Defect limitations in Cu2ZnSn(S, Se)4 solar cells utilizing an In2S3 buffer layer

Abstract

Alternative n-type buffer layer such as In 2 S 3 have been proposed as Cd-free alternative in kesterite Cu 2 ZnSn(S,Se) 4 (CZTSSe) solar cells. In this study, optical and electronic characterisation techniques together with device analysis and simulation were used to assess nanoparticle-based CZTSSe absorbers and solar cells with CdS and In 2 S 3 buffers. Photoluminescence spectroscopy indicated CZTSSe absorbers with In 2 S 3 buffer had a lower density of detrimental non-radiative defects and a higher concentration of copper vacancies V Cu + , responsible for p-type conductivity in CZTSSe, in comparison to the absorber with CdS buffer. Capacitance-voltage (C-V) measurements revealed the In 2 S 3 buffer-based CZTSSe devices had a three times higher apparent doping density and a consequently narrower space charge region than devices with a CdS layer. This resulted in poorer collection of photo-generated charge carriers in the near-IR region despite a more favourable band alignment as determined by X-ray photoelectron and inverse photoelectron spectroscopy. The presence of interfacial defect states in In 2 S 3 devices as determined by C-V and biased quantum efficiency measurements are also responsible for the loss in open-circuit voltage compared with reference devices with CdS.

A Thin In2S3 Interfacial Layer for Reducing Defects and Roughness of Cu2ZnSn(S,Se)4 Thin‐Film Solar Cells

“…These IS results is in good agreement with dark J–V and EQE results. Therefore, it can be concluded that this photo‐doping effect suggests that the In 2 S 3 layer, as an additional buffer layer, forms a high‐energy barrier hindering the interfacial charge transfer if it becomes too thick (e. g., over 20 nm) [17,45] . A brief band diagrams of CZTSSe/In 2 S 3 /CdS is presented for showing the photo‐doping effect (Figure S10).…”

“…The highest certified and reported efficiencies of CZTSSe thin‐film solar cells are 13.0 % (NREL efficiency chart) and 12.6 %, [3,12] respectively; however, this value is much lower than the theoretical efficiency of CZTS‐based single junction cells [i. e., 32.4 % with short‐circuit current density ( J SC )=29.6 mA ⋅ cm −2 , open‐circuit voltage ( V OC )=1.21 V, and fill factor (FF)=89.9 %] calculated using the Shockley–Queisser efficiency limit [13] . In terms of efficiency improvement, CZTSSe thin‐film solar cells mainly suffer from large V OC deficits ( E g / q – V OC ; E g is the band gap energy and q is the electron charge) originating from the severe non‐radiative recombination in the bulk and at the charge separation interfaces [14–18] . In particular, defects such as antisite defects (e. g., Cu Zn , Cu Sn , Sn Cu , and Sn Zn ), vacancies (e. g., V Zn , V Sn , and V Cu ), and defect clusters (e. g., 2Cu Zn +Sn Zn ) along with the secondary phase can easily be created in the CZTSSe bulk and at the interface because of the narrow phase stability region of the pure CZTSSe in the thermodynamic phase diagram [18–20] .…”

“…[13] In terms of efficiency improvement, CZTSSe thin-film solar cells mainly suffer from large V OC deficits (E g /q-V OC ; E g is the band gap energy and q is the electron charge) originating from the severe non-radiative recombination in the bulk and at the charge separation interfaces. [14][15][16][17][18] In particular, defects such as antisite defects (e. g., Cu Zn , Cu Sn , Sn Cu , and Sn Zn ), vacancies (e. g., V Zn , V Sn , and V Cu ), and defect clusters (e. g., 2Cu Zn + Sn Zn ) along with the secondary phase can easily be created in the CZTSSe bulk and at the interface because of the narrow phase stability region of the pure CZTSSe in the thermodynamic phase diagram. [18][19][20] These defects cause serious non-radiative recombination, leading to significant decreases in V OC and FF.…”

“…[18][19][20] These defects cause serious non-radiative recombination, leading to significant decreases in V OC and FF. [17] Therefore, numerous attempts have been made to enhance V OC and the efficiency through various processing methods, including postheat treatment, [2,5] bandgap engineering, [21][22] alloying/doping of (alkali) metal elements, [20,[23][24][25] and adoption of an intermediate layer. [26] However, despite these efforts, conversion efficiencies greater than 12 % are rare in CZTSSe thin-film solar cells; [3] thus, extensive research is needed to improve the device performance, especially through the passivation of deep-level defects, which act as major non-radiative recombination sites.…”

“…In particular, defects such as antisite defects (e. g., Cu Zn , Cu Sn , Sn Cu , and Sn Zn ), vacancies (e. g., V Zn , V Sn , and V Cu ), and defect clusters (e. g., 2Cu Zn +Sn Zn ) along with the secondary phase can easily be created in the CZTSSe bulk and at the interface because of the narrow phase stability region of the pure CZTSSe in the thermodynamic phase diagram [18–20] . These defects cause serious non‐radiative recombination, leading to significant decreases in V OC and FF [17] . Therefore, numerous attempts have been made to enhance V OC and the efficiency through various processing methods, including post‐heat treatment, [2,5] bandgap engineering, [21–22] alloying/doping of (alkali) metal elements, [20,23–25] and adoption of an intermediate layer [26] .…”