Interface Characterization to aid in the Development of alternative Buffer Layers

Transcription

1 Interface Characterization to aid in the Development of alternative Buffer Layers Clemens Heske Institute for Photon Science and Synchrotron Radiation Institute for Chemical Technology and Polymer Chemistry Karlsruhe Institute of Technology Department of Chemistry and Biochemistry, University of Nevada, Las Vegas

2 Marcus Bär Lothar Timo Weinhardt Hofmann Doug Duncan The Group at UNLV Clemens Heske Mike Weir Kyle George Yu Liu Moni Blum Ryan Bugni Michelle Mezher Kim Horsley Sarah Alexander Marc Haeming, Samantha Rosenberg, Chase Aldridge, Dirk Hauschild (KIT) Key partners: W. Yang, J.D. Denlinger, Advanced Light Source, Berkeley Lab Team members: Weinhardt-group at KIT, Bär-group at HZB, Reinert-group at U Wü And former group members: S. Pookpanratana, S. Krause, I. Tran, Y. Zhang, R. Felix, M. Folse, G. Gajjala, S. Sudarshanam, J. White, A. Ranasinghe, A. Luinetti,...

3 Outline Electron and soft X-ray spectroscopies Quick review: CdS/Cu(In,Ga)(S,Se) 2 Chemical structure of annealed In x S y /CuIn(S,Se) 2 Electronic structure of Zn(O,S)/Cu(In,Ga)Se 2

4 UV/Soft X-ray/Electron Spectroscopies UV-Visible Absorption Spectroscopy (UV-Vis) Photoelectron Spectroscopy Photoelectron- (PES, XPS, Spectroscopy UPS) (PES) E e - e - eē Conduction band h h Inverse Photoemission (IPES) e Auger Electron Spectroscopy (AES) X-Ray Emission Spectroscopy ( XES) h h e - Valence band X-ray Absorption Spectroscopy (XAS) Core level

5 Applying soft x-ray/electron spectroscopies to applied questions Experimental approach needs to be customtailored to the actual question Sometimes, cuttingedge and/or unconventional approaches needed Need expertise (know what you are doing) In-situ! scienceblogs.com/zooillogix

6 Surface and Interface Analysis at UNLV Gloveboxes High dyn. range XPS, UPS, Auger, IPES High-res XPS, UPS, Auger Sample preparation and distribution Scanning Probe Microscope

7 Beamline 8.0 Advanced Light Source Lawrence Berkeley National Lab

8 SALSA: Solid And Liquid Spectroscopic Analysis Photoemission In-situ cell RSI 80, (2009) X-ray Emission U Würzburg, UNLV, HZ Berlin, KIT

9 Outline Electron and soft X-ray spectroscopies Quick review: CdS/Cu(In,Ga)(S,Se) 2 Chemical structure of annealed In x S y /CuIn(S,Se) 2 Electronic structure of Zn(O,S)/Cu(In,Ga)Se 2

10 XES of various sulfur compounds XES (1) (2) (3) (4) Peak identification: (5) CdSO 4 (1): S 3s S 2p sulfide (2): Cd 4d S 2p S-Cd bonds (3): S 3s S 2p S-O bonds In 5s S 2p S-In bonds Normalized Intensity CuS Cu 2 S (4): Cu 3d S 2p S-Cu bonds (5): S 3d S 2p S-O bonds CdS Local environment of sulfur atoms can be identified! phys.stat.sol. (a) 187, 13 (2001) Emission Energy (ev) CuIn(S,Se) 2

11 Intermixing at the CdS/CISe interface (XES) Chemical bond between Cd and S Cd-S bond is absent for thin overlayer diffusion of S into the CIGS film APL 74, 1451 (1999)

12 Intermixing at the CdS/CISe interface (PES) detectable Se signal for thick CdS layers Se segregation detectable In signal for (less) thick layers In segregation APL 74, 1451 (1999)

13 Intermixing: Summary ZnO CdS Cu(In,Ga)Se 2 Mo CdS 1-z Se z Cd v In w S 1-y Se y CuInS x Se 2-x Cu(In,Ga)Se 2 Na-lime glass APL 74, 1451 (1999)

14 What grows at the interface? (step 1) (a) Se M 2,3 & S L 2,3 h exc = 200 ev (b) CuInS : 12 hours (total) for 5 nm spectrum 2009: 10 minutes for 1 min spectrum ZnO CdS Allows to see Se M 2,3! Then: Siemens now: NREL (17.6%) Subtract Se M 2,3 and CdSSL 2,3 contribution Residual looks like In 2 S 3 or Ga 2 S 3 CdS 1-z Se z Cd v In w S 1-y Se y Norm. Intensity CdS Ref min (x 1) 8 min (x 2) 4 min (x 8) 2 min (x 30) 1 min (x 65) 0.5 min (x 90) 0 min (x 140) In 2 S 3 Ga 2 S 3 Cu 2 S 4 min Diff (x 17) 2 min Diff (x 2.4) 1 min data 1 min Diff (x 2.7) 0.5 min Diff (x 4.5) Cu(In,Ga)Se 2 CuInS x Se 2-x Mo Na-lime glass Cu(In,Ga)Se 2 APL 97, (2010) Emission Energy [ev] Emission Energy [ev]

15 What grows at the interface? (step 2) Effective CdS Thickness [Å] CIGSe CdS Fraction Difference CdS CBD time [min] Spectral separation allows to draw a depth profile Additional sulfide species is localized at the interface APL 97, (2010)

16 Sulfur gradient-driven Se diffusion at the CdS/CuIn(S,Se) 2 solar cell interface (step 1) x3 Ga 3s Se 3p Ga LMM d) S/Seratio 0 Mg K XPS spectra of AVANCIS (a-c) and NREL (d) absorbers Normalized Intensity x3 c) b) Different S/Se ratios (derived from fits) are given on right ordinate x3 S 2p x3 a) Binding Energy (ev) APL 96, (2010)

17 The CdS/CIGSSe junction CdS/CISe: APL 79, 4482 (2001) CdS/CISSe: EuPVSEC17 (2001), p.1261 CuInSe surface 2 Thin CdS on CuInSe 2 CdS surface (on CuInSe ) 2 CuIn(S,Se) surface 2 CdS/CuIn(S,Se) heterojunction 2 thick CdS/CuIn(S,Se) 2 surface CBM CBO = 0.0 (± 0.2) ev CBM CBM CBO = 0.0 (±0.15) ev CBM E F 1.4 (± 0.15) ev VBM 2.2 (± 0.15) ev E F 1.4 (±0.15) ev VBM 2.4 (±0.15) ev E F VBM VBO = 0.8 (± 0.2) ev VBM VBO = 1.0 (±0.15) ev CdS/CIGS: APL 86, (2005) 0.86 (±0.1) ev Cu(In,Ga)S 2 surface 1.76 (±0.15) ev VBO = (±0.15) ev interface CBO=-0.45 (±0.15) ev CdS/Cu(In,Ga)S 2 surface 0.46 (±0.1) ev E F 2.47 (±0.15) ev Good devices have a flat conduction band offset S-Se intermixing, which can be controlled by S content in CIGSSe CdS/CIGS: cliff in the conduction band

18 Outline Electron and soft X-ray spectroscopies Quick review: CdS/Cu(In,Ga)(S,Se) 2 Chemical structure of annealed In x S y /CuIn(S,Se) 2 Electronic structure of Zn(O,S)/Cu(In,Ga)Se 2

19 Annealing-Induced Effects on the Chemical Structure of the In2S3/CuIn(S,Se)2 Interface D. Hauschild et al., JPC C 119, (2015)

20 Annealing-Induced Effects on the Chemical Intensity (arb. u.) Structure of the In2S3/CuIn(S,Se)2 Interface University of Würzburg, KIT, AVANCIS GmbH, UNLV, Advanced Light Source, HZB, Brandenburgische Technische Universität Cottbus-Senftenberg, ANKA Na 1s Cu 2p 3/2 Se 3d In 2 S 3 / CISSe 80 nm annealed 80 nm 12.5 nm 5.6 nm 3.8 nm 1.8 nm 0.5 nm As-grown: abrupt interface After heat treatment (200 C) to simulate subsequent process steps: strong copper diffusion into the In 2 S 3 layer strong sodium diffusion into the In 2 S 3 layer CISSe Binding Energy (ev) D. Hauschild et al., JPC C 119, (2015)

21 Composition (%) Quantification: Chemical Structure of the In2S3/CuIn(S,Se)2 Interface In S Cu Se exp. decay nom. Buffer Layer Thickness (nm) annealed D. Hauschild et al., JPC C 119, (2015) S 8 Cu In As-grown: formation of a sulfurpoor (indium-rich) In 2 S 3 surface After heat treatment (200 C) to simulate subsequent process steps: copper diffusion into the In 2 S 3 layer; Cu concentration: Cu 1 In concentration near In 5 S concentration near S 8 formation of a copper indium sulfide phase

22 Analysis: S environment at the In2S3/CuIn(S,Se)2 Interface Intensity (arb. u.) XES S L 2,3 a) 0.45 x In 2 S 3 Residual 0.55 x CISSe b) Sum Sum h = 200 ev 10 nm In 2 S 3 /CISSe annealed 80 nm In 2 S 3 /CISSe 0.60 x In 2 S x CISSe Intensity (arb. u.) XES h = 200 ev S L 2,3 (2) (1) e) 80 nm annealed In 2 S 3 / CISSe d) In 2 S 3 c) 80 nm b) 10 nm a) CISSe In 2 S 3 / CISSe Emission Energy (ev) D. Hauschild et al., JPC C 119, (2015) Residual Emission Energy (ev) As-grown: formation of an In 2 S 3 surface sulfur atoms in both In 2 S 3 and CuIn(S,Se) 2 chemical environments After heat treatment (200 C) to simulate subsequent process steps: formation of a copper indium sulfide phase

23 Outline Electron and soft X-ray spectroscopies Quick review: CdS/Cu(In,Ga)(S,Se) 2 Chemical structure of annealed In x S y /CuIn(S,Se) 2 Electronic structure of Zn(O,S)/Cu(In,Ga)Se 2

24 UPS and IPES of the Zn(O,S)/CIGSe interface UPS and IPES spectra of bare absorber (bottom) and thickest Zn(O,S)/CIGSe sample (top) UPS - He I (VBM) IPES (CBM) Error bars are ±0.10 and ±0.15 ev for the VBM and CBM determination, respectively VBM and CBM are determined by linear extrapolation of the leading edge This is not the full picture! Must take interface-induced band bending into account Normalized Intensity Zn(O,S) 2.75 ev ± 0.18 ev E F ev CIGSe 1.55 ev ± 0.18 ev ev 0.45 ev 0.50 ev Mezher et al., Progress in Photovoltaics: Research and Applications, 2016, In Print Binding Energy rel. E F (ev)

25 XPS: Core-Level Peak Positions Core Level CIGSe BE (ev) Thin 5 min Zn(O,S) BE (ev) Shift Se 3d In 3d 5/ Cu 2p 3/ ev Core Level Thin 5 min Zn(O,S) BE (ev) Thick 22.5 min Zn(O,S) BE (ev) Shift S 2p 3/ O 1s (Zn(OH) 2 ) O 1s (ZnO) Zn 2p 3/ ev Core level peak positions of the bare absorber, 5 min, and 22.5 min Zn(O,S)/CIGSe sample Relative shifts shows there is band bending as the interface forms Mezher et al., Progress in Photovoltaics: Research and Applications, 2016, In Print

26 XPS, UPS, IPES: Interface Band Alignment Small interface-induced band bending Very small conduction band offset (CBO) Small spike (essentially flat) conduction band alignment similar to high-efficiency CdS/CIGSe devices Sizable valence band offset (VBO) hole barrier! CIGSe 5Surface Zn(O,S) E5Surface g 0.18 ev Interface 0.18 evee: ev 0.15 ev 1.05 ev 0.10 ev 0.06 ev CBO: 0.09 ev 0.20 ev VBO: 1.11 ev 0.15 ev :2.7g0.45 ev 0.15 ev E F 2.30 ev 0.10 ev Mezher et al., Progress in Photovoltaics: Research and Applications, 2016, In Print 0.20 ev

27 Summary Soft x-ray and electron spectroscopies allow the investigation of surfaces and interfaces in a unique way: Atom-specific and chemically sensitive Chemical properties (intermixing, impurities,...) Electronic structure (gaps, offsets,...) Can help in optimizing manufacturing processes and industrial products Particularly suited for thin film PV materials, and especially CIGSSe! heske@kit.edu, heske@unlv.nevada.edu