Influence of the “second gap” on the transparency of transparent conducting oxides: An ab initio study

Abstract

Transparent conducting oxides (TCOs) are essential to many technologies. These materials are doped (n-or p-type) oxides with a large enough band gap (ideally >3 eV) to ensure transparency. However, the high carrier concentration present in TCOs lead additionally to the possibility for optical transitions from the occupied conduction bands to higher states for n-type materials and from lower states to the unoccupied valence bands for p-type TCOs. The "second gap" formed by these transitions might limit transparency and a large second gap has been sometimes proposed as a design criteria for high performance TCOs. Here, we study the influence of this second gap on optical absorption using ab initio computations for several well-known n-and p-type TCOs. Our work demonstrates that most known n-type TCOs do not suffer from second gap absorption in the visible even at very high carrier concentrations. On the contrary, p-type oxides show lowering of their optical transmission for high carrier concentrations due to second gap effects. We link this dissimilarity to the different chemistries involved in n-versus typical p-type TCOs. Quantitatively, we show that second gap effects lead to only moderate loss of transmission (even in p-type TCOs) and suggest that a wide second gap, while beneficial, should not be considered as a needed criteria for a working TCO.

Transparent conducting materials discovery using high-throughput computing

“…This so-called second band gap effect has been recently studied in a series of typical TCOs. 81 This effect can be slightly detrimental for materials with a second band gap that is significantly lower than the optical band gap. The effect has been shown to be not strong enough to prevent materials with smaller second band gaps to be of interest.…”

“…24 Hybrid functionals such as HSE, or the GW method, both yield better estimates of the absorption onset. 24,81 For example, the G 0 W 0 method (single-shot GW) underestimates the band gap by 0.18 eV on average, 69 and HSE06 has an average error of 0.26 eV for typical semiconductors and 0.41 eV for transition-metal compounds. The error on the band gap can even be further decreased if one uses for example the partially selfconsistent GW 0 (typical errors of 3-5%) or the quasi-particle selfconsistent QSGW (~0.1-0.25 eV errors) techniques.…”

“…If the second band gap is large enough, the absorption edge can shift to blue wavelengths. Reprinted from Ha et al, 81…”

Towards bipolar tin monoxide: Revealing unexplored dopants

“…1, result in moderate effective masses for charge carriers of either flavor [3]. In addition, a large gap of about 4 eV between the lowest and the second lowest unoccupied bands in SnO prevents strong degradation of transparency due to free-carrier absorption, when electron concentrations are increased [20]. All of the above characteristics hint towards an excellent ambipolar TCO candidate.…”

Wide Bandgap Perovskite Oxides with High Room‐Temperature Electron Mobility

“…9,77 Theory predicts BaSnO 3 transmission in the visible range to be independent of doping. 81 SrSnO 3 , another candidate in the family of alkaline-earth stannates, has a wider indirect bandgap with experimental and theoretical values ranging between 4.0 eV -5.0 eV. [82][83][84][85][86] The wider bandgap in SrSnO 3 is argued to be a result of smaller Sn-O-Sn bond angle (< 180 ° as in BaSnO 3 ), which causes a larger overlap between Sn 5s and O 2p orbitals.…”