An in situ-edxrd Study of Reactively co-sputtered Cu(In,Ga)S 2 Layers

Transcription

1 An in situ-edxrd Study of Reactively co-sputtered Cu(In,Ga)S 2 Layers Jonas Krause, Stephan Brunken, Klaus Ellmer Helmholtz-Zentrum Berlin für Materialien und Energie GmbH Hahn-Meitner-Platz Berlin XIX. Erfahrungsaustausch Oberflächentechnologie mit Plasma- und Ionenstrahlprozessen Mühlleithen / Vogtland März 2012

2 CONTENT Introduction and Motivation Cu(In,Ga)S 2 thin film solar cells reactive co-sputtering of CIGS absorber layers in-situ Energy Dispersive X-Ray Diffraction (EDXRD) Temperature Dependent Growth of CIGS Condensation of Indium Intermetallic Phases Summary and Outlook 2

3 Cu(In,Ga)S 2 -THIN FILM SOLAR CELL 1 µm cross sectional SEM image preparation method DC/RF magnetron sputtering chemical bath deposition DC magnetron sputtering 0.6 µm ZnO:Al front contact 0.1 µm CdS buffer layer µm Cu(In,Ga)S 2 absorber layer 0.5 µm Mo back contact float glas absorber deposition sequential process co-evaporation reactive co-sputtering for CuInS 2 thin film solar cells the best efficiencies reached ~11.5% an efficiency of 11.4 % [1] was also achieved with reactively co-sputtered CuInS 2 absorber layers [1] S. Seeger, K. Ellmer, Thin Solid Films, 517 (2009) 3

4 REACTIVE CO-SPUTTERING OF CIGS THIN FILMS reactive sputtering in Ar:H 2 S sputtering chamber with base pressure < Pa two planar magnetrons (Ø 51 mm) equipped with dc power supplies d target-substrate = 6 cm graphite radiation substrate heater process gas: Ar:H 2 S = 2:1 process pressure: 1 Pa substrate: mm 2 Si/SiO 2 (200 nm) Si/SiO 2 /Mo Si/SiO 2 /TiN absorber deposition Cu85Ga15-target Mo Ar/H 2 S substrate heater In-target Cu(In,Ga)S 2 use of a CuGa-Target for Cu(In,Ga)S 2 higher efficiencies (close to 13%) can be expected [2] [2] R. Kaigawa et al.thin Solid Films, 415 (2002) 4

5 ENERGY-DISPERSIVE X-RAY DIFFRACTION (EDXRD): in situ - EXPERIMENTS heater kaptonwindow substrate θ In-target cooled Ge-detector kaptonwindow white x-ray beam 6 60 kev goniometer Argon H 2 S synchrotron radiation DESY/HASYLAB F3 CuGa target time for one spectra 10 s θ 3,6 5

6 EDXRD RESULT Ar/H 2 S = 2:1 ; T sub = 410 C P In = 25 W ; P CuGa = 80 W fluorescence peaks indicator for deposited elements esc esc CuInS 2 (112) diffraction peaks indicator for growing phases and crystallinity t dep = 15 min Cu 1-x Ga x (111) Cu 1-x Ga x (200) 0.01 normalized intensity (a.u.) CuKβ CuKα GaKα GaKβ InKα InKβ 1,2 Cu 9 Ga 4 (330) t dep = 3 min Cu(In,Ga)S 2 (004/200) Cu(In,Ga)S 2 (204/220) Cu(In,Ga)S 2 (112) t dep = 1 min esc esc MoKα MoKβ 1,2 Mo(110) t dep = 0 min photon energy (kev)

7 EDXRD RESULT: DEPOSITION AT DIFFERENT TEMPERATURES higher condensation of In for lower temperatures vapor pressure of In very high p v,in 100 p v,ga > p v,cu CIGS (112) diffraction peak temperature dependent growth behaviour temperature dependent position of CIGS (112): higher Ga content for higher temperature normalized peak area (a.u.) calc. lattice plane space d CIGS(112) (nm) x CuKα 360 C 440 C GaKα InKα CuInS 2 (112) [3] peak position Cu(In,Ga)S 2 (112) CuGaS 2 (112) [4] process condition: co-sputtering in Ar/H 2 S = 2:1 constant power at both targets: P In = 25 W ; P CuGa = 80 W fixed temperature substrate: Si/SiO 2 /Mo [3] JCPDS [4] JCPDS FWHM (kev) normalized peak area (a.u.) : :05 00:10 00:15 deposition time (hh:mm) half width Cu(In,Ga)S 2 (112) peak area Cu(In,Ga)S 2 (112) 7

8 EDXRD RESULT: CONDENSATION OF INDIUM relation of fluorescence peak areas (InKα /CuKα) as indicator for indium condensation indium condensation significantly higher for lower substrate temperature no continuous but sharp change of condensation behavior in the range of T sub C process condition: co-sputtering in Ar/H 2 S = 2:1 constant power at both targets: P In = 25 W ; P CuGa = 80 W fixed temperature substrate: Si/SiO 2 /Mo fluorescence peak relation (a.u.) fluorescence peak relation (a.u.) : : C 380 C 440 C 410 C 440 C 380 C 410 C 440 C A peak (InKα)/A peak (CuKα) 00:05 00:10 00:15 A peak deposition (InKα)/Atime peak (CuKα) (hh:mm) 00:05 00:10 00:15 deposition time (hh:mm) 8

9 EDXRD RESULT: CONDENSATION OF INDIUM indium which has condensed at low temperature on the substrate does not evaporate when temperature is increased T Sub ( C) fluorescence peak relation (a.u.) C 360 C 380 C 410 C 440 C A peak (InKα)/A peak (CuKα) temperature 0.0 process condition: co-sputtering in Ar/H 2 S = 2:1 constant power at both targets: P In = 25 W ; P CuGa = 80 W fixed or variable temperature substrate: Si/SiO 2 /Mo 00: :05 00:10 00:15 deposition time (hh:mm) 9

10 EDXRD RESULT: CONDENSATION OF INDIUM qualitatively same temperature dependence for a TiN substrate surface influence of substrate surface in comparison to Mo: lower condensation on TiN better condensation on SiO 2 fluorescence peak relation (a.u.) C 410 C A peak(inkα)/a peak (CuKα) Mo TiN SiO 2 Mo TiN : :05 00:10 00:15 deposition time (hh:mm) process condition: co-sputtering in Ar/H 2 S = 2:1 constant power at both targets: P In = 25 W ; P CuGa = 80 W fixed temperature different substrates 10

11 EDXRD RESULT: REPRODUCIBILITY INVESTIGATION two nominal identical depositions 17 depositions in between condensation of In, Cu and Ga similar for both depositions crystallization of CIGS (112) phase quite different further investigation necessary normalized peak area (a.u.) normalized peak area (a.u.) CuKα x10-3 GaKα InKα Cu(In,Ga)S 2 (112) H1611 H1628 process condition: co-sputtering in Ar/H 2 S = 2:1 constant power at both targets: P In = 25 W ; P CuGa = 80 W T sub = 410 C substrate: Si/SiO 2 /Mo : :04 00:08 00:12 00:16 deposition time (hh:mm) 11

12 EDXRD RESULT: INTERMETALLIC PHASES 380 C 440 C possible phase identification: Cu 11 In 9 Cu 16 (In,Ga) 9 Cu 9 (In 1-x Ga x ) 4 incomplete sulfurization of the deposited material: indium containing phase for low temperature indium free phase for high temperature no constant growth of intermetallic phases: fast phase transformation, or lateral not homogenously distributed phases phase identification in comparison to [5] R. Mainz, dissertation (2008) 12

13 SUMMARY the growth and crystallization of reactively co-sputtered CIGS films can be monitored in-situ lower temperatures cause a higher condensation of Indium better crystallization of the CIGS for higher temperatures starting at low temperature, increasing the temperature during deposition highest condensation of Indium on SiO 2 lowest on TiN crystallization dynamics still under investigation intermetallic phases: incomplete sulfurization 13

14 OUTLOOK highest reported efficiencies for thin film solar cells 20,3 % [6] with co-evaporated Cu(In,Ga)Se 2 we are starting with experiments of reactive co-sputtering in Ar:H 2 Se to deposit Cu(In,Ga)Se 2 films [6] P. Jackson, et al. Prog Photovoltaics, 19 (2011) 14

15 Vielen Dank für Ihre Aufmerksamkeit! Blick in die Anlage während 15der Abscheidung von Cu(In,Ga)S 2

16 EDXRD RESULT: INTERMETTALIC PHASES 380 C 440 C

17 EDXRD RESULT: CONDENSATION OF INDIUM, HIGHER INDIUM POWER 2.0 fluorescence peak relation (a.u.) P In = 25 W C 360 C 380 C 410 C 440 C P In = 50W 410 C : :10 00:20 00:30 deposition time (hh:mm) 17