Numerical Methods in Quantum Mechanics: Analysis of Numerical Schemes on One-Dimensional Schrodinger Wave Problems.

摘要

The motion and behavior of quantum processes can be described by the Schrodinger equation using the wave function, Psi(x,t). The use of the Schrodinger equation to study quantum phenomena is known as Quantum Mechanics, akin to classical mechanics being the tool to study classical physics. This research focuses on the emphasis of numerical techniques: Finite-Difference, Fast Fourier Transform (spectral method), finite difference schemes such as the Leapfrog method and the Crank-Nicolson scheme and second quantization to solve and analyze the Schrodinger equation for the infinite square well problem, the free particle with periodic boundary conditions, the barrier problem, tight-binding hamiltonians and a potential wall problem.;We discuss these techniques and the problems created to test how these different techniques draw both physical and numerical conclusions in a tabular summary. We observed both numerical stability and quantum stability (conservation of energy, probability, momentum, etc.). We found in our results that the Crank-Nicolson scheme is an unconditionally stable scheme and conserves probability (unitary), and momentum, though dissipative with energy. The time-independent problems conserved energy, momentum and were unitary, which is of interest, but we found when time-dependence was introduced, quantum stability (i.e. conservation of mass, momentum, etc.) was not implied by numerical stability. Hence, we observed schemes that were numerically stable, but not quantum stable as well as schemes that were quantum stable, but not numerically stable for all of time, t. We also observed that second quantization removed the issues with stability as the problem was transformed into a discrete problem. Moreover, all quantum information is conserved in second quantization. This method, however, does not work universally for all problems.