10
$\begingroup$

The title says it all really.

Does this mean that the crystal is moving?

From my notes, I read that

The effect of an external force on an electron in the crystal is to change the crystal momentum $\hbar k$. In the absence of a force, the crystal momentum must be constant and thus conserved.

In a full band the net crystal momentum of electrons should be zero

For an electric field, $\mathcal{E_x}$, we find a change in crystal momentum, $k_x$, $$\hbar\frac{dk_x}{dt}=q\mathcal{E_x}$$

The missing ingredient, required to describe conduction, is scattering, which we assumed to be negligible. But where does the scattering come from? While the force of the electric field on electrons causes a change in crystal momentum, scattering must somehow restore the crystal momentum to their original values. Since Bloch oscillations are incredibly difficult to observe, electrons must be scattered before they can cross the FBZ boundary. The only other entities within the crystal with large crystal momentum are of course phonons.


These quotes are not related and quite honestly I have no idea what any of the quotes mean, as I don't know what is meant by 'crystal momentum'. If anyone wants to know the context from which those quotes came please let me know.

So, my question is as follows: In layman's terms (if possible), what is crystal momentum?


N.B

Before posting this question I looked at this question and I'm finding it hard to understand. I'm only a 2nd year undergraduate, who has begun reading solid state.

$\endgroup$
1

1 Answer 1

7
$\begingroup$

Crystal momentum arises when you consider the allowed energy eigenstates of the electrons which inhabit a crystal. One finds that the allowed energies in crystals are not discrete (as they are for an isolated atom), but rather form continuous bands separated by gaps.

In the plot below, I've plotted the first band in blue and the second band in red; the dotted lines on the vertical axis show the continuous intervals of allowed energies which correspond to each band.

enter image description here These states are labeled by two numbers, $n$ and $k$. $n$ is the band index, which tells you what energy band the state lives in; $k$ (or $\hbar k$, I suppose) is the so-called crystal momentum, which tells you where (left to right) in that energy band you are. In the above diagram, I plotted two different possible states - $n=1$ and $k=-1$, and $n=2,k=0.7$.

You should not think of $\hbar k$ as the physical momentum of the crystal itself. Rather, it is a kind of pseudo-momentum which labels the allowed electronic energy states. Neither should you think of it as the genuine momentum of the particular energy eigenstate which it labels, for the simple reason that the energy eigenstates do not have definite momentum.

Despite not being the genuine momentum, $\hbar k$ has some momentum-like properties. Applying an external force to the electrons in the system causes the crystal momentum of a particular state to change according to $$\frac{d}{dt}(\hbar \mathbf k) = \mathbf F_{ext}$$

Furthermore, if the electronic states interact with external particles (phonons, photons, etc), the sum of the crystal momentum of the electrons and genuine momentum of the external particles is conserved$^\dagger$. This is important when analyzing how electrons can jump from one energy state to the other by interactions with such particles.


$^\dagger$Not quite conserved - rather, conserved up to a reciprocal lattice vector. In other words, if $\mathbf k$ and $\mathbf k'$ are the crystal momenta before and after some event, then we say that $\mathbf k$ is conserved up to a reciprocal lattice vector if

$$\mathbf k' -\mathbf k = \mathbf G$$ for some reciprocal lattice vector $\mathbf G$ (possibly the zero vector).

$\endgroup$
19
  • 2
    $\begingroup$ @Connor It depends on what you mean by "layman." If you are not a student of physics, the best I can do is to say that when an electron travels through a crystal, its momentum is not conserved due to its interactions with the stationary ion lattice. However, there is a related quantity called crystal momentum which is more useful, and which is conserved in a slightly different sense. $\endgroup$
    – J. Murray
    Commented Mar 8, 2022 at 19:45
  • 1
    $\begingroup$ @Connor If you are a student of physics who has some familiarity with quantum mechanics and the terminology of elementary solid state physics (lattice vectors, reciprocal lattice vectors, etc) and still find my answer difficult to understand, then I would be happy to clarify any specific points that are giving you trouble. $\endgroup$
    – J. Murray
    Commented Mar 8, 2022 at 19:48
  • 2
    $\begingroup$ @Connor 1) From a semi-classical perspective, the electron experiences an electrostatic force from the ions, which means that its momentum will not be conserved. Recall that $\frac{d}{dt}\mathbf p = \mathbf F$. 2) No, the description I gave assumes that the lattice is completely rigid and stationary. If we allow lattice vibrations, then the crystal momentum of the electrons will not be conserved. However, in a scattering event the sum of the crystal momentum of the electron and the (genuine) momentum of the phonon will be conserved, which is an example of why crystal momentum is useful. $\endgroup$
    – J. Murray
    Commented Mar 8, 2022 at 19:58
  • 1
    $\begingroup$ 1) In principle, yes. However, in practice we generally take the lattice to be infinitely large and absolutely stationary. In such a model, the lattice is not subject to Newton's laws and momentum is simply not a conserved quantity. This is similar to what we would do when modeling a basketball bouncing on the ground; in principle total momentum is conserved and the Earth recoils by some infinitesimal amount with each bounce, but in practice we don't account for this and simply say that momentum isn't conserved. 2) Not in any direct way I can think of. It's just an analogous [...] $\endgroup$
    – J. Murray
    Commented Mar 8, 2022 at 20:13
  • 1
    $\begingroup$ [...] quantity to momentum which is useful inside a crystal. $\endgroup$
    – J. Murray
    Commented Mar 8, 2022 at 20:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.