Analysis of vertical phase distribution in reactively sputtered zinc oxysulfide thin films

Hydrothermal deposition is emerging as a highly potential route for antimony-based solar cells, in which the Sb2(S,Se)3 is typically in situ grown on a common toxic CdS buffer layer. The narrow band gap of CdS causes a considerable absorption in the short-wavelength region and then lowers the current density of the device. Herein, TiO2 is first evaluated as an alternative Cd-free buffer layer for hydrothermally derived Sb2S3 solar cells. But it suffers from a severely inhomogeneous Sb2S3 coverage, which is effectively eliminated by inserting a Zn(O,S) layer. The surface atom of sulfur in Zn(O,S) uniquely provides a chemical bridge to enable the quasi-epitaxial deposition of Sb2S3 thin film, confirming by both morphology and binding energy analysis using DFT. Then the results of the first-principles calculations also show that Zn(O,S)/Sb2S3 has a more stable structure than TiO2/Sb2S3. The resultant perfect Zn(O,S)/Sb2S3 junction, with a suitable band alignment and excellent interface contact, delivers a remarkably enhanced J SC and V OC for Sb2S3 solar cells. The device efficiency with the TiO2/Zn(O,S) buffer layer boosts from 0.54% to 3.70% compared with the counterpart of TiO2, which has a champion efficiency of Cd-free Sb2S3 solar cells with a structure of ITO/TiO2/Zn(O,S)/Sb2S3/Carbon/Ag by in situ hydrothermal deposition. This work provides a guideline for the hydrothermal deposition of antimony-based films upon a nontoxic buffer layer.

Section: Results and Discussionmentioning

confidence: 97%

Zn(O,S) Buffer Layer for in Situ Hydrothermal Sb₂S₃ Planar Solar Cells

Lin

Guo

Yao

et al. 2021

“…Following the deposition of the CBD-Zn(O,S) buffer layer shown in Figure b, the O 1s spectra show a prominent change near the lower binding energy related to the O-I states. Considering the CBD process, the O-II phase (532.6 eV) stems from the Zn(OH) 2 states and the O-III phase clearly disappears . On the basis of the detection of the In 3d 5/2 and Se 3d 5/2 peaks, the O-I peak can be deconvoluted into two states, namely, O–In (530.6 eV) and O–Zn (531.6 eV), corresponding to In-oxide and Zn-oxide, respectively.…”

Section: Results and Discussionmentioning

confidence: 99%

“…To understand the solar-cell functionality produced by KF PDT, we determined the band alignment of a p–n junction between CIGS and CBD-Zn(O,S). Each E g of the CIGS and CBD-Zn(O 1– x ,S x ) buffer layer was calculated from the relative concentrations of the XPS depth profiles acquired at a specific sputtering time (86 and 1.5 min, respectively) by substituting the x values in the following equations: E g,CIGS = 1.044 + 0.735 x – 0.223(1 – x ), x = Ga/(In + Ga), and E g,Zn(O 1– x ,S x ) = xE g,ZnS + (1 − x ) E g,ZnO – bx (1 – x ). , The E g of CIGS is 1.24 eV, which is considered a reasonable value reflecting the bulk property. The E g of CBD-Zn(O,S) is similar (∼2.6 eV) regardless of the KF PDT process time (see Table S1).…”

Section: Results and Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Evolution of Morphological and Chemical Properties at p–n Junction of Cu(In,Ga)Se₂ Solar Cells with Zn(O,S) Buffer Layer as a Function of KF Postdeposition Treatment Time

Lee

Cho

et al. 2021

Self Cite

We carried out KF postdeposition treatment (PDT) on a Cu(In,Ga)Se 2 (CIGS) layer with a process time varying from 50 to 200 s. The highest CIGS solar-cell efficiency was achieved at a KF PDT process time of 50 s; in this condition, we observed the highest level of K element at the near-surface of the CIGS layer and the perfectly passivated pinholes on the CIGS surface. At process times above 150 s, the oversupplied KF agglomerated into large islands and was subsequently eliminated during the deposition of the chemical bath deposition (CBD)-Zn(O,S) buffer layer owing to the islands' water-soluble characteristics. As a result, the growth mechanism of the CBD-Zn(O,S) layer varied as a function of KF PDT process time. X-ray photoemission spectroscopy (XPS) measurements were used to examine the dependency of the chemical state on the KF PDT process time, and from the results, we formulated a chemical reaction model based on the shift in the elemental binding energy following deposition of the CBD-Zn(O,S) buffer layer. The chemical states of the K−In−Se phase, which have a beneficial effect on the solar-cell performance owing to the formation of durable and improved p−n junctions, are formed only at a KF PDT process time of 50 s. We derived band alignments from the XPS depth profiles by extracting the conduction-and valence-band offsets, and we used opticalpump−THz-probe spectroscopy to measure the ultrafast photocarrier lifetimes related to the defect states following KF PDT. Our key findings can be summarized as follows: (i) photocarrier transport is beneficial at a low barrier height, and (ii) the photocarrier lifetime increases when the K−In−Se phases are formed on the CIGS surface, which allows K + ions to be effectively substituted into Cu vacancies.

“…The linear dependence for the lattice parameters of ZnS x O 1−x alloys on the S-composition is consistent with experimental results. 28 The lattice mismatch (Δa/a ̅ ) between ZnO and ZnS is as large as 19.5%, which causes the difficulty of alloying and affects the electronic properties of ZnS x O 1−x alloys.…”

Section: CDmentioning

confidence: 99%

Interface Engineering of Cu(In,Ga)Se₂ Solar Cells by Optimizing Cd- and Zn-Chalcogenide Alloys as the Buffer Layer

Wang

Lan

Zheng

et al. 2021

The optimization of band alignment at the buffer/ absorber interface is realized by tuning compositions of Cd and Zn chalcogenides as the buffer layer toward high-efficiency Cu(In,Ga)-Se 2 (CIGS) solar cells. Using the special quasi-random structure (SQS) approach, we construct randomly disordered Zn x Cd 1−x S y Se 1−y alloys and ZnS x O 1− x alloys as alternatives to the traditional CdS buffer layer. The compositional dependence of formation energies, lattice parameters, band-gap energies, and band alignments of Zn x Cd 1−x S y Se 1−y and ZnS x O 1−x alloys is investigated by first-principles density functional theory calculations. For quaternary Zn x Cd 1−x S y Se 1−y alloys, we find that the miscibility temperatures and the bandgap bowing coefficients are proportional to the lattice mismatch between the mixing elements. The linear dependence of lattice parameters, trinomial dependence of band-gap energies and band-edge positions on the alloy-composition of Zn x Cd 1−x S y Se 1−y alloys are established. For ZnS x O 1−x alloys, we find the lattice parameters also exhibit a linear dependence on its composition. Because of the large lattice mismatch and the chemical disparity between ZnO and ZnS, the bowing coefficient for the bandgap energies of ZnS x O 1−x alloys is composition dependent, and is larger for dilute ZnS x O 1−x alloys. With the optimization criteria of moderate spike-like conduction band offset, large valance band offset, sufficiently wide bandgap, and lattice match with respect to the CIGS absorber, we illustrate the optimal composition range of both Zn x Cd 1−x S y Se 1−y alloys and ZnS x O 1−x alloys as the buffer layer of the CIGS solar cells. Our work demonstrates that Zn x Cd 1−x S y Se 1−y alloys and ZnS x O 1−x alloys are promising buffer layers for high-efficiency CIGS solar cells.