In this paper [1], an arithmetic error led to the incorrect value of the terahertz electric-field amplitude. This was caused by incorrect calibration of the balance detection signal in the electro-optic (EO) sampling with a GaP EO crystal. All the electric-field amplitudes measured by the EO sampling method should be read as 0.6 times those indicated in the paper. Accordingly, the estimated pulse energy and all the pump fluences should be read as 0.36 times those indicated in the paper. The corrected maximum value of the electric field amplitude on the sample is 0.4 MV/cm at 1 THz, which corresponds to the Keldysh parameter $\gamma ~0.08$ for the gap energy of 1 eV, i.e. still far less than unity. Two of the fitting parameters in Eq. (2) should be corrected as A=0.2 (MV/cm)⁻² and E_th=$1.4\left(\pm 0.1\right)$ MV/cm. The nanotube diameter (d) dependence of the threshold electric-field (E_th) for the terahertz-pulse induced exciton generation is re-estimated as E_th (MV/cm)=0.9/d ² (nm²). This correction gives a better agreement with the theoretically predicted value E_th (MV/cm)~0.81/d ² (nm²) for the impact excitation mechanism considered in the paper.

These corrections do not change at all the conclusion of the paper on the experimentally observed terahertz pulse-induced exciton generation in carbon nanotubes.