Surface states of 3D topological insulators

[EdB]

I have a question (more like a curiosity) related to three-dimensional topological insulators, which support Dirac-like states at their surfaces. From the theory, it is well known that these states are immune to scattering from non-magnetic impurities, i.e. impurities that do not break time-reversal symmetry. Therefore, they are topologically protected surface states.

Now, when one performs an experiment to image these surface states, the best and clearest signature is provided by angle-resolved photoemission spectroscopy (ARPES). This technique is performed under ultra-high vacuum, which means under a vacuum of < 10^-10 torr. This is done to minimize the collisions between the photoemitted electrons and the remaining particles in the ARPES chamber environment.
To image the surface states, a high-quality single crystal is cleaved in situ, which means that the material is cracked inside the ARPES chamber to expose a clean and fresh surface that can be imaged by the spectrometer. Who has performed this technique, knows that the surface states of 3D topological insulators can survive only up to a few hours/days and eventually the spectra will get blurred over time. So, after some amount of time, the spectra degrade because of the dirty environment in the ARPES chamber. What is the mechanism behind this blurring? If the surface states are immune to non-magnetic impurities and in the absence of any non-magnetic impurity, these states should survive over time. What am I missing here?

 

[Henryk]

What am I missing here?

Adsorption!. Even at such high vacuum, there will be adsorption of molecules (mostly water) on the cleaved surface. The adsorbed molecules alter surface electronic states.

 

[EdB]

Thanks for the answer, Henryk. I do agree with you in the case of trivial surface states in a band semiconductor, while I find more difficult to imagine how adsorbed non-magnetic molecules can microscopically interact to "destroy" topologically-protected electronic states. Should I interpret your answer as "adsorption masks any surface state in the photoemission process"? Does this imply that the topologically-protected states survive but they cannot be simply imaged?

 

[Henryk]

it is my understanding that adsorbed molecules alter the surface states. I am not quite sure about the mechanism, it could as simple as an electric field of an adsorbed molecule shifts the energy of the state. I do not thing that a monolayer of any species is actually capable of screening high energy electron beam.