Boundary conditions for the Schrödinger equation in the numerical simulation of quantum systems

This paper demonstrates that while closed quantum systems can be defined by local boundary conditions, open systems cannot due to the uncertainty principle, and proposes a method using a small numerical lattice to simulate open systems while preserving the physical reality of infinitely extended waves.

The Gist

Imagine you are trying to film a movie of a tiny, invisible particle (like an electron) moving through a world. In the real universe, this particle can travel forever, bouncing off walls, passing through barriers, or spreading out like a ripple in a pond.

But in a computer simulation, you have a problem: Your computer screen is finite. You can't show an infinite universe on a 10-inch monitor. You have to draw a box, a "grid," to do the math.

This paper, written by Marco Patriarca, tackles a tricky puzzle: How do you simulate a particle that is supposed to be traveling from "infinity" without the edges of your computer screen ruining the scene?

Here is the story of the paper, broken down into simple concepts and analogies.

1. The Two Types of Worlds: The Cage vs. The Highway

The author distinguishes between two types of quantum systems:

• The Closed System (The Cage): Imagine a particle trapped inside a box with solid, reflective walls. It bounces back and forth forever.

  • The Solution: This is easy to simulate. You just tell the computer, "If the particle hits the edge, stop it." In math, this is a local boundary condition. It's like saying, "At this specific point on the grid, the value must be zero." It works perfectly because the particle is contained.

• The Open System (The Highway): Imagine a particle zooming down an infinite highway. It might hit a bump (a barrier) and bounce back, or it might keep going.

  • The Problem: You cannot simulate an infinite highway on a finite grid. If you try to start a "plane wave" (a perfect, endless ripple) at the very edge of your screen, you run into a physics paradox called the Uncertainty Principle.
  • The Analogy: Think of the Uncertainty Principle like a rule that says, "You can't know exactly where a car is and exactly how fast it's going at the same time."
    • If you try to inject a perfect, endless wave at a single point on your grid, you are saying, "This wave exists only here and only now." But a perfect wave needs to exist everywhere to have a perfect speed.
    • Trying to force a perfect wave into a tiny box is like trying to fit a 100-mile-long train onto a 10-foot platform. The physics breaks.

2. The Old Way vs. The New Trick

The Old Way (The Wave Packet):
Usually, scientists simulate these open systems by creating a "wave packet"—a short, fuzzy blob of a wave that looks like a train car. They shoot this blob across the screen.

• The Flaw: To make this blob look like a perfect, endless wave (a plane wave), the blob has to be huge. If you want to simulate a slow-moving wave, the blob needs to be miles long. Your computer would need a grid the size of a galaxy to hold it. This is too slow and requires too much memory.

The New Trick (The "Magic Injection" Point):
Patriarca proposes a clever workaround. Instead of trying to fill the whole screen with a wave, you create a special injection point on the grid.

• The Analogy: Imagine a river (the wave) flowing toward a dam (the barrier).
  • Normal Simulation: You try to fill the whole riverbed with water before you start the movie.
  • Patriarca's Method: You just put a faucet at a specific spot on the riverbank. You turn the faucet on, and water flows out.
  • The Magic: The computer math is tweaked so that at the faucet, the water looks like it's coming from an infinite distance upstream, even though the riverbed (the grid) is short.

3. How the Trick Works (The "Subtraction" Magic)

Here is the secret sauce of the method, explained simply:

1. The Setup: You pick a point on your grid (let's call it Point S) to be your "source."
2. The Injection: You force the wave to appear at Point S.
3. The Right Side (Transmission): To the right of Point S, the wave moves forward. When it hits a barrier, some bounces back, and some goes through. To stop the "through" wave from hitting the right edge of your screen and bouncing back (which would ruin the simulation), you put up an Imaginary Wall.
   • Analogy: This isn't a real wall; it's like a "black hole" or a sponge at the end of the screen that swallows the wave so it never comes back.
4. The Left Side (Reflection): This is the hardest part. When the wave hits the barrier, it bounces back toward the source.
   • The Problem: At the source, you have the incoming wave (from the faucet) and the reflected wave (bouncing back) mixing together. It looks like a mess of interference.
   • The Solution: The computer math is modified to subtract the known incoming wave from the total mess.
   • Analogy: Imagine you are watching a tennis match. The ball (reflected wave) comes back at you, but there's also a fan blowing wind (the incoming wave) that you know about. The computer says, "I know the wind is blowing at 10 mph. Let me subtract that wind from the total air movement so I can see only the ball."
   • By doing this subtraction, the computer isolates the "reflected wave" and lets it travel back to the left edge, where another "sponge" (imaginary potential) swallows it.

4. Why This Matters

This method is a game-changer because:

• It saves space: You don't need a massive grid. You can use a small, manageable computer screen.
• It handles the impossible: You can simulate perfect, endless waves (plane waves) and complex, changing barriers without needing to wait for a giant wave packet to travel across the screen.
• It's flexible: It works for static barriers (like a wall) and dynamic ones (like a wall that vibrates or changes shape over time).

Summary

The paper solves the problem of simulating infinite quantum worlds on finite computers.

• Closed systems are like fish in a bowl; easy to simulate with simple walls.
• Open systems are like birds flying in an infinite sky; hard to simulate because you can't draw the whole sky.
• The Solution: Instead of drawing the whole sky, you create a "magic faucet" that injects the bird's path, and you use "mathematical sponges" to catch the bird when it flies off-screen. You also use a "mathematical eraser" to separate the bird's return flight from the wind that pushed it there.

This allows scientists to study how particles scatter, tunnel, and interact with time-changing forces with high precision, using much less computer power than before.

Technical Summary

1. Problem Statement

The paper addresses a fundamental challenge in the numerical simulation of the time-dependent Schrödinger equation: the definition of appropriate boundary conditions for open quantum systems.

• Closed Systems: For closed systems (where probability is conserved within a finite region), standard local boundary conditions (e.g., $\psi=0$ at the boundaries) are well-defined and easily implemented.
• Open Systems: For open systems (e.g., scattering problems involving incident plane waves or wave packet trains), standard local boundary conditions fail.
  • The Uncertainty Principle Constraint: A true plane wave extends infinitely in space. To simulate it on a finite numerical lattice using standard methods, one would need a grid size proportional to $1/k$ (where $k$ is the wave vector). For small $k$ (long wavelengths), the required grid size becomes computationally prohibitive.
  • The "Injection" Paradox: Attempting to inject a plane wave at a specific point $x$ with a defined momentum creates a contradiction with the Heisenberg Uncertainty Principle. A perfectly defined momentum implies infinite position uncertainty, meaning the wave cannot be "localized" at an injection point without violating physical laws. Consequently, standard local boundary conditions cannot simulate a plane wave entering a system from infinity.

2. Methodology

The author proposes a novel numerical method that modifies the finite-difference equation to simulate open systems on a small, finite lattice while maintaining physical correspondence with infinitely extended waves.

A. The Numerical Framework

• Equation: The 1D Schrödinger equation is discretized in space (using a second-order finite difference approximation) but kept continuous in time for the derivation.
  $$i \frac{\partial \psi_j}{\partial t} = -\frac{1}{\Delta x^2}(\psi_{j+1} + \psi_{j-1} - 2\psi_j) + V_j \psi_j$$
• Absorbing Boundaries: To prevent unphysical reflections from the edges of the finite lattice, imaginary potentials ($V_i(x) = -ic(x-x_i)^2$) are introduced near the left and right boundaries. These absorb outgoing waves (reflected and transmitted), effectively simulating an infinite domain.

B. The Core Innovation: Wave Injection and Separation

The method allows for the injection of an incident wave $\Phi_0(x,t)$ at a specific lattice point $x_s$ and the subsequent separation of the total wave function $\Psi_{Tot}$ into incident, reflected, and transmitted components.

1. Injection Point ($x_s$): The incident wave $\Phi_0$ is generated at $x_s$.
2. Left Region ($j < s$): The lattice points to the left of the injection point are used to track only the reflected wave ($\Psi_R$).
   • Mechanism: At the boundary $j=s$, the incident wave value $\Phi_0$ is subtracted from the total wave function in the finite-difference equation. This ensures that the evolution equation for $j < s$ only "sees" the reflected component.
3. Right Region ($j > s$): The lattice points to the right track the total wave function ($\Psi_{Tot} = \Phi_0 + \Psi_R + \Psi_T$).
   • Mechanism: At $j=s+1$, the incident wave value $\Phi_0$ is added to the neighbor term to ensure the total wave evolves correctly through the potential region.
4. Handling Scattering:
   • The Transmitted Wave is observed in the region $x > x_{potential}$ before being absorbed by the right imaginary potential.
   • The Reflected Wave is isolated in the region $x < x_s$ by the subtraction technique, allowing for its direct study without interference from the incident wave.

3. Key Contributions

• Resolution of the Boundary Condition Paradox: The paper demonstrates that while local boundary conditions are impossible for open systems due to the uncertainty principle, a modified finite-difference scheme can simulate the effect of an infinite wave entering a finite system.
• Decoupling of Components: The method provides a rigorous way to numerically separate incident, reflected, and transmitted waves in real-time simulations, which is difficult with standard wave packet approaches.
• Applicability to Non-Stationary Systems: Unlike perturbation theory or partial wave analysis (which often assume stationary states), this method handles time-dependent potentials and transient phenomena directly.
• Efficiency: It allows the simulation of plane waves and large wave packets using a small computational grid, avoiding the need for massive lattices required by traditional wave packet approximations.

4. Results and Validation

The author validates the method through several numerical examples using the Crank-Nicolson implicit difference method:

• Plane Wave Propagation: Successfully simulated a plane wave ($k=2.4$) injected at $x=-15$ and absorbed at the right boundary. The quantum current converged to the expected incident current $J_0 = 2k|A|^2$.
• Static Scattering (Square Barrier):
  • Simulated scattering off a square potential barrier ($V_0=5$).
  • Visualized the interference pattern between incident and reflected waves in the region $x_s < x < \text{barrier}$.
  • Confirmed that for $x < x_s$, the interference vanished, leaving only the pure reflected wave, validating the subtraction technique.
  • Quantitative Agreement: Calculated reflection ($R$) and transmission ($T$) coefficients matched theoretical analytical solutions for both $k^2 > V_0$ and $k^2 < V_0$ cases.
• Time-Dependent Scattering: Simulated a plane wave scattering off a barrier with an oscillating height ($V(x,t) = V_0[1 + \alpha \cos(\omega t)]$). The method successfully captured the time-evolution of currents and spatial oscillations without approximation.

5. Significance

This work provides a robust numerical tool for quantum dynamics that bridges the gap between theoretical idealizations (infinite plane waves) and computational limitations (finite grids).

• Versatility: It is applicable to linear and nonlinear potentials (e.g., Hartree-Fock approximations) and self-consistent equations.
• Physical Insight: By isolating reflected and transmitted components, it allows for a detailed study of scattering phenomena, including transient effects and chaotic behavior in nonlinear tunneling, which are difficult to analyze with other methods.
• Computational Efficiency: It eliminates the need for excessively large grids or complex wave packet constructions, making the simulation of scattering by time-dependent potentials feasible on standard hardware.

In summary, Patriarca's method transforms the numerical simulation of open quantum systems by redefining how boundary conditions are applied, effectively "injecting" waves from infinity into a finite lattice while mathematically separating the physical components of the scattering process.