Optical properties of reactively sputtered Cu2ZnSnS4 solar absorbers determined by spectroscopic ellipsometry and spectrophotometry

https://doi.org/10.1016/j.solmat.2016.01.014Get rights and content

Highlights

  • The optical constants of CZTS are determined by spectroscopic ellipsometry (SE).

  • The absorption coefficients (α) yielded by SE and spectrophotometry are in agreement.

  • The band gap (Eg) determined from Tauc plots was ~1.57±0.02 eV.

  • α is a few 104 cm−1 just above Eg and exceed 105 cm−1 above photon energy ~2.5 eV.

  • Samples have equivalent efficiencies of 2.8% (500 nm) and 5.3% (800 nm).

Abstract

We have determined for the first time the device-relevant optical constants of 500 nm and 800 nm-thick Cu2ZnSnS4 absorbers, grown on bare and Mo-coated soda-lime glass (SLG), using spectroscopic ellipsometry (SE). The composition, structure, phase purity and morphology were characterized by X-ray fluorescence, X-ray photoelectron spectroscopy depth profiling, X-ray diffraction, Raman spectroscopy, scanning-electron microscopy and atomic force microscopy. For the SE analysis, carefully determined sample characteristics were utilized to build a multilayer stack optical model, in order to derive the dielectric functions and refractive indices. The SE-derived absorption coefficients from CZTS/SLG samples were compared with those derived from complementary spectrophotometry measurements and found to be in good agreement. The bandgap determined from Tauc plots was Eg=1.57±0.02 eV. The absorption coefficients just above the bandgap were found to be a few 104 cm−1 and to exceed 105 cm−1 at energies above ~2.5 eV, which is much higher than previously found. The sub-bandgap k-value was found to be k~0.05 or less, suggesting that a moderate band tail is present. Separate device characterization performed on identical samples allowed us to assign device efficiencies of, respectively, 2.8% and 5.3% to the 500 nm and 800 nm-thick samples featured in this study.

Introduction

The quaternary compound kesterite Cu2ZnSnS4 (CZTS) is a promising earth-abundant choice for thin film solar cell applications [1], owing to its direct bandgap of optimal size ~1.5 eV [2], [3], [4], [5], its high absorption coefficient >104 cm−1 in the visible region [2], [3], as well as crystallographic and electronic structures similar to that of high-efficiency Cu(In,Ga)Se2 [6] based absorbers. Since the first 0.66%-efficient CZTS solar cell device made by Katagiri et al. in 1996, great progress was seen over the years [1], [7]. The current efficiency records are ~9.2% for CZTS [8] and ~12.6% for its close relative CZTSSe [9]. However, more improvements are needed for CZTS to reach above ~15% in order to satisfy the requirements for practical applications. To understand the loss mechanisms and further improve the efficiency, electrical and optical device modeling [10], [11] are needed. High-quality wavelength-dependent optical constants that accurately represent the device-relevant CZTS material are in demand as a crucial and a fundamental input if correct conclusions are to be drawn from such analyses. In general, optical constants, being fundamental material properties, are worth being repeatedly determined for an important emerging material like CZTS.

So far, the existing literature ellipsometry data for the pure sulfide CZTS are limited [12], [13], [14] and have been deficient in one way or another: sample composition and description of surface preparation methods were sometimes lacking; the importance of careful material characterization was frequently overlooked; possible influences from secondary phases were often neglected; the analyzed absorbers were rarely grown in the same way or on the same substrates as those for devices; none have compared ellipsometry data with that from transmittance and reflectance measurements in the same study even though their transparent substrates permit such comparison. Furthermore, there is a general lack of comparable device performance values provided in the same study despite the generally observed spread in performance from different CZTS processing routes or even between nominally identical processes. To reach accurate conclusions from the complex optical analysis, careful complementary characterization of the studied samples is key – as demonstrated by Ahn et al. in their detective work that settled the controversy over the bandgap value of CZTSe [15]. In this work, we attempt to rectify these oversights, providing much sought-after ellipsometry data determined from device-relevant material. We analyzed ~500 nm and ~800 nm thick Cu-poor and Zn-rich CZTS absorbers grown on bare as well as Mo-coated soda-lime glass (SLG), which yielded solar cells having 2.8% (~500 nm) and 5.3% (~800 nm) solar conversion efficiency, measured on virtually identical samples. The choice of substrates ensures that the as-fabricated CZTS/Mo/SLG absorbers have the same characteristics as those in a device in terms of e.g. grain structure and Na incorporation, while the CZTS/SLG samples offer the possibility to compare the optical absorption coefficients. The composition, structure, phase purity and morphology were carefully examined and the substrate effects were evaluated. Knowledge gained from these characterizations was then used for establishing a realistic optical model for the spectroscopic ellipsometry (SE) analysis, in order to determine the dielectric functions. The SE-derived absorption coefficients were compared with those extracted from complementary spectral transmittance and reflectance measurements done on CZTS/SLG samples.

Section snippets

Preparation

CZTS films were prepared on bare and Mo-coated SLG by DC magnetron reactive co-sputtering of CuS, Zn and Sn targets in Ar and H2S atmosphere in a Von Ardenne CS600 system, followed by a rapid annealing process [16]. The detailed fabrication procedure can be found elsewhere [17]. Samples pairs S1 (CZTS/Mo/SLG) and S2 (CZTS/SLG) were produced in the precursor-deposition batch A and annealed together in a graphite box. Samples pairs S3 (CZTS/Mo/SLG) and S4 (CZTS/SLG) were produced in the

SE measurements

Spectroscopic ellipsometry (SE) characterizes the changes of the polarization state of light as a result of reflection from the sample. In this work, by using a variable-angle spectroscopic ellipsometer with a rotating analyzer equipped with an auto-retarder (J.A. Woollam Co., Inc.), standard ellipsometric functions ψ and Δ were recorded in the wavelength range of 300–1700 nm with 10 nm steps, at angles of incidence of 65°, 70° and 75°. For acquisitions on samples S2 and S4 with SLG substrates, a

Optical constants

The dielectric functions and refractive indices extracted from the modeling analysis are shown in Fig. 8. (The tabulated refractive indices are included in “Supplementary materials”). Also shown in Fig. 8 are the data derived from previous work from Levcenko et al. [12], Li et al. [13] as well as the computational work using density functional theory (DFT) from Persson [6], which serves as a valuable indication for general trends. All the samples in the present work share similar features: an

Concluding remarks

In summary, we have presented for the first time the determination of the optical constants of reactively sputtered CZTS films prepared on bare and Mo-coated SLG by spectroscopic ellipsometry (SE). Careful characterization was performed in terms of composition, phase purity and morphology to ensure the correctness of the optical layer stack model used for the SE analysis, which successfully included ZnS and SnS2 segregations. It was found that SLG substrates tend to induce larger grain growth

Acknowledgments

S.-Y. Li is grateful for assistance on AFM from Nina Shariati Nilsson, discussion with Tove Ericson regarding XRF measurements and Olivier Donzel-Gargand regarding SEM measurements. We would like to thank Knut and Alice Wallenberg foundation, Swedish research council, the Foundation for Strategic Research, and The People Program (Marie Curie Actions) of the European Union׳s Seventh Framework Program FP7/2007-2013/ under REA Grant agreement no. 316488 (KESTCELLS) for financial support.

References (42)

  • S. Siebentritt

    What limits the efficiency of chalcopyrite solar cells?

    Sol. Energy Mater. Sol. Cells

    (2011)
  • I. Kentaro et al.

    Electrical and optical properties of stannite-type quaternary semiconductor thin films

    Jpn. J. Appl. Phys.

    (1988)
  • C. Persson

    Electronic and optical properties of Cu2ZnSnS4 and Cu2ZnSnSe4

    J. Appl. Phys.

    (2010)
  • H. Hiroi, N. Sakai, T. Kato, H. Sugimoto, High voltage Cu2ZnSnS4 submodules by hybrid buffer layer, in: Proceedings of...
  • M.A. Green et al.

    Solar cell efficiency tables (Version 45)

    Prog. Photovolt.: Res. Appl.

    (2015)
  • T. Gokmen et al.

    Semi-empirical device model for Cu2ZnSn(S,Se)4 solar cells

    Appl. Phys. Lett.

    (2014)
  • M.T. Winkler et al.

    Optical designs that improve the efficiency of Cu2ZnSn(S,Se)4 solar cells

    Energy Environ. Sci.

    (2014)
  • S. Levcenko et al.

    Optical constants of Cu2ZnSnS4 bulk crystals Moldavian

    J. Phys. Sci.

    (2009)
  • J. Li et al.

    Spectral optical properties of Cu2ZnSnS4 thin film between 0.73 and 6.5 eV

    Opt. Express

    (2012)
  • S. Ahn et al.

    Determination of band gap energy (Eg) of Cu2ZnSnSe4 thin films: on the discrepancies of reported band gap values

    Appl. Phys. Lett.

    (2010)
  • J.J. Scragg et al.

    Effects of back contact instability on Cu2ZnSnS4 devices and processes

    Chem. Mater.

    (2013)
  • Cited by (36)

    • Pulsed laser deposited and sulfurized Cu<inf>2</inf>ZnSnS<inf>4</inf> thin film for efficient solar cell

      2021, Solar Energy Materials and Solar Cells
      Citation Excerpt :

      Kesterite-quaternary material Cu2ZnSnS4 (CZTS)-based solar cells have become an admirable substitute and the most attractive absorber material on because of their excellent performance, benefitting from non-toxic and earth-abundant elements. CZTS is a p-type semiconductor with a bandgap of 1.5 eV and a high absorption coefficient (>104 cm −1) [3,4]. Thus, a micron-level film could absorb almost all photons with energy greater than the bandgap.

    • Heterojunction post-heat treatment process driving high efficiency for Cu<inf>2</inf>ZnSnS<inf>4</inf> solar cell

      2021, Solar Energy Materials and Solar Cells
      Citation Excerpt :

      Raman spectroscopy can further determine whether there are secondary phases in the samples. In addition, the penetration depth of excitation wavelength 532 nm laser is about 100 nm [29]. Since the thickness of CdS is ~60 nm, the CdS layer, CdS/CZTS interface and the surface region of CZTS can be well detected by using excitation wavelength of 532 nm.

    View all citing articles on Scopus
    View full text