Efficiency limitations of polycrystalline thin film solar cells: case of Cu(In,Ga)Se2

Small-area Cu(In,Ga)Se2 thin film solar cells have reached more than 19% efficiencies. Compared to other polycrystalline materials this efficiency value is remarkable. Nevertheless, the 19% for Cu(In,Ga)Se2 range more than 6% (absolute) below the world's best single-crystalline Si cells and almost 14% below the upper theoretical limit of 33% for an ideal black body cell with infinitely large mobility and radiative recombination only. About 4% out of the 14% are of optical nature, additional 3% stem from the limited mobility/diffusion length and from band gap fluctuations with a standard deviation σg≈50 meV due to spatial variations of composition and stoichiometry of the quaternary compound Cu(In,Ga)Se2. Thus, about 26% efficiency would be possible if there were only these band gap fluctuations. Additional, voltage-dependent electrostatic potential fluctuations push down the efficiency further to 19%: The polycrystalline Cu(In,Ga)Se2 which is unavoidably structurally inhomogeneous due to dislocations, grain boundaries, point defects, etc. is also electrostatically inhomogeneous because of charged defects. Electrostatic potential fluctuations at the valence and conduction band edge may be not only responsible for a high saturation current density but also for the ideality factor in the current/voltage curve. The band gap and electrostatic potential fluctuations make the effective band gap which controls the intrinsic carrier density smaller than the average optical gap. The (zero bias) electrostatic potential fluctuations are here derived from the ideality factors of the current/voltage curve. The ideality factor reflects the voltage-induced electrostatic homogenization of the sample. For the world's best Cu(In,Ga)Se2 cells with an ideality factor of nid=1.5, we estimate zero bias electrostatic potential fluctuations with a standard deviation σelec≈140 meV.