Conductivity and transparency are two properties mutually exclusive in nature. Thus, transparent conductors are complex engineered systems that required fine levels of defect control and tuning.
The basic requirements for TCOs have been covered by many authors but were originally presented by Hamberg and Granqvist: The first requirement is a host oxide with a band gap in excess of 3.1 eV. Smaller band-gap hosts can be used, relying upon the well-known Burstein-Moss shift with doping to increase to the effective optical gap to greater than 3.1 eV. Second, there can be no interband transitions less than 3.1 eV in energy. This limits consideration to cations with filled d-shells, such as 3d10 Cu+, Zn2+ and Ga3+ and 4d10 Ag+ , Cd2+ , In3+ and Sn4+. These criteria ensures visible transparency, but not conductivity. The third TCO requirement is the ability to degenerately dope the oxide hosts with carrier contents in excess of 1020 cm-3. But carrier content alone does not guarantee high electrical conductivity. The final requirement is a highly dispersed conduction band (for n-types) or valence band (for p-types), leading to high electron or hole mobilities.
For the purpose of achieving high efficiency heterojunction solar cells with cumulative series resistance loses < 1.5 Ohms.cm2 the sheet resistance of the front TCO needs to be below 45 Ohms/sq. The typical approach use to achieve these values is to vary the doping concentration of the TCO, thus tuning the electron concentration. The ultimate problem with this approach is the free-carrier absorption that rises in the near IR product of the high electron concentration. Based on this one can conclude that the best parameter to adjust is mobility, however, this has proven the most difficult parameter to control since 1902 when the first TCO was synthesized.