport efficiency. The results show how to tune the coin parameter and initial state to maximize or minimize absorption at specific targets.
• Experimental Relevance: The authors suggest that these models are directly testable in photonic quantum systems (using time-multiplexing and programmable modulators to simulate sinks), providing a roadmap for experimental verification of these asymptotic behaviors.
• Methodological Impact: The successful application of two-scale convergence to quantum walk integrals offers a powerful new tool for analyzing asymptotic behaviors in other quantum lattice models.

In summary, this work provides a comprehensive analytical framework for understanding how quantum walkers interact with absorbing boundaries, resolving previous ambiguities and offering precise predictions for both theoretical limits and finite-size experimental setups.