Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2025 Oct 8;8(20):15258–15267. doi: 10.1021/acsaem.5c02266

From Solutions to Photovoltaic Devices: Assessing the Impact of Zn Complexation for Bifacial Cu2ZnSn(S,Se)4 Thin Films

Alice Sheppard †,‡, Jacques Kenyon §, Nada Benhaddou §, Lila Mahmoud †, Alexander W Black †, Jake W Bowers §, David J Fermin †,*
PMCID: PMC12569970  PMID: 41169693

Abstract

Solution-based deposition of high-quality inorganic compound semiconductors onto a variety of substrates is a key challenge toward integrating photovoltaic technologies into a wide range of infrastructures. Cu2ZnSn­(S,Se)4 (CZTSSe) is one of the most promising solar absorbers processable by solution-based methods, however there is a substantial knowledge gap linking the chemistry of cation complexes and chalcogen precursors, e.g. thiourea (TU), to the microstructure and opto-electronic properties of the thin-films. In this study, we focus our attention on the complexation of zinc chloride (ZnCl2) and zinc acetate (ZnAc2) in dimethylformamide (DMF)-based CZTS precursor inks and how this ultimately affects the performance of CZTSSe thin-film devices on F/SnO2 (FTO) substrates. Acetate coordination not only improves the overall CZTSSe composition and structural uniformity but also lowers the zinc salt decomposition temperature, from 519 to 284 °C, which significantly affects the rate of grain growth during selenization and final microstructure. ZnAc2-based CZTSSe films show densely packed grain growth due to the fast rate of selenium incorporation during annealing, while ZnCl2-based CZTSSe displays slower rates due to the high decomposition temperature of ZnCl2. Upon the incorporation of 25 nm Mo at the CZTSSe/FTO interface, a champion power conversion efficiency of 6.02% was achieved with ZnAc2 precursor salt, over two times greater than the equivalent device architecture prepared with ZnCl2, at 2.62%. This investigation illustrates the significant role of the molecular complexes in tuning the grain growth kinetics and final microstructure and therefore improving device performance on semitransparent substrates.

Keywords: photovoltaic devices, thin-film devices, selenium incorporation, dimethylformamide

graphic file with name ae5c02266_0011.jpg

graphic file with name ae5c02266_0009.jpg

Introduction

The decarbonization of electricity generation is pivotal to meet legally binding commitments to net carbon zero by 2050. A key area of industrial growth is the implementation of building-integrated photovoltaic (BIPV) modules, typically semitransparent, flexible, and bifacial applications. A promising thin-film PV material for such technologies is Cu2ZnSn­(S,Se)4 (CZTSSe) due to its earth-abundant base elements, low cost, and ability to be deposited on a wide range of substrates. −

Since 2024, several studies have reported solution-processed CZTSSe devices with certified power conversion efficiencies (PCEs) above 14%, reigniting efforts toward developing efficient, stable, scalable, and cost-effective PV technology. − Solution-processing commonly involves precursor inks based on Cu, Sn and Zn metal salts in solvents such as dimethylformamide (DMF) and dimethyl sulfoxide (DMSO), in the presence of thiourea (TU) acting sulfur precursor and complexing agent. Significant improvements to CZTSSe efficiencies have been brought about by tuning the properties in the molecular precursor ink, including the choice of solvent, Cu–Sn oxidation states, cation alloying, or extrinsic doping. ,− The solution processing of the thin film involves sequential spin coating and drying steps of inks onto Mo-coated glass, generating a dry Cu2ZnSnS4 (CZTS) precursor film, which is subsequently selenized via reactive annealing. This raises a key fundamental question, how solvation and complexation chemistry in the inks can have an impact on the elemental ordering of polycrystalline CZTSSe annealed at temperatures in the range of 550 °C? We have recently shown that molecular complexes in the ink assemble into submicron-size aggregates in solution, which affect the microstructure of dry precursor, the morphology of the polycrystalline thin-film, and its PV performance. We have also demonstrated how blending various proportions of DMF and isopropanol (IPA) directly affected Sn distribution, local work function landscape, and device performance of F/SnO2 (FTO)-based CZTSSe.

In addition to the nature of the solvent, the chemistry of the metal salts has also been linked to significant changes in the PV performance. Common counterions for the metal precursor include chloride (Cl–), acetate (Ac–), and nitrate. − TU stabilizes Cu+ in solution, forming Cu­(TU)1–4Cl through S, − while ZnCl2(TU)2, ZnAc2(TU)2, and SnCl2(TU) complexes have also been reported in the literature. , Instead of complexing with TU, SnCl4 coordinates with DMF or DMSO, forming SnCl4(DMF)2 and SnCl4(DMSO)2, respectively. , Trifiletti et al. proposed that metal acetates form an intermetallic network between cation centers linked by TU in DMSO, generating a sol–gel with homogeneous distribution of metals upon deposition. Ahmad et al. compared metal chloride and acetate precursors in the synthesis of Cu2ZnSnS4 nanoparticles (NPs) at 250 °C. This study concluded that there was negligible incorporation of Zn into the NPs for chloride-based metal precursors due to the strong adsorption of Cl– ions onto ZnS NPs, whereas acetate-based metal salts lead to the formation of Cu2ZnSnS4 NPs. A recent work by Moser et al. concluded that Sn loss during the reactive annealing was mitigated by substituting zinc chloride (ZnCl2) for zinc acetate (ZnAc2). These observations appear consistent with the fact that high-efficiency CZTSSe devices often use CuCl, SnCl4, and ZnAc2 as precursors. − However, a link between the coordination chemistry of the precursor salts in solution and the PV performance is fundamentally missing.

In this work, we provide the first in-depth analysis of the link between complexation in molecular precursors featuring ZnCl2 and ZnAc2, their thermochemistry, and their thin-film microstructure, composition, and PV device performance. Considering the applications of CZTSSe in BIPV, FTO substrates have been utilized in this work. The chemical nature of the Zn salt has clear effects on cation complexation and thermochemical properties of the precursor inks, which are linked to stark differences in grain growth kinetics and CZTSSe film microstructure. Interestingly, these differences have little effect on the surface composition and electronic landscape as probed by XPS and characterized by energy-filtered photoemission electron microscopy (EF-PEEM). These observations suggest that the nature of the Zn counterions has an influence on the bulk properties of the materials. We also observed that the deposition of a 25 nm Mo layer at the FTO surface leads to a significant improvement in device PCE from 2.62% to 6.02%, which further emphasizes the complexity of nucleating high-quality CZTSSe.

Results and Discussion

Complexation with Chloride and Acetate Zinc­(II) Precursor Counterions

The precursor solutions are prepared with CuCl2, SnCl2, either ZnCl2 or ZnAc2, and TU salts dissolved in a 75:25 ratio of dimethylformamide/isopropanol (DMF/IPA) (Supporting InformationExperimental Methods) and a TU/(Cu+Zn+Sn) of 5, as developed from our previous study for CZTSSe on an FTO substrate. Figure displays the Fourier transform infrared (FTIR) spectra of individual metal salts with TU (TU+MX2, where M is the metal ion and X is the precursor counterion −Cl or Ac) as well as the complete molecular precursor solutions (ZnX2-pre). The various IR modes of DMF, TU, and IPA are also summarized in Table S1. , Figure a shows FTIR spectra over the fingerprint region (600–1800 cm–1), indicating that the main changes in complexation occur through the TU CS bond at 740 cm–1. With the addition of CuCl2·2H2O, the appearance of the peak at 720 cm–1 and decreased intensity of free TU at 740 cm–1 indicate the formation of Cu-TU and Lewis acid–base interaction. This bathochromic shift is due to a decrease in the CS bond order. Furthermore, no significant changes in the TU H–N–H bending mode at 1615 cm–1 are observed, confirming that TU–metal complexation takes place via sulfur bonds. In Figure b, no change in free TU peak is detected with the addition of SnCl2, showing evidence that SnCl2 is not complexed by TU. Assessing TU+ZnX2 in Figure b shows a small change in the relative intensity between free TU CS at 740 cm–1 and the emergence of a peak at 720 cm–1. Compared with TU+CuCl2, there is only a slight decrease in TU intensity, indicating that most TU is uncoordinated in solution. The shoulder peak at 720 cm–1 may be associated with ZnCl2(TU)2 or ZnAc2(TU)2 complexes, as reported in the literature. ,,− This may allow additional coordination environments to be present, as discussed below.

1.

1

Open in a new tab

FTIR spectroscopy analysis, between 600 and 1800 cm–1, of the individual metal salts with TU (TU+MX2) and the complete molecular precursor solutions prepared with ZnCl2 (ZnCl2-pre) and ZnAc2 (ZnAc2-pre) precursor salts (a). TU CS stretching in DMF/IPA solvent mixture in presence of each individual metal salts (TU +MX2) and the complete precursor solutions (ZnX2-pre) (b). TU, M–TU, and SnCl4(DMF)2 peaks are labeled at 740, 720, and 693 cm–1, respectively.

Yang et al. characterized the coordination of ZnAc2 in water based on the wavenumber spacing between the symmetric (ν s) and asymmetric stretching (ν a) of the carboxyl-group in FTIR spectra (Δν a–s). Δν a–s can be related to various coordination environments, one of which is bidentate bridging coordination, which describes where the two acetate oxygen atoms bond to two different metal centers. According to Trifiletti et al., bidentate bridging coordination of acetate ions is preferred and is solvent-independent. Due to this, we propose that acetate ions will have the same bridging coordination environment in DMF as described in the literature. This additional Zn coordination environment can directly impact the speciation with free TU, CuCl­(TU)3, and Sn in the sol–gel. An additional peak at 692 cm–1 can be observed for ZnCl2-pre, which correlates to the SnCl4(DMF)2 complex. The existence of this SnCl4(DMF)2 complex in ZnCl2-pre and not in ZnAc2-pre implies that Sn is more freely able to coordinate with DMF without bridged acetate counterions present in the solution. This suggests that ZnAc2 bidentate bridging coordination facilitates an interaction between Sn metal centers, CuCl­(TU)3, and ZnAc2(TU)2. This is an important observation and has an impact on the compositional inhomogeneity in the CZTS precursor films, as discussed further below.

Based on our findings, as well as those reported in the literature, we propose the complexes formed in the ZnCl2- and ZnAc2-based precursor solutions, which are displayed in Figure . While CuCl­(TU)3, ZnCl2(TU)2, and ZnAc2(TU)2 are tetrahedrally coordinated (Figure a–c), ,,, SnCl4(DMF)2 show an octahedral coordination (Figure e). Figure d shows the proposed structure of ZnAc2 complex featuring bidentate bridging coordination, which is in stark contrast to the ZnCl2 complex in which the Cl– groups remain strongly bound.

2.

2

Open in a new tab

Proposed structure of CuCl­(TU)3 (a), ZnCl2(TU)2 (b), ZnAc2(TU)2 (c), ZnAc2-bridged coordination (d), and SnCl4(DMF)2 (e) complexes. ,,,,

Thermogravimetric analysis (TGA) responses of the dried ZnCl2- and ZnAc2-based precursor solutions (methodology described in Experimental Details) are shown in Figure , with the TGA of individual precursor salts displayed in Figure S1 and summarized in Table S2. Three general regimes can be identified in the thermal processing of CZTS precursors: (1) dehydration of complexes from room temperature to 180 °C; (2) decomposition of TU-M complexes and metal sulfide formation between 200 and 270 °C, and (3) metal oxide formation from 275 °C. According to the trend in Figure , DMF, TU, and TU-M complexes decompose from room temperature to 250 °C (see also Figure S1e). From 250 to 550 °C, chloride and acetate impurities are removed. , For the dried precursor salts, there are additional mass losses that cannot be associated solely with the removal of organic complexes and impurities, at ∼285 and ∼510 °C for ZnAc2-pre and ZnCl2-pre, respectively. When assessing the decomposition of the individual salts (Figure S1 and Table S2), ZnAc2 decomposes at a lower temperature (284 °C) than ZnCl2 (519 °C). ZnCl2 and ZnAc2 decompose into ZnO, − which accounts for the 7% and 6% residue in Figures S1c,d. The additional mass losses in the dried precursor solutions can therefore be related to the decomposition of ZnAc2 and ZnCl2. These results are consistent with FTIR spectroscopy, as most of the ZnCl2 and ZnAc2 salts do not complex with TU (Figure b). During spin coating, a hot plate temperature of 350 °C was used, which is lower than the decomposition temperature of ZnCl2 but higher than of ZnAc2. This key difference is likely to affect the CZTSSe film formation during reactive annealing. No additional decomposition peaks can be directly associated with CuCl2·2H2O and SnCl2 salts in either ZnCl2-pre or ZnAc2-pre (Figures S1a,b). The mass loss at 600–650 °C can be associated with the removal of SnS and S x . Interestingly, SnS loss occurs at slightly higher temperatures in the case of ZnAc2-pre, which provides further evidence of the stronger interaction of Sn complexes with other cations in this formulation.

3.

3

Open in a new tab

TGA profiles of the dry precursors ZnCl2-pre and ZnAc2-pre (a). DMF, H2O, and TU decomposition is shaded in purple, chloride and acetate decomposition is shaded in red, SnS and S x are shaded in green, and CZTS and metal sulfide formation is shaded in yellow. Normalized derived mass (Der. Mass) of each precursor solution with respect to temperature (b), with the decomposition of ZnAc2 (red) and ZnCl2 (pink) highlighted.

Precursor Counterion Dependence on Grain Growth Kinetics

Figure shows a collection of top-down and cross-sectional scanning electron microscope (SEM) images obtained at various annealing times at 530 °C of ZnCl2- and ZnAc2-based CZTSSe absorbers. The direct formation from amorphous CZTS to crystalline CZTSSe is demonstrated for both precursor counterions, without evidence of secondary phases, such as Cu2–x (S,Se), Cu2Sn­(S,Se)3 (CTS), and Zn­(S,Se). , In the case of ZnCl2–CZTSSe, grain growth begins after 5 min of selenization (Figure a), where the grain size gradually increases with increasing annealing time. For ZnAc2–CZTSSe, grain growth begins immediately after reaching 530 °C (0 min) (Figure c). Unlike ZnCl2–CZTSSe, grains increase in size up to 5 min, with a low grain packing density, before decreasing from 10 to 30 min of annealing. After 20 min of selenization, films prepared with either Zn counterion reach a compact structure, minimizing the risk of shunting during device fabrication, with ZnCl2–CZTSSe having larger grains than ZnAc2–CZTSSe.

4.

4

Open in a new tab

Top-down (a, c) and cross-sectional (b, d) SEM of ZnCl2- (a, b) and ZnAc2-based (c, d) CZTSSe absorbers on FTO substrates during selenization. The number in each panel represents the annealing time in minutes at 530 °C. The scale bar for each condition is the same as that for the as-deposited (As) image.

Cross-sectional SEM images show that the as-deposited ZnAc2 precursor film is thicker than the ZnCl2 precursor film, at 0.82 and 0.66 μm, respectively, despite using the same spin coating conditions and solution concentration (Supporting InformationExperimental Methods). As shown in Figure a, the residual mass of ZnAc2-pre is greater than that of ZnCl2-pre, at 20% and 14%, respectively, indicating that using a ZnAc2 salt leads to a more thermally stable precursor solution. This higher residual mass may be contributing to the greater overall thickness of the ZnAc2 precursor film after spin coating and drying at 350 °C. Top-down grain growth occurs in both cases, as shown in Figure b,d, initially forming sharp-faceted grains at the top and a fine-grain bottom layer, leading to top-to-bottom grains. Figure S2 illustrates the evolution of the surface morphology at various annealing times, as probed by high-resolution confocal laser scanning microscopy (CLSM). These images identify quantifiable changes in surface roughness at the nanoscale, in a field of view of 130 μm × 130 μm. We can see that the root-mean-square height (Sq) of the films is below 100 nm over the field of view, demonstrating the homogeneous nature of the film deposition. Figure a contrasts the time evolution of Sq, showing an increase from 64 to 116 nm in the case of ZnCl2–CZTSSe through the course of selenization. On the other hand, the ZnAc2–CZTSSe precursor shows a sharp increase in Sq from 46 to 110 nm after 5 min of annealing, followed by a decrease to 52 nm after 30 min of annealing. Based on the evolution of the surface roughness (Figure a) and grain microstructure (Figure ), we propose the grain growth mechanism illustrated in Figure b. The observations suggest that the same ZnCl2–CZTSSe film morphology obtained after 10 min annealing is achieved only under 2 min in the case of ZnAc2–CZTSSe. However, we must stress the fact that this analysis is qualitative, given the complex configuration of the reactive annealing step.

5.

5

Open in a new tab

Surface roughness (a) and grain growth mechanism (b) of ZnCl2- and ZnAc2-based CZTSSe absorbers on FTO substrates during selenization.

We propose that the faster selenization rate of ZnAc2–CZTSSe is linked to the difference in complexation and extent of decomposition of the Zn precursor salts achieved during drying between spin coating steps (Figures , , and S1c,d). As described in the Experimental Methods (Supporting Information), drying steps were carried out at a hot plate temperature of 350 °C. This temperature is significantly higher than the decomposition temperature of ZnAc2 (284 °C), enabling Zn2+ to associate with other cations and sulfur to form a disordered CZTS layer, which is swiftly transformed into high-quality CZTSSe upon annealing at 530 °C. On the other hand, the decomposition temperature of ZnCl2 is 519 °C, which is considerably higher than the drying temperature, hindering the formation of CZTS in the precursor layer and slowing down the selenization process and grain growth.

Figure shows the X-ray diffraction (XRD) patterns and Raman spectra for the ZnCl2- and ZnAc2–CZTSSe absorbers during selenization. Polycrystalline CZTSSe with (112), (204)/(220) and (312)/(116) crystallographic planes at approximately 2θ = 27.3°, 45.3°, and 53.7° (JCPDS 04–019–1847), respectively, can be observed for both absorber conditions after 20 min of selenization. The enlarged image of the (112) diffraction peak (Figure c,d) shows a shift to lower diffraction angles with increasing selenization time. This shift is more gradual for ZnCl2–CZTSSe compared to ZnAc2–CZTSSe. This trend is also observed in the (112) diffraction peak full-width at half-maximum (fwhm) (Table S3). For ZnCl2–CZTSSe, the fwhm gradually decreases from 0.64° to 0.14° from 2 to 20 min, respectively. The fwhm of ZnAc2–CZTSSe sharpens after 5 min of selenization, at 0.16°, which further supports the notion of increased selenization rate when using an acetate-based Zn counterion. Both absorber conditions display a peak at 2θ = 14.5°, which can be associated with SnSe2 and is commonly detected for CZTSSe on FTO substrates. − Raman spectra shown in Figure e,f indicates that the reaction pathway is independent of the zinc precursor counterion, directly transforming from amorphous CZTS to highly crystalline CZTSSe, rather than through binary or ternary phases, shown by the coexistence of low- (Se modes173, 196, 233, and 245 cm–1) and high-frequency Raman peaks (S modes328 cm–1).

6.

6

Open in a new tab

XRD patterns (a–d) and Raman spectra excited with 488 nm laser (e, f) of ZnCl2- (a, c, e) and ZnAc2-based (b, d, f) CZTSSe absorbers on FTO substrates during selenization: As (red), 0 min (orange), 2 min (yellow), 5 min (green), 10 min (light blue), 15 min (blue), 20 min (purple), and 30 min (magenta).

The orbital splitting and binding energy positions of Cu 2p, Zn 2p, and Sn 3d, measured with X-ray photoelectron spectroscopy (XPS), confirmed Cu+, Zn2+, and Sn4+. Using ZnAc2 leads to an increase in Zn/Sn surface composition compared to ZnCl2, at 1.53 and 1.09, respectively (Figure a–c and Table S4). Energy-filtered photoemission electron microscopy (EF-PEEM) analyzed the surface spatial variation of the work function (WF) of the CZTSSe absorber (Figure d–g). The maps show a relatively smooth WF distribution across the field of view for both precursor formulations. In previous studies, we have observed that surface Sn disorder can generate areas of low WF. ,, The distribution of WF in the case of ZnCl2–CZTSSe appears slightly narrower than the ZnAc2–CZTSSe case, with the center shifted by approximately 30 meV toward lower values. The lower WF center could be linked to the lower Zn/Sn ratio, which agrees with our previous study where we observed the WF centre decreasing to values as low as 4.9 eV at Zn/Sn < 1 as a result of changes in DMF/IPA ratio.

7.

7

Open in a new tab

XPS spectra of Cu 2p (a), Zn 2p (b), and Sn 3d (c) and 3D work function (WF) maps (d, f) and corresponding distributions (e, g), measured by EF-PEEM, of ZnCl2- (orange) and ZnAc2–CZTSSe (aqua).

The Effect of Zn Precursor and Substrate Composition in Device Performance

Following the procedure described in the Experimental Methods (Supporting Information), 0.25 cm2 devices were fabricated in the substrate configuration stack SLG/FTO (500 nm)/Mo (0 or 25 nm)/CZTSSe/CdS (50 nm)/i-ZnO (50 nm)/Al/ZnO (500 nm)/Ag (500 nm). In addition to assessing the effect of the Zn precursor, we also investigated the role of a 25 nm Mo layer sputtered on the FTO substrate. Figure S3 shows the morphology of the CZTSSe film and devices prepared with the ZnCl2 and ZnAc2 formulation, without (0) and with 25 nm Mo (25) at the back contact, labeled as 0-ZnCl2, 25-ZnCl2, 0-ZnAc2, and 25-ZnAc2, respectively. Interestingly, the grain size decreases in the presence of 25 nm Mo for both precursor counterions. Analyzing over a larger topographical area and with CLSM, there are inhomogeneities across the surface for ZnCl2-based absorbers which are not present for ZnAc2-based absorbers (Figure S3). Engberg et al. demonstrated the existence of surface patterning while using an all-chloride CZTS molecular ink. These inhomogeneities may be related to the Marangoni effect, which occurs due to variations in the surface tension. The lateral flow of the solution on the surface is dependent on ink properties, such as solute concentration, viscosity, and solvent evaporation rate. The reduction of such patterning for ZnAc2-based absorbers indicates that ZnAc2 stabilizes wettability gradients during drying, possibly due to the bidentate bridging coordination between acetate groups discussed earlier. Figure S3e–h shows the cross-sectional morphology of ZnCl2- and ZnAc2-based devices, indicating the CdS and transparent conducting oxide coatings uniformly cover the CZTSSe absorbers. With the inclusion of 25 nm Mo, a thin fine-grain layer forms at the back contact (Figure S3f,h), suggesting that Mo impacts the CZTSSe grain growth at the rear interface.

The J–V curves for the front-illuminated champion cells are shown in Figure a and summarized in Table S5, with the corresponding external quantum efficiency (EQE) spectra in Figure b. As these devices have nominally two window layers, “front illumination” refers to illumination through the CdS side of the stack. ZnAc2–CZTSSe devices, without and with Mo, show a clear improvement in front illumination PCE compared to ZnCl2–CZTSSe devices, with champion PCEs of 3.27% and 4.49% for 0-ZnCl2 and 0-ZnAc2 devices, respectively. This is further supported by the average cell performance (Figure c and Table S6). The combination of a 25 nm Mo insertion layer and ZnAc2 precursor salt (25-ZnAc2) boosted the PCE to a champion of 6.02%, with J sc of 26.1 mA cm–2, V OC of 0.403 V, and FF of 57.3%. 25-ZnAc2 shows improvements in all PV metrics (Figure c–f and Table S6), particularly V OC and FF, resulting in an average PCE two times greater than that for 0-ZnAc2. The series resistance (R s) is approximately 3 Ω cm2 for all CZTSSe conditions (Figure S4a), likely originating from the FTO substrate. For 25-ZnAc2, the shunt resistance (R sh) improves considerably to an average value of 212 Ω cm2 (Figure S4b), showing a suppression of leakage current and improved carrier collection when combining both ZnAc2 and a 25 nm Mo insertion layer. The improvement of R sh for 25-ZnAc2 greatly reduces the ideality factor from 2.15 for 0-ZnAc2 to 1.63, as shown in Table S5. In addition, Figure S4c compares the champion CZTSSe devices illuminated from the front and back. Both ZnAc2-based devices show improved diode behavior and performance under back illumination compared to ZnCl2-based devices, with PCEs of ∼0.6%. Despite the decrease in J sc due to reduced light transmission, the inclusion of 25 nm Mo enhances the quality of the rear interface of ZnAc2–CZTSSe, shown by improved V OC and FF (Table S7), further illustrating the interplay between the precursor composition and substrate properties. The overall enhancement of device performance metrics with the nanoscale Mo layer on FTO substrates is consistent with previous reports. ,,

8.

8

Open in a new tab

J–V and EQE measurements of champion cells (0.25 cm2) (a, b) for 0-ZnCl2 (orange), 25-ZnCl2 (green), 0-ZnAc2 (aqua), and 25-ZnAc2 (purple) CZTSSe devices. Box plots of PCE (c), J sc (d), V OC (e), and FF (f).

As shown in Figure S3b, the 25-ZnCl2 absorber has severe Marangoni patterning and porosity, shown by the high surface roughness of 209 nm, over 2.5 times greater than 25-ZnAc2 (Figure S3). This high roughness can be associated with localized areas of porous grain growth, leading to shunting pathways to the FTO back contact and reducing R sh to 27.6 ± 9.7 Ω cm2, despite having 25 nm Mo to act as a hole-selective transport layer and improve carrier collection. The overall uniformity is also poor for 0-ZnCl2, to a lesser extent (Figure S3a), indicating that the combination of this layer and the ZnCl2-based precursor solution significantly impedes the absorber microstructure and overall uniformity and therefore the PCE. This demonstrates the importance of tuning the precursor solution through the Zn­(II) precursor counterion to improve overall uniformity, CZTSSe microstructure, and grain growth kinetics.

Finally, the EQE response is shown in Figure b. The higher EQE response between 400 and 500 nm for ZnAc2-based CZTSSe devices indicates an improvement in the CZTSSe/CdS interface. This may be associated with the reduced Sq of ZnAc2-based absorbers compared to ZnCl2-based CZTSSe (Figure S3) that improves CdS growth. Furthermore, carrier collection is improved through the CZTSSe bulk prepared with a ZnAc2 salt. The bandgap for all devices is approximately 1.1 eV, which is consistent with a S/(S+Se) ratio of 0.2.

Conclusions

In this work, we have uncovered the impact of the nature of Zn salt precursor on the solution processing of CZTSSe thin films, from chemical speciation in solution, to grain growth kinetics, film morphology, and photovoltaic performance. Our infrared studies show that most TU does not coordinate with ZnCl2 and ZnAc2, and their decomposition can be followed by thermogravimetric analysis. The considerably lower decomposition temperature of ZnAc2 compared to ZnCl2, at 284 and 519 °C, respectively, directly increases the rate of grain growth during selenization, significantly improving the grain microstructure of ZnAc2–CZTSSe compared to ZnCl2–CZTSSe. Surface patterning in the thin-film morphology, originating from inhomogeneous surface tension (Marangoni effect), is more significant in the ZnCl2 than the ZnAc2 formulation. We linked this macroscopic observation to the bidentate bridging coordination of acetate counterions in solution, which led to improvement in the uniformity of the film composition and microstructure.

Finally, coupling ZnAc2–CZTSSe with a 25 nm Mo insertion layer led to a champion PCE of 6.02% on an FTO substrate, with improvement in all device metrics with respect to the ZnCl2–CZTSSe-based devices. Our analysis clearly shows how the composition of the molecular precursor can have a substantial impact across different length scales, from complexation and interactions with other precursors, to the surface wettability, thermolysis, and crystal growth. These observations are key, not only in the formulation of precursors for thin-film PV, but also in the broader context of solution processing of high specification compound semiconductors.

Supplementary Material

ae5c02266_si_001.pdf (1.1MB, pdf)

Acknowledgments

A.S. acknowledges the financial support by the Engineering and Physical Sciences Research Council (EPSRC, DTA grant EP/T517872/1). A.S. and D.J.F. acknowledge the support from EPSRC through the SolPV grant EP/V008676/1. N.B. and J.W.B. are grateful for the support via EP/V008692/1. The authors acknowledge the use of the University of Bristol NanoESCA II Laboratory. A.S. would like to acknowledge Dr. Yuancai Gong and Dr. Edgardo Saucedo for inspiring this work through a scientific discussion.

Data are available at the University of Bristol data repository, data.bris, at https://doi.org/10.5523/bris.1qflk9pgu1wcz2q62iytqxxjl6.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsaem.5c02266.

  • Experimental Details; summary of FTIR modes of DMF, TU and IPA; TGA measurements of precursor salts; surface roughness analysis of CZTSSe absorbers prepared with ZnCl2- and ZnAc2-based precursor solutions; summary of CZTSSe (112) diffraction peak during selenization; summary of surface composition, WF and WF distribution; SEM and surface roughness images of CZTSSe absorbers without and with a 25 nm Mo layer; summary of photovoltaic parameters for champion devices; summary of mean photovoltaic parameters over a range of devices; and series and shunt resistances of solar cells (PDF)

A.S. wrote the manuscript, designed the experiments, characterized the precursor ink solutions, fabricated and characterized the CZTSSe absorbers, and performed data analysis. J.K., N.B., and J.W.B. contributed to Mo insertion layer deposition, CZTSSe device fabrication, and JV and EQE measurements. L.A.M.M. aided in the acquisition of the TGA measurements. A.W.B. provided scientific discussions and reviewed and edited the manuscript. All authors were involved in the discussions and approved the manuscript. D.J.F. contributed to the manuscript preparation, data analysis, and supervised the research project.

The authors declare no competing financial interest.

References

  1. Nations, U . NDC Registry; United Nations, 2025. https://unfccc.int/NDCREG (accessed February 20, 2025). [Google Scholar]
  2. Peng C.-Y., Dhakal T. P., Garner S., Cimo P., Lu S., Westgate C. R.. Fabrication of Cu2ZnSnS4 solar cell on a flexible glass substrate. Thin Solid Films. 2014;562:574–577. doi: 10.1016/j.tsf.2014.03.054. [DOI] [Google Scholar]
  3. Yang K. J., Kim S., Kim S. Y., Ahn K., Son D. H., Kim S. H., Lee S. J., Kim Y. I., Park S. N., Sung S. J., Kim D. H., Enkhbat T., Kim J., Jeon C. W., Kang J. K.. Flexible Cu2ZnSn­(S,Se)4 solar cells with over 10% efficiency and methods of enlarging the cell area. Nat. Commun. 2019;10(1):2959. doi: 10.1038/s41467-019-10890-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Sun Q., Shi C., Xie W., Li Y., Zhang C., Wu J., Zheng Q., Deng H., Cheng S.. Defect Synergistic Regulations of Li&Na Co-Doped Flexible Cu2ZnSn­(S,Se)4 Solar Cells Achieving over 10% Certified Efficiency. Adv. Sci. 2024;11(6):e2306740. doi: 10.1002/advs.202306740. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Espindola-Rodriguez M., Sylla D., Sánchez Y., Oliva F., Grini S., Neuschitzer M., Vines L., Izquierdo-Roca V., Saucedo E., Placidi M.. Bifacial Kesterite Solar Cells on FTO Substrates. ACS Sustainable Chem. Eng. 2017;5(12):11516–11524. doi: 10.1021/acssuschemeng.7b02797. [DOI] [Google Scholar]
  6. Becerril-Romero I., Sylla D., Placidi M., Sanchez Y., Andrade-Arvizu J., Izquierdo-Roca V., Guc M., Perez-Rodriguez A., Grini S., Vines L., Pusay B., Almache R., Puigdollers J., Pistor P., Saucedo E., Espindola-Rodriguez M.. Transition-Metal Oxides for Kesterite Solar Cells Developed on Transparent Substrates. ACS Appl. Mater. Interfaces. 2020;12(30):33656–33669. doi: 10.1021/acsami.0c06992. [DOI] [PubMed] [Google Scholar]
  7. Xu A., Li X., Liu H., Xiang C., Ma C., Li Y., Pan X., Yan W., Wang S., Huang W., Xin H.. Over 10% Efficiency Pure Sulfide Kesterite Solar Cells on Transparent Electrode with Cd-Ag Co-Alloying. Small. 2025;21(1):e2407075. doi: 10.1002/smll.202407075. [DOI] [PubMed] [Google Scholar]
  8. Zhou Y., Xiang C., Dai Q., Xiang S., Li R., Gong Y., Zhu Q., Yan W., Huang W., Xin H.. 11.4% Efficiency Kesterite Solar Cells on Transparent Electrode. Adv. Energy Mater. 2023;13(19):2300253. doi: 10.1002/aenm.202300253. [DOI] [Google Scholar]
  9. Wei H., Cui C., Li Y., Wu Z., Wei Y., Han Y., Han L., Lu B., Wang X., Pang S., Shao Z., Cui G.. Regulating Hetero-Nucleation Enabling Over 14% Efficient Kesterite Solar Cells. Small. 2024;20(19):e2308266. doi: 10.1002/smll.202308266. [DOI] [PubMed] [Google Scholar]
  10. Gong Y., Jimenez-Arguijo A., Medaille A. G., Moser S., Basak A., Scaffidi R., Carron R., Flandre D., Vermang B., Giraldo S., Xin H., Perez-Rodriguez A., l Saucedo E.. Li-Doping and Ag-Alloying Interplay Shows the Pathway for Kesterite Solar Cells with Efficiency Over 14% Adv. Funct. Mater. 2024;34(42):2404669. doi: 10.1002/adfm.202404669. [DOI] [Google Scholar]
  11. Shi J., Wang J., Meng F., Zhou J., Xu X., Yin K., Lou L., Jiao M., Zhang B., Wu H., Luo Y., Li D., Meng Q.. Multinary alloying for facilitated cation exchange and suppressed defect formation in kesterite solar cells with above 14% certified efficiency. Nat. Energy. 2024;9:1095–1104. doi: 10.1038/s41560-024-01551-5. [DOI] [Google Scholar]
  12. Li Y., Cui C., Wei H., Shao Z., Wu Z., Zhang S., Wang X., Pang S., Cui G.. Suppressing Element Inhomogeneity Enables 14.9% Efficiency CZTSSe Solar Cells. Adv. Mater. 2024;36(25):e2400138. doi: 10.1002/adma.202400138. [DOI] [PubMed] [Google Scholar]
  13. Xie T., Han L., Chu L., Han M., Jian Y., Chi J., Zhong X., Liu T., Kou D., Zhou W., Zhou Z., Yuan S., Meng Y., Qi Y., Wu S.. Passivation of Deep-Level Defects through Chemical Environment Regulation for Efficient CZTSSe Solar Cells Based on Active Selenium Adsorption–Desorption Process. Adv. Funct. Mater. 2025;35:2414940. doi: 10.1002/adfm.202414940. [DOI] [Google Scholar]
  14. Zhong X., Wang J., Han L., Chi J., Liu T., Kou D., Zhou W., Zhou Z., Yuan S., Meng Y., Qi Y., Meng Q., Wu S.. Li, Ag Co-Doping Enables Efficient Kesterite Solar Cell with a High Fill Factor of 74.30% Adv. Funct. Mater. 2025;35:2418548. doi: 10.1002/adfm.202418548. [DOI] [Google Scholar]
  15. Clark J. A., Murray A., Lee J. M., Autrey T. S., Collord A. D., Hillhouse H. W.. Complexation Chemistry in N,N-Dimethylformamide-Based Molecular Inks for Chalcogenide Semiconductors and Photovoltaic Devices. J. Am. Chem. Soc. 2019;141(1):298–308. doi: 10.1021/jacs.8b09966. [DOI] [PubMed] [Google Scholar]
  16. Gong Y., Zhang Y., Jedlicka E., Giridharagopal R., Clark J. A., Yan W., Niu C., Qiu R., Jiang J., Yu S., Wu S., Hillhouse H. W., Ginger D. S., Huang W., Xin H.. Sn4+ precursor enables 12.4% efficient kesterite solar cell from DMSO solution with open circuit voltage deficit below 0.30 V. Sci. China Mater. 2021;64(1):52–60. doi: 10.1007/s40843-020-1408-x. [DOI] [Google Scholar]
  17. Su Z., Sun K., Han Z., Cui H., Liu F., Lai Y., Li J., Hao X., Liu Y., Green M. A.. Fabrication of Cu2ZnSnS4 solar cells with 5.1% efficiency via thermal decomposition and reaction using a non-toxic sol–gel route. J. Mater. Chem. A. 2014;2(2):500–509. doi: 10.1039/C3TA13533K. [DOI] [Google Scholar]
  18. Xin H., Katahara J. K., Braly I. L., Hillhouse H. W.. 8% Efficient Cu2ZnSn­(S,Se)4 Solar Cells from Redox Equilibrated Simple Precursors in DMSO. Adv. Energy Mater. 2014;4(11):1301823. doi: 10.1002/aenm.201301823. [DOI] [Google Scholar]
  19. Tiwari D., Koehler T., Lin X., Harniman R., Griffiths I., Wang L., Cherns D., Klenk R., Fermin D. J.. Cu2ZnSnS4 Thin Films Generated from a Single Solution Based Precursor: The Effect of Na and Sb Doping. Chem. Mater. 2016;28(14):4991–4997. doi: 10.1021/acs.chemmater.6b01499. [DOI] [Google Scholar]
  20. Tiwari D., Cattelan M., Harniman R. L., Sarua A., Fox N., Koehler T., Klenk R., Fermin D. J.. Impact of Sb and Na Doping on the Surface Electronic Landscape of Cu2ZnSnS4 Thin Films. ACS Energy Lett. 2018;3(12):2977–2982. doi: 10.1021/acsenergylett.8b02081. [DOI] [Google Scholar]
  21. Tiwari D., Cattelan M., Harniman R. L., Sarua A., Abbas A., Bowers J. W., Fox N. A., Fermin D. J.. Mapping Shunting Paths at the Surface of Cu2ZnSn­(S,Se)4 Films via Energy-Filtered Photoemission Microscopy. iScience. 2018;9:36–46. doi: 10.1016/j.isci.2018.10.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Cui Y., Wang M., Dong P., Zhang S., Fu J., Fan L., Zhao C., Wu S., Zheng Z.. DMF-Based Large-Grain Spanning Cu2ZnSn­(Sx,Se1‑x)4 Device with a PCE of 11.76. Adv. Sci. 2022;9(20):e2201241. doi: 10.1002/advs.202201241. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Gong Y., Zhang Y., Zhu Q., Zhou Y., Qiu R., Niu C., Yan W., Huang W., Xin H.. Identifying the origin of the Voc deficit of kesterite solar cells from the two grain growth mechanisms induced by Sn2+ and Sn4+ precursors in DMSO solution. Energy Environ. Sci. 2021;14(4):2369–2380. doi: 10.1039/D0EE03702H. [DOI] [Google Scholar]
  24. Romanyuk Y. E., Haass S. G., Giraldo S., Placidi M., Tiwari D., Fermin D. J., Hao X., Xin H., Schnabel T., Kauk-Kuusik M., Pistor P., Lie S., Wong L. H.. Doping and alloying of kesterites. J. Phys.: Energy. 2019;1(4):044004. doi: 10.1088/2515-7655/ab23bc. [DOI] [Google Scholar]
  25. Agbenyeke R., Sheppard A., Keynon J., Benhaddou N., Fleck N., Corsetti V., Alkhalifah M. A., Tiwari D., Bowers J. W., Fermin D. J.. Correlating molecular precursor interactions with device performance in solution-processed Cu2ZnSn­(S,Se)4 thin-film solar cells. ACS Appl. Mater. Interfaces. 2024;16(27):35315–35322. doi: 10.1021/acsami.4c05321. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Sheppard A., Agbenyeke R., Laverock J., King L., Kenyon J., Benhaddou N., Fleck N., Harniman R. L., Sarua A., Tiwari D., Bowers J. W., Fox N. A., Fermin D. J.. Utilizing Solvent Repulsion between Dimethylformamide and Isopropanol to Manipulate Sn Distribution for Bifacial Cu2ZnSn­(S,Se)4 Solar Cells. ACS Appl. Energy Mater. 2024;7(24):11766–11774. doi: 10.1021/acsaem.4c01905. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Cabas-Vidani A., Haass S. G., Andres C., Caballero R., Figi R., Schreiner C., Márquez J. A., Hages C., Unold T., Bleiner D., Tiwari A. N., Romanyuk Y. E.. High-Efficiency (LixCu1–x)2ZnSn­(S,Se)4 Kesterite Solar Cells with Lithium Alloying. Adv. Energy Mater. 2018;8(34):1801191. doi: 10.1002/aenm.201801191. [DOI] [Google Scholar]
  28. Moser S., Tiwari A. N., Carron R.. Interplay between Li and Na amid co-doped solution-processed Cu2ZnSn­(S,Se)4 absorbers for solar cells. Sol. Energy Mater. Sol. Cells. 2023;250:112094. doi: 10.1016/j.solmat.2022.112094. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Moser S., Krummenacher J., Aribia A., Morzy J., Carron R.. Mitigating Sn loss via anion substitution in the Cu2+–Sn2+ precursor system for Cu2ZnSn­(S,Se)4 solar cells. J. Mater. Chem. A. 2024;12(46):32424–32435. doi: 10.1039/D4TA04539D. [DOI] [Google Scholar]
  30. Niu C., Gong Y., Qiu R., Zhu Q., Zhou Y., Hao S., Yan W., Huang W., Xin H.. 11.5% efficient Cu2ZnSn­(S,Se)4 solar cell fabricated from DMF molecular solution. J. Mater. Chem. A. 2021;9(22):12981–12987. doi: 10.1039/D1TA01871J. [DOI] [Google Scholar]
  31. Xu X., Zhou J., Yin K., Wang J., Lou L., Jiao M., Zhang B., Li D., Shi J., Wu H., Luo Y., Meng Q.. Controlling Selenization Equilibrium Enables High-Quality Kesterite Absorbers for Efficient Solar Cells. Nat. Commun. 2023;14(1):6650. doi: 10.1038/s41467-023-42460-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Spofford W. A., Amma E. L.. The crystal structure of bis­(thiourea)­copper­(I) chloride. Acta Crystallogr. 1970;26:1474–1483. doi: 10.1107/S0567740870004351. [DOI] [Google Scholar]
  33. Krzewska S., Podsiadły H.. Studies on the Reaction of Copper­(II) with Thiourea-I. J. Inorg. Nucl. Chem. 1980;42(1):83–85. doi: 10.1016/0022-1902(80)80048-0. [DOI] [Google Scholar]
  34. Bowmaker G. A., Hanna J. V., Pakawatchai C., Skelton B. W., Thanyasirikul Y., White A. H.. Crystal Structures and Vibrational Spectroscopy of Copper­(I) Thiourea Complexes. Inorg. Chem. 2009;48(1):350–368. doi: 10.1021/ic801310r. [DOI] [PubMed] [Google Scholar]
  35. Yamaguchi A., Penland R. B., Mizushima S., Lane T. J., Curran C., Quagliano J. V.. Infrared Absorption Spectra of Inorganic Coordination Complexes. XIV. Infrared Studies of Some Metal Thiourea Complexes1a . J. Am. Chem. Soc. 1958;80(3):527–529. doi: 10.1021/ja01536a005. [DOI] [Google Scholar]
  36. Madarász J., Bombicz P., Okuya M., Kaneko S.. Thermal Decomposition of Thiourea Complexes of Cu­(I), Zn­(II), and Sn­(II) Chlorides as Precursors for the Spray Pyrolysis Deposition of Sulfide Thin Films. Solid State Ionics. 2001;141-142:439–446. doi: 10.1016/S0167-2738(01)00740-8. [DOI] [Google Scholar]
  37. Kannan V., Rajesh N. P., Bairava Ganesh R., Ramasamy P.. Growth and characterization of Bisthiourea-zinc Acetate, a new nonlinear optical material. J. Cryst. Growth. 2004;269(2–4):565–569. doi: 10.1016/j.jcrysgro.2004.05.051. [DOI] [Google Scholar]
  38. Trifiletti V., Mostoni S., Butrichi F., Acciarri M., Binetti S., Scotti R.. Study of Precursor-Inks Designed for High-Quality Cu2ZnSnS4 Films for Low-Cost PV Application. ChemistrySelect. 2019;4(17):4905–4912. doi: 10.1002/slct.201900170. [DOI] [Google Scholar]
  39. Ahmad R., Saddiqi N. U., Wu M., Prato M., Spiecker E., Peukert W., Distaso M.. Effect of the Counteranion on the Formation Pathway of Cu2ZnSnS4 (CZTS) Nanoparticles under Solvothermal Conditions. Inorg. Chem. 2020;59(3):1973–1984. doi: 10.1021/acs.inorgchem.9b03338. [DOI] [PubMed] [Google Scholar]
  40. Shundalau M. B., Chybirai P. S., Komyak A. I., Zazhogin A. P., Ksenofontov M. A., Umreiko D. S.. Modeling of structures and calculation of IR vibrational spectra of N,N-dimethylformamide dimers by density functional theory. J. Appl. Spectrosc. 2011;78:326–336. doi: 10.1007/s10812-011-9466-1. [DOI] [Google Scholar]
  41. Sivakumar N., Kanagathara N., Marchewka M. K., Drozd M., Jayavel R., Anbalagan G.. The theoretical and experimental vibrational studies of thiourea and silver nitrate (2:1) complex. Spectrochim. Acta, Part A. 2018;204:717–725. doi: 10.1016/j.saa.2018.06.035. [DOI] [PubMed] [Google Scholar]
  42. Alía J. M., Edwards H. G. M., Stoev M. D.. A systematic FT-Raman spectroscopic study of twelve bis-thiourea complexes, A­(tu)­2B2 (A = Zn, Cd, Hg; B = Cl, Br, I, SCN) Spectrochim. Acta, Part A. 1999;55(12):2423–2435. doi: 10.1016/S1386-1425(99)00028-1. [DOI] [Google Scholar]
  43. Swaminathan K., Irving H. M. N. H.. Infra-red absorption spectra of complexes of thiourea. J. Inorg. Nucl. Chem. 1964;26(7):1291–1294. doi: 10.1016/0022-1902(64)80212-8. [DOI] [Google Scholar]
  44. Cavalca L., Gasparri G., Andreetti G., Domiano P.. The crystal structure of bisthiourea-zinc acetate. Acta Crystallogr. 1967;22:90–98. doi: 10.1107/S0365110X67000167. [DOI] [Google Scholar]
  45. Rajasekaran R., Ushasree P. M., Jayavel R., Ramasamy P.. Growth and characterization of zinc thiourea chloride (ZTC): a semiorganic nonlinear optical crystal. J. Cryst. Growth. 2001;229:563–567. doi: 10.1016/S0022-0248(01)01229-5. [DOI] [Google Scholar]
  46. Györyová K., Balek V.. Thermal Stability of New Zinc Acetate-based Complex Compounds. J. Therm. Anal. 1993;40:519–532. doi: 10.1007/BF02546621. [DOI] [Google Scholar]
  47. Yang M. M., Crerar D. A., Irish D. E.. A Raman spectroscopic study of lead and zinc acetate complexes in hydrothermal solutions. Geochim. Cosmochim. Acta. 1989;53:319–326. doi: 10.1016/0016-7037(89)90384-0. [DOI] [Google Scholar]
  48. Uhl A. R., Katahara J. K., Hillhouse H. W.. Molecular-ink route to 13.0% efficient low-bandgap CuIn­(S,Se)2 and 14.7% efficient Cu­(In,Ga)­(S,Se)2 solar cells. Energy Environ. Sci. 2016;9(1):130–134. doi: 10.1039/C5EE02870A. [DOI] [Google Scholar]
  49. Mar, G. L. ; Timbrell, P. Y. ; Lamb, R. N. . Formation of Zinc Oxide Thin Films by the Thermal Decomposition of Zinc Acetate; Springer Proceedings, 1993. [Google Scholar]
  50. D F. P., L W. E., N D. R. A., Andrade E., Miki-Yoshida M.. Growth, structure and optical characterization of high quality ZnO thin films obtained by spray pyrolysis. Thin Solid Films. 1999;350:192–202. doi: 10.1016/S0040-6090(99)00050-4. [DOI] [Google Scholar]
  51. Jones F., Tran H., Lindberg D., Zhao L., Hupa M.. Thermal Stability of Zinc Compounds. Energy Fuels. 2013;27(10):5663–5669. doi: 10.1021/ef400505u. [DOI] [Google Scholar]
  52. Clark J. A., Uhl A. R., Martin T. R., Hillhouse H. W.. Evolution of Morphology and Composition during Annealing and Selenization in Solution-Processed Cu2ZnSn­(S,Se)4 . Chem. Mater. 2017;29(21):9328–9339. doi: 10.1021/acs.chemmater.7b03313. [DOI] [Google Scholar]
  53. Qin X., Xu B., Lin J., Chen J., Tong H., Chen Y., Yang P., Chu J., Sun L.. Above 10% efficient electrodeposited Cu2ZnSn­(S,Se)4 solar cell achieved by modifying precursor. Sol. Energy Mater. Sol. Cells. 2022;242:111781. doi: 10.1016/j.solmat.2022.111781. [DOI] [Google Scholar]
  54. Temgoua S., Bodeux R., Mollica F., Naghavi N.. Comparative study of Cu2ZnSnSe4 solar cells growth on transparent conductive oxides and molybdenum substrates. Sol. Energy. 2019;194:121–127. doi: 10.1016/j.solener.2019.10.050. [DOI] [Google Scholar]
  55. Ruiz-Perona A., Palma-Lafuente D., Sánchez Y., Guc M., Kodalle T., Salem M. O., Placidi M., Serna R., Pérez-Rodríguez A., Merino J. M., Caballero R.. Semitransparent Wide Bandgap Cu2ZnGe­(S,Se)4 Thin-Film Solar Cells: Role of the Sulfurization Process. Solar RRL. 2024;8(4):2300947. doi: 10.1002/solr.202300947. [DOI] [Google Scholar]
  56. Liu H., Li Y., Xu A., Li X., Xiang C., Zhou S., Wang S., Yan W., Xin H.. 8.36% Efficient CZTS Solar Cells on Transparent Electrode via Solution Processing. Solar RRL. 2024;8(21):2400588. doi: 10.1002/solr.202400588. [DOI] [Google Scholar]
  57. Moulder, J. F. a. ; S, W. F. ; Sobol, P. E. ; Bomben, K. D. . Handbook of X-ray Photoelectron Spectroscopy; Perkin-Elmer Corp: Eden Prairie, 1992. [Google Scholar]
  58. Tiwari D., Yakushev M. V., Koehler T., Cattelan M., Fox N., Martin R. W., Klenk R., Fermin D. J.. Mapping the Energetics of Defect States in Cu2ZnSnS4 films and the Impact of Sb Doping. ACS Appl. Energy Mater. 2022;5(4):3933–3940. doi: 10.1021/acsaem.1c03729. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Engberg S., Martinho F., Gansukh M., Protti A., Kungas R., Stamate E., Hansen O., Canulescu S., Schou J.. Spin-coated Cu2ZnSnS4 solar cells: A study on the transformation from ink to film. Sci. Rep. 2020;10(1):20749. doi: 10.1038/s41598-020-77592-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Haight R., Barkhouse A., Gunawan O., Shin B., Copel M., Hopstaken M., Mitzi D. B.. Band alignment at the Cu2ZnSn­(SxSe1–x)4/CdS interface. Appl. Phys. Lett. 2011;98(25):253502. doi: 10.1063/1.3600776. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ae5c02266_si_001.pdf (1.1MB, pdf)

Data Availability Statement

Data are available at the University of Bristol data repository, data.bris, at https://doi.org/10.5523/bris.1qflk9pgu1wcz2q62iytqxxjl6.

Articles from ACS Applied Energy Materials are provided here courtesy of American Chemical Society

RESOURCES