I am taking an introductory course in condensed matter physics, and am absolutely stumped by the concept of the Brillouin zone and backfolding of the dispersion curve in the nearly free electron model.

Specifically, when we assume that the lattice is periodic IN REAL SPACE with a period a, then we can represent this as a fourier series and the significant k vectors (or G vectors) i.e. the reciprocal lattice vectors, are spaced $2\pi /a$ apart. Now with my understanding, we have no fourier components which have k vectors intermediate to these G vectors. These are the permitted wavevectors to describe any function (e.e the electron wavefunction) in order to ensure that it has the correct periodicity. It should have no components with k smaller than the first reciprocal lattice vector $2\pi /a$ (or if we centre our reciprocal space intermediate to two lattice points, we can go from $\pm \pi /a$.

Why then do we backfold the free electron dispersion curve into the first Brillouin zone when we introduce periodicity? This si the biggest mystery to me, and I'm certainly not getting something important here. To me, periodicity in real space means that all of the information is contained in a fucntion defined on the small interval which is the space occupied by a unit cell. In reciprocal space, small intervals in real space translate to large distances in reciprocal space. So it makes absolutely no sense to me as to why we then say that all of the information is contained within a small interval of the first Brilluin zone in k space. It should be the other way around! If the only unique structure in real space is that contained within the unit cell, then the description of this in k space should require the fourier components at all of the recirprocal lattice points going to infinity!

I would greatly appreciiate if someone could clarify why we talk of all of the information being contained in the first Brillouin zone, and why supposedly k vectors outside of the first brillouin zone are equivalent to those within the first brillouin zone.

  • $\begingroup$ The electron wave function is under no obligation to have the periodicity of the lattice. It is only the potential through which it travels, that has this periodicity. $\endgroup$ – LLlAMnYP May 17 '18 at 10:55
  • $\begingroup$ @LLlAMnYP in my lecture notes it says , and i quote, "the translatiinal symmetry of the lattice means tgat the electron probability density must have the same periodicity" $\endgroup$ – Meep May 17 '18 at 12:13
  • 1
    $\begingroup$ Oh, sure, for delocalized electrons that's true, though not necessarily so for excitations. But the probability density ($|\Psi^2|$) is not the same thing as the wavefunction $\Psi$, which can have a phase multiplier $\exp(i \mathbf{k r})$ where $\mathbf{k}$ isn't necessarily a reciprocal lattice vector. $\endgroup$ – LLlAMnYP May 17 '18 at 12:22
  • $\begingroup$ @LLlAMnYP Okay I think I get that- the k vector for the electron wavefunction is restricted only by the whole crystal lattice dimensions, and this is the k vector that is represented in the first Brillouin zone. However still don't understand why the electron wavefunction plane wavelength has a smallest wavelength given by $2\pi /a$ where $a$ is the unit cell distance in the periodic lattice. In the case of phonons, this is because there is nothing on a smaller scale to carry the wave. If we tried to imagine a wave with a smaller wavelength, it would physically correspond (in terms of the $\endgroup$ – Meep May 17 '18 at 12:58
  • $\begingroup$ lattice point ibrations) to a wave of a longer wavelength that fits inside the first Brillouin zone (essentially, aliasing occurs). However when it comes to electron wavefunctions, this is dependent on the electron density and this can vary within a unit cell of the lattice . I can imagine fitting two electron wavefunction wavelengths within the distance of ion separation. The key difference here for me is that the electrons, which carry the waves, don't have the same scale restrictions as phonons, which are intrinsically oscillations of the lattice points and not within the lattice points. $\endgroup$ – Meep May 17 '18 at 13:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.