First principles analysis of electronic structure evolution and the indirect- to direct-gap transition in Ge$_{1-x}$Pb$_{x}$ group-IV alloys (1911.05679v1)

Abstract: We present a theoretical analysis of electronic structure evolution in the group-IV alloy Ge${1-x}$Pb${x}$ based on density functional theory. For ordered alloy supercells we demonstrate the emergence of a singlet conduction band (CB) edge state, suggesting the emergence of a direct band gap for Pb compositions as low as $x \approx 1$%. However, application of hydrostatic pressure reveals Pb-induced hybridisation, with the CB edge state in a Ge${63}$Pb${1}$ ($x = 1.56$%) supercell retaining primarily indirect (Ge L${6c}$) character. For an ordered Ge${15}$Pb${1}$ ($x = 6.25$%) supercell we find that the CB edge has acquired primarily direct (Ge $\Gamma{7c}$) character, confirming the presence of an indirect- to direct-gap transition. The importance of alloy disorder is highlighted by investigating the impact on the electronic structure of the formation of a nearest-neighbour Pb-Pb pair. Having established the importance of short-range disorder, we analyse the electronic structure evolution as a function of $x$ using a series of 128-atom special quasi-random structures (SQSs). Our calculations reveal a strong reduction (increase) of the band gap (spin-orbit splitting energy), by $\approx 100$ meV ($\approx 40$ meV) per % Pb replacing Ge. We find an indirect- to direct-gap transition occurring in a narrow composition range centred about $x \approx 7$%, close to which composition we calculate that the alloy becomes semimetallic. Further analysis suggests that long-range order introduced by Born von Karman (supercell) boundary conditions leads to overestimated energy splitting of the Ge L${6c}$-derived CB states in the 128-atom SQSs. Accounting for these finite-size effects, we expect a direct band gap to emerge in Ge${1-x}$Pb$_{x}$ for $x \gtrsim 3 - 4$%.