Why are kesterite solar cells not 20% efficient?

Although kesterite solar cells show the same range of band gaps as the related chalcopyrites, their efficiencies have so far reached only 10%, compared with 20% for the chalcopyrites. A review of the present literature indicates that several non-ideal recombination channels pose the main problem: (i) recombination at the interface between the kesterite and the CdS buffer. This is very likely due to an unfavourable cliff-like band alignment between the absorber and the buffer. However, for pure selenide absorbers, this recombination path is not dominating, which could be due to a spike-like band alignment at the absorber–buffer interface. (ii) A second major recombination becomes obvious in a photoluminescence maximum well below the band gap, even in record efficiency absorbers. This is either due to a very high density of defects, comparable to the density of states in the band, or to stannite inclusions. In view of the phase diagram, secondary phases are not likely the source of the low energy emission. Only in sulphide kesterite a non-stoichiometric SnS phase could also cause this low energy radiative recombination.

► The recombination paths in kesterite solar cells are reviewed.
► S containing cells are dominated by recombination at the kesterite–CdS interface.
► Pure selenide devices do not show interface recombination.
► Low energy radiative recombination indicates problematic bulk recombination.