Dominik M. Berg

The Consequences of Kesterite Equilibria for Efficient Solar Cells

Copper-zinc-tin-chalcogenide kesterites, Cu(2)ZnSnS(4) and Cu(2)ZnSnSe(4) (CZTS(e)) are ideal candidates for the production of thin film solar cells on large scales due to the high natural abundance of all constituents, a tunable direct band gap ranging from 1.0 to 1.5 eV, a large absorption coefficient, and demonstrated power conversion efficiencies close to 10%. However, Sn losses through desorption of SnS(e) from CZTS(e) at elevated temperatures (above 400 °C) impede the thorough control of film composition and film homogeneity. No robust and feasible fabrication process is currently available. Here we show that understanding the formation reaction of the kesterite absorber is the key to control the growth process and to drastically improve the solar cell efficiency. Furthermore, we demonstrate that this knowledge can be used to simplify the four-dimensional parameter space (spanned by the four different elements) to an easy and robust two-dimensional process. Sufficiently high partial pressures of SnS(e) and S(e) (a) prevent the decomposition reaction of the CZTS(e) at elevated temperatures and (b) introduce any missing Sn into a Sn-deficient film. This finding enables us to simplify the precursor to a film containing only Cu and Zn, whereas Sn and S(e) are introduced from the gas phase by a self-regulating process.

In-depth resolved Raman scattering analysis for the identification of secondary phases: Characterization of Cu2ZnSnS4 layers for solar cell applications

This work reports the in-depth resolved Raman scattering analysis with different excitation wavelengths of Cu2ZnSnS4 layers. Secondary phases constitute a central problem in this material, particularly since they cannot be distinguished by x-ray diffraction. Raman spectra measured with 325 nm excitation light after sputtering the layers to different depths show peaks that are not detectable by excitation in the visible. These are identified with Cu3SnS4 modes at the surface region while spectra measured close to the back region show peaks from ZnS and MoS2. Observation of ZnS is enhanced by resonant excitation conditions achieved when working with UV excitation.

Raman analysis of monoclinic Cu2SnS3 thin films

Secondary phases like Cu2SnS3 are major obstacles for kesterite thin film solar cell applications. We prepare Cu2SnS3 using identical annealing conditions as used for the kesterite films. By x-ray diffraction, the crystal structure of Cu2SnS3 was identified as monoclinic. Polarization-dependent Raman investigations allowed the identification of the dominant peaks at 290 cm−1 and 352 cm−1 with the main A′ symmetry vibrational modes from the monoclinic Cu2SnS3 phase. Furthermore, micro-resolved Raman investigations revealed local variations in the spectra that are attributed to a secondary phase (possibly Cu2Sn3S7). This exemplifies the abilities of micro-resolved Raman measurements in the detection of secondary phases.

Route Toward High-Efficiency Single-Phase Cu

Thin-film chalcogenide kesterites Cu2ZnSnS4 and Cu2 ZnSnSe4 (CZTSSe) are promising candidates for the next-generation thin-film solar cells. They exhibit a high natural abundance of Cu, Zn, Sn and S2, a high absorption coefficient, and a tunable direct bandgap between 1.0 and 1.5 eV. A prerequisite for the use of CZTSSe as absorber layers in photovoltaic applications on large scales is a detailed knowledge of the formation reaction. Recently, we have shown that a decomposition/formation equilibrium governs the formation reaction. The presence of Sn(S,Se) during the high-temperature preparation steps is essential to prevent decomposition. This improves the solar cell efficiency from 0.02% to 6.1%. In this paper, we show that the decomposition is universal. Absorbers produced by high-temperature coevaporation and samples produced by low-temperature precursor fabrication followed by annealing in a tube furnace in S or Se atmosphere are compared in order to elucidate that in all cases, the loss of Sn(S,Se) forms a degraded surface region. We demonstrate that the degraded surface of CZTSe absorbers contains grains of ZnSe. These new insights can be used to explain why some of the synthesis routines described in the literature yield much better efficiencies than others.

_{\bf 2}

Electrochemical deposition is shown to be a novel technique to deposit films of N,N′-dibutylperylene-3,4:9,10-bis(dicarboximide) (BuPTCDI) dye that avoids the need for high vacuum or solubilising side chains on the molecule. The technique exploits the higher solubility of the reduced ionic form of the dye over the neutral form. BuPTCDI was chemically reduced to solubilise and then electrochemically oxidised to form a film on various substrates. The properties of the films were investigated by UV/Vis spectroscopy, Photoluminescence, Raman spectroscopy, X-ray diffraction, SEM and photoconductivity showing the successful deposition of the BuPTCDI molecules. The technique was also used to deposit films on interdigitated-electrode substrates enabling measurement of field-effect mobility.

ZnSn(S,Se)

The addition of Ag to Cu-Ga-In precursors for reaction to form (AgCu)(InGa)Se2 has shown benefits including improved adhesion, greater process tolerance and potential for improved device performance. In this study, sequential layer sputtering of Ag-Ga, Cu-Ga, and In targets with different layer sequences is characterized. Ag/(Cu+Ag) and (Ag+Cu)/(Ga+In) ratios were fixed at 0.25 and 0.9, respectively. The most uniform morphology is achieved in Ag-Ga/Cu-In-Ga co-sputtered precursors. No metallic In phase was found in this case, and only the Ag(In,Ga)2 phase was detected. Varying the sputtering sequence for stacked layers results in dissimilar morphologies and structural phases. X-ray diffraction (XRD) analyses reveal that the Ag-Ga and In layers intermix when they are in contact, forming a Ag(In,Ga)2 phase. Incorporating a Cu-Ga layer between the Ag-Ga and In layers prevents the formation of such a phase. Finally, solar cells fabricated from the Cu-Ga/In/Ag-Ga metal precursor sequence showed the highest overall performance.

_{\bf 4}

The precursor reaction process for the formation of Cu(In,Ga)(S,Se)2 (CIGSSe) absorber layers offers great potential for both high efficiency and large scale module production. Nevertheless, long reaction times can create a barrier for cost reduction. In this work, we report on a pathway to a 70 % faster precursor reaction process using a Se-capped metallic precursor. The effects of the precursor reaction time and Se-layer thickness on the through-film Ga- and S-profiles are characterized. These parameters allow control of the band gap profile and the device performance. The process further yields reduced void formation at the Mo/CIGSSe interface and improved adhesion.