Since all Eigenstates of the Hamiltonian are stable, one needs at least a short perturbation of the system to get an excited state; e.g.:
A beam of
electrons
electromagnetic radiation
positrons
neutrons
...
The scattering may be elastic or inelastic, i.e. momentum and/or energy may be exchanged.
Electrons can be scattered by core electrons
There exists an interaction between electrons and phonons
The thermal energy of the system itself causes an excited state since only at \(T = 0\) the Fermi energy describes the highest occupied energy state. We find a dynamic equilibrium between states steadily increasing or decreasing the energy.
Surfaces and defects in a crystal change the Hamiltonian of the perfect solid leading to additional electronic transitions.
For all these processes according to Eq. (3.18) the transfer matrix element can be calculated; they quantize the probability for these transitions.
The excited state may loose energy by the same processes as described above:
by electromagnetic radiation
the same particles, which moved into the solid, loose energy and momentum and will be detected in the scattered beam.
thermal processes (e.g. phonons)
at defects
on surfaces
REMARK: Principally you do not need superconductivity to find zero ohmic resistance; once
a current flow started (described by the \(k\)-distribution of electrons, which are Eigenstates of the Hamiltonian)
it would be stable for infinite time. You need defects within the crystal or phonons to reduce the current by scattering.
For every well defined measurement and if knowing the underlying processes, quantum
mechanics provides the adequate calculation rule.
Which is the starting state?
Which is the end state?
Which expectation value do I have to calculate?
Examples:
Scattering experiment:
incoming free particle
outgoing free particle
Scattering cross section
Recombination processes:
State in the first band
State in the second band
Cross section of a trap (a surface)/Probability for a radiative transition
This allows to calculate which processes have the highest transition probability
Often approximation methods are used to calculate matrix elements
Essential for all calculations is the Hamiltonian of the perfect system, i.e. the Eigenvalues and Eigenvectors. Quantum mechanical Hamiltonians almost always look like:
| \begin{equation*} H = \sum_i E_i |f_i\rangle \langle f_i| \end{equation*} | (3.21) |
© J. Carstensen (Quantum Mech.)