Impact of Annealing-Induced Intermixing on the Electronic Level

Dec 30, 2015 - ... University of Nevada, Las Vegas (UNLV), 4505 Maryland Parkway, Las ... Institute for Photon Science and Synchrotron Radiation, Karl...
0 downloads 0 Views 742KB Size
Research Article www.acsami.org

Impact of Annealing-Induced Intermixing on the Electronic Level Alignment at the In2S3/Cu(In,Ga)Se2 Thin-Film Solar Cell Interface Marcus Bar̈ ,*,†,‡,§ Nicolas Barreau,∥ François Couzinié-Devy,∥ Lothar Weinhardt,§,⊥,#,∇ Regan G. Wilks,† John Kessler,∥ and Clemens Heske§,⊥,#,∇ †

Renewable Energy, Helmholtz-Zentrum Berlin für Materialien und Energie GmbH, Hahn-Meitner-Platz 1, 14109 Berlin, Germany Institut für Physik und Chemie, Brandenburgische Technische Universität Cottbus-Senftenberg, Platz der Deutschen Einheit 1, 03046 Cottbus, Germany § Department of Chemistry and Biochemistry, University of Nevada, Las Vegas (UNLV), 4505 Maryland Parkway, Las Vegas, Nevada 89154-4003, United States ∥ Institut des Matériaux Jean Rouxel (IMN)-UMR 6502, Université de Nantes, Centre National de la Recherche Scientifique (CNRS), 2 rue de la Houssinière, BP 32229, 44322 Nantes Cedex 3, France ⊥ Institute for Photon Science and Synchrotron Radiation, Karlsruhe Institute of Technology, Hermann-von-Helmholtz-Platz 1, 76344 Eggenstein-Leopoldshafen, Germany # ANKA Synchrotron Radiation Facility, Karlsruhe Institute of Technology, 76344 Eggenstein-Leopoldshafen, Germany ∇ Institute for Chemical Technology and Polymer Chemistry, Karlsruhe Institute of Technology, Engesserstraße 18/20, 76128 Karlsruhe, Germany ‡

S Supporting Information *

ABSTRACT: The interface between a nominal In2S3 buffer and a Cu(In,Ga)Se2 (CIGSe) thinfilm solar cell absorber was investigated by direct and inverse photoemission to determine the interfacial electronic structure. On the basis of a previously reported heavy intermixing at the interface (S diffuses into the absorber; Cu diffuses into the buffer; and Na diffuses through it), we determine here the band alignment at the interface. The results suggest that the pronounced intermixing at the In2S3/CIGSe interface leads to a favorable electronic band alignment, necessary for high-efficiency solar cell devices. KEYWORDS: chalcopyrite solar cell, In2S3 buffer, band offset, intermixing, inverse photoemission



INTRODUCTION One main goal of past and current research on Cu(In,Ga)Se2 (“CIGSe”) chalcopyrite-based thin-film solar cells, for which several groups have reached efficiencies in excess of 20%,1−4 is the replacement of the CdS buffer layer with a Cd-free, more transparent material. Furthermore, it would be desirable to substitute the conventional chemical bath deposition (CBD) of the buffer with an in-line processing approach. To address both concerns, In2S3 thin films deposited by, e.g., physical vapor deposition,5,6 sputtering,7 atomic layer deposition,8 or spray ion layer gas reaction9 are promising candidates to replace the conventional CBD−CdS buffer in CIGSe solar cells. For an insight-driven buffer layer optimization, a detailed knowledge of the chemical and electronic interface properties is a prerequisite. Studies of the chemical structure of the In2S3/ CIGSe thin-film solar cell interface find that a pronounced diffusion of Cu and Na from the CIGSe/Mo/glass substrate into the (nominal) In2S3 buffer layer takes place5,8,10−15 at postannealing/deposition temperatures of 200−250 °C. Solar cell studies investigating in detail the impact of the In2S3 annealing/ deposition temperature on device performance show that this is also the temperature range necessary for high device efficiencies (>16%16). In an earlier non-destructive depth-resolved study of © XXXX American Chemical Society

the In2S3/CIGSe interface, using a set of samples with varying In2S3 thicknesses and a combination of different X-ray spectroscopy techniques with different probing depths, we were able to reveal a detailed picture of this chemical interface structure.17 This structure is summarized in Figure 1. Briefly, during co-evaporation of In2S3 onto a CIGSe substrate (held at 200 °C), a “CuIn5S8”-like buffer material is formed, the absorber surface/interface region is chemically modified by a

Figure 1. Schematic presentation of the chemical structure of the “In2S3”/CIGSe interface. Reproduced with permission from ref 17. Copyright 2010 AIP Publishing LLC. Received: November 4, 2015 Accepted: December 30, 2015

A

DOI: 10.1021/acsami.5b10614 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 2. (a) UPS (He II, left) and IPES (right) spectra of the bare NH3-treated CIGSe absorber (top), the 1/32, 1/16, 1/8, 1/2, and 1/1 In2S3/ CIGSe samples, and the In2S3/Mo “reference” sample (bottom). The red solid lines represent the linear extrapolation of the leading edges to derive the VBM and CBM positions with respect to the Fermi level, EF. (b) VBM and CBM values derived from the data in panel a. (c) Surface band gap (Egsurf = CBM − VBM) values, derived from the VBM and CBM values in panel b.

Table 1. VBM, CBM, and Egsurf Values at the Surface of the Bare (NH3-Treated) CIGSe Absorber, the 1/32, 1/16, 1/8, 1/2, and 1/1 In2S3/CIGSe Samples, and the In2S3/Mo “Reference” Sample, Derived from the UPS and IPES Data Presented in Figure 2 sample

VBM (eV) ± 0.10 eV

CBM (eV) ± 0.10 eV

Egsurf (eV) ± 0.14 eV

CIGSe 1/32 In2S3/CIGSe 1/16 In2S3/CIGSe 1/8 In2S3/CIGSe 1/2 In2S3/CIGSe 1/1 In2S3/CIGSe In2S3/Mo

−0.73 −0.97 −1.27 −1.31 −1.23 −1.29 −1.84

0.78 0.82 0.97 0.91 0.99 0.89 0.61

1.51 1.79 2.24 2.22 2.22 2.18 2.45

buffer layer by thermal co-evaporation of elemental indium and sulfur at 200 °C substrate temperature.17,18 To vary the In2S3 thickness for this experiment, different deposition times were used. The standard (80 nm nominal thickness) buffer layer used in solar cells is prepared with a deposition time of 10 min (termed “1/1”). Additional In2S3 layers were prepared with deposition times of 5 min (“1/2”), 2.5 min (“1/4”), 75 s (“1/8”), 40 s (“1/16”), 20 s (“1/32”), and 10 s (“1/64”). As a reference, a nominally 40 nm thick In2S3 layer was deposited on a Mo/glass substrate. Note that, according to XPS measurements, the indium sulfide material deposited on molybdenum can be considered to be stoichiometric.17 After preparation, all samples were sealed in polyethylene bags filled with dry N2 and desiccant for transport. At UNLV, the samples were transferred into the analysis chamber (base pressure of