Solid-State NMR and Raman Spectroscopy To Address the Local Structure of Defects and the Tricky Issue of the Cu/Zn Disorder in Cu-Poor, Zn-Rich CZTS Materials

The material Cu2ZnSn(S,Se)(4) (CZTS) offers a promising indium-free alternative to Cu(In,Ga)Se-2 for the absorber layer in thin-film solar cells. It is known that the highest solar energy conversion efficiencies are reached for Cu-poor, Zn-rich CZTS compositions and that too much disorder at the Cu and Zn sites can have a negative impact on the device performance. In this article, we investigate the structures of [V-Cu + Zn-Cu] A-type and [2Zn(Cu) + Zn-Sn] B-type defect complexes and their impact on the long-range Cu/Zn disorder. To that end, we use Sn-119, Cu-65, and Zn-67 solid-state NMR and Raman spectroscopy to characterize powdered CZTS samples. For both A- and B-type substitutions, our NMR investigations demonstrate the clustering of the complexes. Moreover, we show that (A+B)-type compounds should be considered as A-type and B-type compounds, since no interaction seems to exist between [V-Cu + Zn-Cu] and [2Zn(Cu) + Zn-Sn] defect complexes. In addition, it is worth noting that [2Zn(Cu) + Zn-Sn] complexes have only a minor impact on the level of disorder at the Cu and Zn sites. In contrast, [V-Cu + Zn-Cu] complexes seem to restrain the random distribution of Cu at the Zn site and of Zn at the Cu site; i.e., the long-range Cu/Zn disorder. Raman characterization of the CZTS samples was also conducted. The Q = I-287/I-303 and the newly introduced Q' = I-338/(I-366 + I-374) ratios determined from Raman spectra collected at 785 nm turn out to be very sensitive to the level of Cu/Zn disorder. Moreover, they can be used to differentiate the nature of the substitution in slow-cooled materials.