Synthesis and characterization of copper zinc tin sulfide nanoparticles and thin films.

Abstract

Copper zinc tin sulfide (Cu2ZnSnS4, or CZTS) is emerging as an alternative material to the present thin film solar cell technologies such as Cu(In,Ga)Se2 and CdTe. All the elements in CZTS are abundant, environmentally benign, and inexpensive. In addition, CZTS has a band gap of ~1.5 eV, the ideal value for converting the maximum amount of energy from the solar spectrum into electricity. CZTS has a high absorption coefficient (>104 cm-1 in the visible region of the electromagnetic spectrum) and only a few micron thick layer of CZTS can absorb all the photons with energies above its band gap. CZT(S,Se) solar cells have already reached power conversion efficiencies >10%. One of the ways to improve upon the CZTS power conversion efficiency is by using CZTS quantum dots as the photoactive material, which can potentially achieve efficiencies greater than the present thin film technologies at a fraction of the cost. However, two requirements for quantum-dot solar cells have yet to be demonstrated. First, no report has shown quantum confinement in CZTS nanocrystals. Second, the syntheses to date have not provided a range of nanocrystal sizes, which is necessary not only for fundamental studies but also for multijunction photovoltaic architectures. We resolved these two issues by demonstrating a simple synthesis of CZTS, Cu2SnS3, and alloyed (Cu2SnS3)x(ZnS)y nanocrystals with diameters ranging from 2 to 7 nm from diethyldithiocarbamate complexes. As-synthesized nanocrystals were characterized using high resolution transmission electron microscopy, X-ray diffraction, Raman spectroscopy, and energy dispersive spectroscopy to confirm their phase purity. Nanocrystals of diameter less than 5 nm were found to exhibit a shift in their optical absorption spectra towards higher energy consistent with quantum confinement and previous theoretical predictions. Thin films from CZTS nanocrystals deposited on Mo-coated quartz substrates using drop casting were found to be continuous but highly porous. Annealing CZTS nanocrystal films at temperatures as low as 400°C led to an intense grain growth; however, thin films from CZTS nanocrystals cracked on annealing due to their high porosity. Although quantum confinement in CZTS is only accessible in nanocrystals of diameters less than 5 nm, the high volume of the ligands as compared to the volume of the nanocrystals makes it a challenge to deposit continuous compacted thin films from small nanocrystals. Films deposited from thermal decomposition of a stoichiometric mix of metal dithiocarbamate complexes were found to be predominantly CZTS. These films from complexes were found to be continuous but microporous. The diameter of the spheres making up the microporous structure could be changed by changing the anneal temperature. The structural composition of the final film could be altered by changing the heating rate of the complexes. CZTS exists in three different crystal structures: kesterite, stannite, and pre-mixed Cu-Au (PMCA) structures. Due to the similarity in the crystal structures, it is extremely difficult to distinguish them based on X-ray diffraction. We computed the phonon dispersion curves for the three structures using ab-initio calculations, and found characteristic discontinuities at the Γ-point which can potentially be used to distinguish the three. In addition, the Γ-point phonon frequencies, which correspond to the Raman peak positions, for the three structures were found to be shifted from each other by a few wavenumbers. By deconvoluting the experimental Raman spectra for both CZTS and Cu2ZnSnSe4 (CZTSe) using Gaussian peaks, we observed that the most intense Raman scattering peak in both CZTS and CZTSe is a sum of two different peaks which correspond to scattering from their respective kesterite and stannite phases. The electronic, structural, and vibrational properties of a series of CZTS-CZTSe alloys (CZTSSe) were studied using ab-initio calculations. The S-to-Se ratio and the spatial distribution of the anions in the unit cell were found to determine the energy splitting between the electronic states at the top of the valence band and the hole mobility in CZTSSe alloys and solar cells. X-ray diffraction patterns and phonon distribution curves were found to be sensitive to the local anion ordering. The predicted Raman scattering frequencies and their variation with x agree with experimentally determined values and trends.