4.2 Hamiltonian
We shall use the correspondence principle to derive the quantum Hamiltonian from the classical one. Unfortunately electrodynamics is most readily formulated in the Lagrangian formalism, where one uses the variables
and
, rather than
and
. We shall have to do a little work to end up with a Hamiltonian, which we can then quantise using the correspondence principle.
4.2.1 Lagrangian approach
It is straightforward to write the Lagrangian for a charged particle in an electromagnetic field in terms of the scalar and vector potentials, see Ref.
[12], pp. 23.
We use the definition of the electromagnetic potentials in vacuum as
This allows us to write an expression for the Lagrangian for a particle in the e.m. field as
Indeed the Lagrangian equation of motion (the generalisation of
)
becomes
which can be written as
which is the standard result you have seen many times before. There are two important algebraic steps in the derivation above that deserve some attention: First of all, the total time derivative
in (
↓) acts both implicitely on the argument
of
, since
is time dependent, and explictely on the argument
. This gives rise to the two last expressions on the left-hand side of Eq. (
↓). Secondly, from the first to the second line of (
↓) we have used the standard result (
↓).
4.2.2 Canonical momentum
The important reason to write down the complicated expression (
↓) above, is to derive a Hamiltonian, which is formulated in terms of momenta; we follow the standard procedue, and find that the form of the canonical momentum is simple but surprising. From the definition we find
The canonical momentum in an electromagnetic field thus
includes the vector potential.
The classical Hamiltonian, defined as
can be written as
This is the form we shall use this course.
4.2.3 Quantum Hamiltonian
The quantum Hamiltonian is constructed by the correspondence principle, substituting operators where we had classical variables.
where the electro-magnetic potentials
and
are simply functions of the coordinate operator, and as always
A choice had to be made in writing down (
↓), since
and
do in general not commute, so the order of multiplication matters. The form used here is called the “minimal coupling substitution”, and is the one commonly used.
4.2.4 Conservation of momentum
There is a problem with Eq. (
↓), which we notice by looking at the commutator of the Hamiltonian in zero electric potential,
, with (a component of) the momentum operator. We use Eq. (
↓), and the fact that
does not commute with
, since the latter depends functionally on
,
and this is obviously not zero! But it is trivial to show that the momentum-like quantity
commutes with
and is thus a constant of the motion (conserved). This should not surprise us; e.g., if
describes a homogeneous magnetic field we know classically that the particle moves in a spiral about the magnetic field lines. We are just seeing the quantum analogue.