Abstract
Original article: EPL, 126 (2019) 67004.
Export citation and abstract BibTeX RIS
Due to an error in our numerical calculations for the matrix element in eq. (10), the optical conductivity ![$\Re[\sigma_{xx}]$](https://content.cld.iop.org/journals/0295-5075/128/3/39901/revision1/epl19907ieqn1.gif)
of fig. 1(a) does not have the correct shape. Instead of a slight decrease as a function of ω, once the transitions 
and 
become active, the optical conductivity is slightly increased (see corrected fig. 1(a) below). In fact, the final exact behavior of the optical conductivity as a function of the photon energy is determined by the interplay between the matrix element and JDOS. Due to the ω-dependence in the matrix element, the steps can be curved upwards as in 
and downwards as in 
.
However, the mathematical expressions of our paper and the physical conclusions remain valid.
Fig. 1: ![$\Re[\sigma_{zz}]$](https://content.cld.iop.org/journals/0295-5075/128/3/39901/revision1/epl19907ieqn6.gif)
without (blue) and with (orange) a parallel magnetic field given by 
, measured by unity of 
and excitation energy by the half bulk gap. The corresponding optical transitions of each peak are labeled above. The blue and orange peaks in the light-purple shaded area, where the bulk absorption signals are plotted in purple, are immersed in the bulk spectrum and thus cannot be seen experimentally.
Download figure:
Standard image