Underbarrier interference (Q531310)

The paper considers tunneling through 2D barriers occurring in a different scenario than Wentzel-Kramers-Brillouin (WKB). In difference to the traditional approach in which a decay of a state with a tangent momentum is not considered since it does not correspond to a saddle point and the net contribution is supposed to be averaged down to a small value, it is shown that the conclusion is not correct and states with tangent momentum can play a crucial role in tunneling processes. First, into framework of the formulation problem the tunneling occurs in the \(x\) direction from a straight infinitely long quantum wire aligned along the \(y\) axis, and a particle is described by the Schrödinger equation. It is found a wave function for a wide barrier when the electric field is zero or finite. The author uses a semiclassical approach when only the exponential part of the wave function is calculated. In the main exponential approximation the action \(S\) is large and the prefactor \(C\) is less important. Quantum fluctuations around the saddle determine the preexponential factor. A semiclassical method of the solution of the Schrödinger equation in the case of a finite electric field is reduced to the Hamilton-Jacobi formalism. The classical action is calculated as a solution of the Hamilton-Jacobi equation. The action is also tracked along classical trajectories in imaginary time. It is shown that the both methods lead to the same tunneling probability. When two branches of a wave function exist at the same spatial domain, the smaller one can unexpectedly disappear. From the semiclassical standpoint it happens as a jump in space of the prefactor \(C(x,y)\) down to zero, stating the Stokes phenomenon. It is shown that there are no Stokes jumps in the semiclassical solution obtained. While the exact solution is unknown, it is proved that exponential multiplier is the main part of the exact wave function and the prefactor plays a secondary role. Then there are studied the classical trajectories in imaginary time. It is shown that the path obtained lies in the complex plane providing a real tangent velocity. This path can be treated as a saddle point with fluctuation around it. Further, there are discussed the problems of numerical calculations and how a particle penetrates through the barrier. A semiclassical method based on the Hamilton-Jacobi equation is applied to tunneling through a barrier with an impurity having a particular form of the impurity potential. It is considered tunneling from a homogeneous quantum wire through the barrier with the impurity. The author outlines, by general arguments, the nature of the underbarrier interference. To draw more exact conclusions it is used the rigorous method of summation of various propagating waves. This method is generic with a saddle formalism which allows one to collect rapidly oscillating functions. Finally, the problem of a particle in the 2D potential is considered from the standpoint in which two different particles described by spatial coordinates move in identical 1D potentials (two particles to be in the entangled state).