For the proposed Gaussian wave packet, differentiation gives
Inserting these in the Time-dependent Schrödinger equation with Hamiltonian , its right-hand side becomes
Equating the constant and quadratic coefficients with yields
Conversely these two equations make the substitution an identity, so they are sufficient for the Gaussian evolution in an inverted harmonic oscillator.
Set with real constant . Then
Thus solves the Riccati equation wherever it is finite. Integrating the amplitude equation also gives
with the square root chosen continuously from . For the normalizable solutions used in part (c), ; equivalently one can choose modulo . The unrestricted real parameter solves the differential equation locally, but not every value defines a normalizable wavefunction.
Write with . Normalization of the Gaussian wave packet gives , hence . The supplied Gaussian integral ratio then gives
The momentum operator obeys , and direct second differentiation gives . Taking the expectation value and substituting the first result gives
since . These are the second moments of a complex Gaussian wave packet.
For the solution in part (b), the trigonometric identity
gives, writing ,
Thus normalizability requires , and the exact expectation values are
Since ,
Their difference is the constant , so the mean energy stays constant despite exponential spreading in both observables.

Articles by others on the same topic (0)

There are currently no matching articles.