Vapour phase alkali species for Cu(In,Ga)Se2 solar cells

Alkalis are essential in Cu(In,Ga)Se2 absorber layers for efficient solar cells. Current doping
methods rely on solid state diffusion of an alkali through to the absorber layer, e.g. a thin NaF layer
on Mo or NaCl dissolved in a metal precursor ink[1]. The apparent concentration of alkali in the
final absorber is determined by the initial alkali dosing and the use of an interfacial barrier to stop
alkali diffusion from the substrate. Until now the vapor–absorber interface as a source or sink of
alkali doping has been largely ignored. We show that device efficiency improves from 2 to 8% by
gas phase Na adsorption alone. Conversely initial results show that Na can also be desorbed to
the gas phase. Although these efficiencies are lower than those obtained by including Na directly
in the precursor (device efficiency 13.3% [1]), the findings are relevant to all chalcogenide growers
as they show that exact doping, and thus control of device efficiency, is only possible when gas
phase adsorption/desorption processes are controlled.