Adam Hultqvist

Inline Cu(In,Ga)Se

In this paper, co-evaporation of Cu(In,Ga)Se2 (CIGS) in an inline single-stage process is used to fabricate solar cell devices with up to 18.6% conversion efficiency using a CdS buffer layer and 18.2% using a Zn1-xSnxOy Cd-free buffer layer. Furthermore, a 15.6-cm2 mini-module, with 16.8% conversion efficiency, has been made with the same layer structure as the CdS baseline cells, showing that the uniformity is excellent. The cell results have been externally verified. The CIGS process is described in detail, and material characterization methods show that the CIGS layer exhibits a linear grading in the [Ga]/([Ga]+[In]) ratio, with an average [Ga]/([Ga]+[In]) value of 0.45. Standard processes for CdS as well as Cd-free alternative buffer layers are evaluated, and descriptions of the baseline process for the preparation of all other steps in the Ångström Solar Center standard solar cell are given.

_{2}

To improve the conduction band alignment and explore the influence of the buffer-absorber interface, we here investigate an alternative buffer for Cu2ZnSnS4 (CZTS) solar cells. The Zn(O, S) system was chosen since the optimum conduction band alignment with CZTS is predicted to be achievable, by varying oxygen to sulfur ratio. Several sulfur to oxygen ratios were evaluated to find an appropriate conduction band offset. There is a clear trend in open-circuit voltage (Voc), with the highest values for the most sulfur rich buffer, before going to the blocking ZnS, whereas the fill factor peaks at a lower S content. The best alternative buffer cell in this series had an efficiency of 4.6% and the best CdS reference gave 7.3%. Extrapolating Voc values to 0 K gave activation energies well below the expected bandgap of 1.5 eV for CZTS, which indicate that recombination at the interface is dominating. However, it is clear that the values are affected by the change of buffer composition and that increasing sulfur content of the Zn(O, S) increases the activation energy for recombination. A series with varying CdS buffer thickness showed the expected behavior for short wavelengths in quantum efficiency measurements but the final variation in efficiency was small.

Co-evaporation for High-Efficiency Solar Cells and Modules

Zinc tin oxide (Zn(1-x)Sn(x)O(y)) has been proposed as an alternative buffer layer material to the toxic, and light narrow-bandgap CdS layer in CuIn(1-x),Ga(x)Se(2) thin film solar cell modules. In this present study, synchrotron-based soft X-ray absorption and emission spectroscopies have been employed to probe the densities of states of intrinsic ZnO, Zn(1-x)Sn(x)O(y) and SnO(x) thin films grown by atomic layer deposition. A distinct variation in the bandgap is observed with increasing Sn concentration, which has been confirmed independently by combined ellipsometry-reflectometry measurements. These data correlate directly to the open circuit potentials of corresponding solar cells, indicating that the buffer layer composition is associated with a modification of the band discontinuity at the CIGS interface. Resonantly excited emission spectra, which express the admixture of unoccupied O 2p with Zn 3d, 4s, and 4p states, reveal a strong suppression in the hybridization between the O 2p conduction band and the Zn 3d valence band with increasing Sn concentration.

Zn(O, S) Buffer Layers and Thickness Variations of CdS Buffer for Cu

In this paper, Cu(In,Ga)Se2 (CIGS) thin-film solar cells are prepared on nominally alkali-free glass substrates using an in-line CIGS growth process. As compared with, for example, borosilicate glass or quartz, the glass is engineered to have similar thermal expansion coefficient as soda-lime glass (SLG) but with alkali content close to zero. Na is incorporated in the CIGS material using an ex-situ deposited NaF precursor layer evaporated onto the Mo back contact. Several thicknesses of the NaF layer were tested. The results show that there is a process window, between 15 and 22.5 nm NaF, where the solar cell conversion efficiency is comparable with or exceeding that of SLG references. The effect of an NaF layer that is too thin on the solar cell parameters was mainly lowering the open-circuit voltage, which points to a lower effective dopant concentration in the CIGS layer and is also consistent with presented C - V measurements and modeling results. For excessively thick NaF layers, delamination of the CIGS layer occurred. Additional measurements, such as scanning electron microscopy (SEM), secondary ion mass spectrometry, capacitance-voltage analysis (C - V), time-resolved photoluminescence (TRPL), external quantum efficiency (EQE), current-voltage analysis (J-V), and modeling, are presented, and the results are discussed.

_{2}

Na plays an important role in the electrical performance of Cu(In,Ga)Se 2 (CIGS) thin-film solar cells. Traditionally, Na has been introduced during the growth of CIGS by thermal diffusion from the soda-lime glass (SLG) substrate; however, better control of the amount of Na is needed to have a more precise control of growth conditions. The introduction of Na into CIGS was studied in three different ways: from the SLG, from a NaF precursor, and from a Na-doped Mo (MoNa) back contact. The most successful approaches were obtained by using the conventional SLG and the NaF precursor. Different growth temperatures of CIGS were tested in an attempt to diffuse more Na from the MoNa layer.

ZnSnS

The use of Na-free or low Na content glass substrates is observed to enhance the resiliency to potential-induced degradation, as compared with glass substrates with high Na content, such as soda lime glass (SLG). The results from stress tests in this study suggest that degradation caused by a combination of heat and bias across the SLG substrate is linked to increased Na concentration in the CdS and Cu(In,Ga)Se2 (CIGS) layers in CIGS-based solar cells. The degradation during the bias stress is dramatic. The efficiency drops to close to 0% after 50 h of stressing. On the other hand, cells on Na-free and low Na content substrates exhibited virtually no efficiency degradation. The degraded cells showed partial recovery by resting at room temperature without bias; thus, the degradation is nonpermanent and may be due to Na migration and accumulation rather than chemical reaction.

_{4}

In this report, Cu(In,Ga)Se2, CIGS, solar cell devices have been fabricated on nominally alkali free glasses with varying coefficients of thermal expansion (CTE) from 50 to 95 * 10−7/ °C. A layer of NaF deposited on top of the Mo was used to provide Na to the CIGS film. Increasing the glass CTE leads to a change of stress state of the solar cell stack as evidenced by measured changes of stress state of the Mo layer after CIGS deposition. The open circuit voltage, the short circuit current density, and the fill factors, for solar cells made on the various substrates, are all found to increase with CTE to a certain point. The median energy conversion efficiency values for 32 solar cells increases from 14.6% to the lowest CTE glass to 16.5% and 16.6%, respectively, for the two highest CTE glasses, which have CTE values closest to that of the soda lime glass. This is only slightly lower than the 17.0% median of soda lime glass reference devices. We propose a model where an increased defect density in the CIGS...

Solar Cells

The presence of Na in Cu(In,Ga)Se2 layers increases the electrical performance of this type of thin-film solar cell. A detailed comparison of incorporating Na in the CIGS layer by two different methods is performed by evaluating several hundred devices fabricated under similar conditions. The first method is based on the conventionally used Na diffusion from the soda-lime glass substrate, whereas the second method is based on a NaF precursor layer deposited on a Mo-coated alkali-free glass substrate. The sample where Na is introduced by using a NaF precursor layer shows an orientation weighted toward (2 0 4)/(2 2 0) and a net acceptor concentration of 3.4 × 1016 cm-3, while SLG shows a (1 1 2) orientation with a 2.9 × 1016 cm-3 acceptor concentration. Both sample types show close identical elemental depth profiles, morphology, and electrical performance.

Soft X-ray characterization of Zn1−xSnxOy electronic structure for thin film photovoltaics

ZnO-based compounds are of interest as buffer layers in Cu(In,Ga)Se2 (CIGS) solar cells, due to the ability to change the electrical and optical properties of ZnO by addition of other elements. The device structure of a CIGS solar cell is; soda-lime glass/Mo/CIGS/buffer layer/ZnO/ZnO:Al. This contribution treats growth and characterization of Zn1-xMgxO and Zn(O,S) on glass substrates and as buffer layers in CIGS solar cell devices. The ZnO-based compounds are grown by atomic layer deposition at deposition temperatures below 200 °C using metal-organic precursors.

Cu(In,Ga)Se

The compatibility of atomic layer deposition directly onto the mixed halide perovskite formamidinium lead iodide:methylammonium lead bromide (CH(NH2)2, CH3NH3)Pb(I,Br)3 (FAPbI3:MAPbBr3) perovskite films is investigated by exposing the perovskite films to the full or partial atomic layer deposition processes for the electron selective layer candidates ZnO and SnOx. Exposing the samples to the heat, the vacuum, and even the counter reactant of H2O of the atomic layer deposition processes does not appear to alter the perovskite films in terms of crystallinity, but the choice of metal precursor is found to be critical. The Zn precursor Zn(C2H5)2 either by itself or in combination with H2O during the ZnO atomic layer deposition (ALD) process is found to enhance the decomposition of the bulk of the perovskite film into PbI2 without even forming ZnO. In contrast, the Sn precursor Sn(N(CH3)2)4 does not seem to degrade the bulk of the perovskite film, and conformal SnOx films can successfully be grown on top of it using atomic layer deposition. Using this SnOx film as the electron selective layer in inverted perovskite solar cells results in a lower power conversion efficiency of 3.4% than the 8.4% for the reference devices using phenyl-C70-butyric acid methyl ester. However, the devices with SnOx show strong hysteresis and can be pushed to an efficiency of 7.8% after biasing treatments. Still, these cells lacks both open circuit voltage and fill factor compared to the references, especially when thicker SnOx films are used. Upon further investigation, a possible cause of these losses could be that the perovskite/SnOx interface is not ideal and more specifically found to be rich in Sn, O, and halides, which is probably a result of the nucleation during the SnOx growth and which might introduce barriers or alter the band alignment for the transport of charge carriers.