17. Ladder Operators
Daniel V. Schroeder
Operators in quantum mechanics aren’t merely a convenient way to keep track of
eigenvalues (measurement outcomes) and eigenvectors (definite-value states). We
can also use them to streamline calculations, stripping away unneeded calculus (or
explicit matrix manipulations) and focusing on the essential algebra.
As an example, let’s now go back to the one-dimensional simple harmonic oscillator, and use operator algebra to find the energy levels and associated eigenfunctions.
Recall that the harmonic oscillator Hamiltonian is
1 2 1
p + mωc2 x2 ,
where p is the momentum operator, p = −ih̄d/dx (in this lesson I’ll use the symbol
p only for the operator, never for a momentum value). As in Lesson 8, it’s easiest
to use natural units in which we set the particle mass m, the classical oscillation
frequency ωc , and h̄ all equal to 1. Then the Hamiltonian is simply
H = p 2 + x2 ,
with p = −id/dx.
The basic trick, which I have no idea how to motivate, is to define two new
operators that are linear combinations of x and p:
a− = √ (x + ip),
a+ = √ (x − ip).
These are called the lowering and raising operators, respectively, for reasons that
will soon become apparent. Unlike x and p and all the other operators we’ve worked
with so far, the lowering and raising operators are not Hermitian and do not represent any observable quantities. We’re not especially concerned with their eigenvalues
or eigenfunctions; instead we’ll focus on how they mathematically convert one energy eigenfunction into another. (Note: Most authors use the notation a and a†
instead of a− and a+ , but Griffiths uses the −/+ notation, and I too think it’s
more appropriate. In conventional units, by the way, we would insert factors of m,
ωc , and h̄ into these definitions as needed to make a− and a+ dimensionless. It’s a
good exercise to figure out exactly how to do this, especially if you’re even a little
uncomfortable with my use of natural units.)
The first thing to notice about a− and a+ is that if you multiply them together,
you almost get the Hamiltonian:
a− a+ = 21 x2 + p2 − i(xp − px)
= H − [x, p]
= H + 12 ,
where I’ve used the basic commutator relation [x, p] = ih̄ (with h̄ = 1) that you
worked out for homework. Similarly, multiplying them together in the other order
gives a+ a− = H − 21 . We can therefore write the Hamiltonian as
H = a− a+ −
= a+ a− + 12 .
Now suppose that ψ is any energy eigenfunction, so that Hψ = Eψ for some E.
Then I claim that if we act on ψ with either the raising or the lowering operator,
we get another energy eigenfunction. To prove this, I need to check what happens
when the Hamiltonian operator acts on this new function. Here’s what happens in
the case of a+ ψ:
H(a+ ψ) = (a+ a− + 12 )(a+ ψ)
= a+ (a− a+ + 12 )ψ
= a+ (H + 1)ψ
= a+ (E + 1)ψ
= (E + 1)(a+ ψ).
So a+ ψ is an energy eigenfunction, and its eigenvalue is exactly one unit greater than
that of ψ itself. (In ordinary units, this one unit of energy would be h̄ωc .) This is why
a+ is called the raising operator: it mathematically raises any energy eigenstate ψ
up the quantum ladder by one rung, to the next-highest energy eigenstate. There’s
no limit to how many times we can apply the raising operator, so this proves that
a quantum harmonic oscillator has an infinite ladder of energy eigenstates, with
equally spaced levels separated in energy by h̄ωc .
By a completely analogous calculation you can show that a− lowers any energy
eigenstate ψ by one rung down the quantum ladder:
H(a− ψ) = (E − 1)(a− ψ).
This is why a− is called the lowering operator. This result would seem to imply
that our infinite ladder of energy eigenstates continues downward in energy without
limit—but that can’t possibly be the case, because there can’t be any states with
negative energy when the potential energy is positive everywhere. The only way
out is if, when a− acts on the lowest-energy eigenstate (call it ψ0 ), it gives zero:
a− ψ0 = 0.
Then equation 7 can still be true for ψ0 even though there’s no lower-energy state.
And from this simple equation we immediately see that Hψ0 = 12 , so the groundstate energy is half a unit, or 12 h̄ωc . The remaining energies are integer steps above
this, so we’ve proved that the energies of a quantum harmonic oscillator are as
claimed in Lesson 8:
En = (n + 12 )h̄ωc ,
n = 1, 2, 3, . . . .
(If you’re reading carefully you may have noticed a couple of loopholes in the
logic I just described. How do we know the quantum ladder goes up indefinitely,
rather than ending with a state ψmax for which a+ ψmax = 0? And how do we know
that the ladder that starts with ψ0 includes all the states? Could there be others
that we’ve missed, perhaps lying in between the ones we’ve found? These loopholes are actually pretty easy to close, and I’d encourage you to think about them,
but only after you’re comfortable with the rest of the calculations and reasoning
To find the energy eigenfunctions that correspond to the eigenvalues En , we still
need to use calculus—but at least the method is intuitive and straightforward. To
find the ground-state eigenfunction we can use equation 8, which becomes an ordinary differential equation for ψ0 (x) when we express the lowering operator in terms
of the explicit momentum operator, p = −id/dx. The solution to the differential
equation is ψ0 (x) = Ae−x /2 , just as we saw in Lesson 8 (and you can work out
the normalization constant A if you like). Then, to find the first excited state, just
apply the raising operator, also written in terms of p = −id/dx, to the ground state
(and again work out the normalization constant if you need it). Keep applying the
raising operator to work your way up the quantum ladder until the novelty wears
off. As you might guess, it gets pretty tedious to work out more than the first few
eigenfunctions by hand.
I hope you agree that the ladder-operator method is by far the most elegant way
of solving the TISE for the simple harmonic oscillator. The bad news, though, is that
no such elegant method exists for solving the TISE for other one-dimensional potential functions; the method worked here only because the Hamiltonian is quadratic
in both p and x, allowing it to be factored, aside from an additive constant, into
the product a− a+ . So why spend time learning a method—however elegant—whose
applicability is so restricted? There are two reasons.
The first reason is that harmonic oscillators really are ubiquitous in nature,
from vibrating molecules to elastic solids to the electromagnetic field and other
fundamental fields of elementary particle physics. In continuous vibrating systems
we refer to the units of excitation energy as particles, such as photons in the case of
the electromagnetic field, or as quasiparticles, such as phonons in the case of elastic
vibrations. The study of these continuous vibrating quantum systems is called
quantum field theory, and ladder operators are a fundamental tool of quantum field
theorists. But we won’t have time to explore quantum field theory in this course.
The second reason, though, is that ladder operators will come up again in this
course in a somewhat different context: angular momentum. Instead of adding and
removing energy, the ladder operators in that case will add and remove units of
angular momentum along the z axis. They will therefore be an extremely useful
tool in our study of systems with spherical symmetry, especially atoms.