The sublinearity of the pressure dependence of the band gap of tetrahedral semiconductors is not well represented by the usual quadratic fit to pressure, nor by the linear fit to lattice constant, and this is responsible for much of the scatter in reported results. Linear fitting to the change in density eliminates this scatter. Thus high-pressure experimentation is shown to be much more accurate than is supposed. We propose that the linear pressure coefficient should be defined and reported as the rate of change of band gap with pressure at zero pressure, obtained using a suitable equation of state and performing a linear fit for the density deformation potential. Data for III-V and II-VI semiconductors are reassessed accordingly and recommended linear pressure coefficients are given.

• Received 22 August 1997

DOI:https://doi.org/10.1103/PhysRevB.58.12579