Over 15% efficient wide-band-gap Cu(In,Ga)S2 solar cell: Suppressing bulk and interface recombination through composition engineering

•Cu deficiency suppresses the bulk recombination losses•Theoretical understanding of defects and its relation to antisite defects•Effective interface passivation using Zn(O,S) buffer layer•High power conversion efficiency from non-toxic H2S-free Cd-free process Cu(In,Ga)S2 is a high-potential material for usage in tandem solar cells; however, has remained limited so far. High losses both account performance limitation. In this work, we adopt holistic approach address losses. We show that can be substantially suppressed by controlling Cu material. From theoretical calculations, argue reduces are probably most detrimental defects. Additionally, effectively passivate through layer, thereby minimizing at interface. This leads cell device over 15% 1.6-eV-band-gap completely process. The path further improvement discussed increase viability toward application. progress remains significantly mainly due photovoltage (Voc) interfaces. Here, via combination photoluminescence, cathodoluminescence, electrical measurements, ab initio modeling, improve ∼1.6-eV-band-gap (Eg) Cu(In,Ga)S2. optoelectronic quality absorber improves upon reducing [Cu]/[Ga+In] (CGI) ratio, as manifested suppression deep defects, higher quasi-Fermi level splitting (QFLS), improved charge-carrier lifetime, Voc. identify CuIn/CuGa major performance-limiting defect comparing formation energies various intrinsic Interface layer Cu-poor devices, which activation energy equal Eg. demonstrate an 15.2% with Voc 902 mV H2S-free, Cd-free, KCN-free state-of-the-art photovoltaic technologies improving every year coming close limits. For improvements, new concepts will needed, among which, cells considered viable way forward.1Albrecht S. Rech B. Perovskite cells: on top commercial photovoltaics.Nat. Energy. 2017; 2: 16196Crossref Scopus (75) Google Scholar, 2Jošt M. Kegelmann L. Korte Albrecht Monolithic perovskite review present status advanced characterization methods 30% efficiency.Adv. Energy Mater. 2020; 10: 1904102Crossref (156) 3Al-Ashouri A. Köhnen E. Li Magomedov Hempel H. Caprioglio P. Márquez J.A. Morales Vilches A.B. Kasparavicius Smith et al.Monolithic perovskite/silicon >29% enhanced hole extraction.Science. 370: 1300-1309Crossref PubMed (408) Scholar main criteria suitable band gap, efficiency, stability.4Zafoschnig L.A. Nold Goldschmidt J.C. race lowest costs electricity production: techno-economic analysis silicon, cells.IEEE J. Photovoltaics. 1632-1641Crossref (21) Therefore, sulfide chalcopyrite Cu(In,Ga)S2, variable gap between 1.5 2.4 eV, receiving considerable interest device.5Lomuscio Rödel T. Schwarz Gault Melchiorre Raabe D. Siebentritt Quasi-fermi-level Cu-rich CuInS2 layers.Phys. Rev. Applied. 2019; 11: 054052Crossref (23) 6Hiroi Iwata Y. Adachi Sugimoto Yamada New World-record pure-sulfide Cu(in,Ga)S2 thin-film With process.IEEE 2016; 6: 760-763Crossref (74) 7Kim Nagai Tampo Ishizuka Shibata Large open-circuit voltage boosting pure prepared Cu-deficient metal precursors.Prog. Photovolt. Res. Appl. 28: 816-822Crossref (13) 8Kaigawa R. Funahashi K. Fujie Wada Merdes Caballero Klenk Tandem ZnO coated substrates.Sol. Sol. Cells. 2010; 94: 1880-1883Crossref (31) 9Lewerenz H.J. Goslowsky Husemann K.-D. Fiechter Efficient CuInS2.Nature. 1986; 321: 687-688Crossref (91) adopts structure similar high-efficiency Cu(In,Ga)Se2. Despite high photoconversion (PCE) 23.35% achieved Cu(In,Ga)(S,Se)2,10Nakamura Yamaguchi Kimoto Yasaki Kato Cu(In,Ga)(Se,S)2 record 23.35%.IEEE 9: 1863-1867Crossref (514) Scholar,11NRELBest research-cell chart.https://www.nrel.gov/pv/cell-efficiency.htmlDate: 2020Google certified PCE 15.5% thus far.6Hiroi Hence, determining their underlying origin paramount importance and, consequently, chalcopyrite. A good requires efficient generation photocarriers followed sustained build-up density. latter directly correlated (QFLS) lifetime (τ), often used analyze quality. Non-radiative reduce maximum achievable QFLS increasing deficit. long time, large deficit (described VocSQ – Voc, where stands Shockley-Queisser Voc)12Shockey-Queisser limit single junction under AM G spectrum calculated principle detailed balance absorption emission radiation, assuming creation one e-h pair/photon only radiative recombination.Google more than 300 mV. corresponds 75% S-Q even best cell.6Hiroi Scholar,13Merdes Mainz Klaer Meeder Rodriguez-Alvarez Schock H.W. Lux-Steiner M.C. 12.6% CdS/Cu(In,Ga)S2-based open circuit 879mV rapid thermal process.Sol. 2011; 95: 864-869Crossref 14Hiroi Progress 1000-mV 1630-1634Crossref (18) 15Barange N. Chu V.B. Nam Ahn I.-H. Kim Y.D. Han I.K. Min B.K. Ko D.-H. Ordered nanoscale heterojunction architecture solution-based CuInGaS2 thin film performance.Adv. 1601114Crossref (7) 16He G. Yan C. Yuan X. Sun Huang He Zhang Stride al.11.6% process.ACS 3: 11974-11980Crossref (5) Moreover, rather low short lifetimes (∼hundreds ps) typically observed Cu(In,Ga)S2.7Kim Scholar,14Hiroi implies significant non-radiative lies (front- back-contact) defects.17Scheer Open questions after 20 years research.Prog. Photovolt.: 2012; 20: 507-511Crossref Concerning bulk, QLFS (and photoluminescence quantum yield [PLQY]) caused states. regard, role point been highlighted studies analysis, although, much unexplored Cu(In,Ga)S2.18Chen Wang C.-Y. J.-T. Hu X.-P. Zhou S.-X. First-principles study semiconductor CuInS2.J. Phys. 112: 084513Crossref (27) 19Lomuscio Sood Phonon coupling shallow CuInS2.Phys. 101: 085119Crossref 20Binsma J.J.M. Giling L.J. Bloem Luminescence CuInS2: I.J. Lumin. 1982; 27: 35-53Crossref (141) 21Hofhuis Schoonman Goossens Elucidation excited-state dynamics films.J. Chem. 2008; 15052-15059Crossref (39) secondary phases depends composition.22Thomere Guillot-Deudon Caldes M.T. Bodeux Barreau Jobic Lafond Chemical crystallographic investigation Cu2S-In2S3-Ga2S3 ternary system.Thin Solid Films. 2018; 665: 46-50Crossref (14) or CuInS2, grown Cu-excess conditions (referred Cu-rich) extensively explored, owing better grain growth superior properties. Lomuscio al. showed (1) lower (2) value was temperature consistent reduction PL intensity. However, it did not translate into losses.5Lomuscio Recently, found absorbers doping (D.A. al., unpublished data). Hiroi demonstrated 920 mV, exploiting optimized ZnMgO layer.6Hiroi same group later enhancement annealing (RTA) process, while underscoring morphology Ga back-grading.14Hiroi compared low-band-gap (1.52 eV) (1.51-eV) absorbers. They attributed reduced former.7Kim All approaches relied toxic H2S, Cd- and/or KCN-based physical insights still missing. Based above-discussed results, see limitation immediate attention. chemical these limiting relationship composition need understood Apart bulk-related losses, devices suffer recombination, limits Factors front surface include density formed during KCN etching Cu2-xS phase absorbers, recently shown case CuInSe2;23Elanzeery Babbe F. Werner Brammertz Challenge CuInSe2 etching.Phys. Materials. 055403Crossref (24) mismatch electronic alignment absorber/buffer leading effective gap.24Ghorbani Erhart Albe Cu(In,Ga)(S,Se)2 compounds In2S3, NaIn5S8, CuIn5S8 materials.Phys. 075401Crossref (16) Scholar,25Weinhardt Fuchs O. Groß Storch Umbach Dhere N.G. Kadam A.A. Kulkarni S.S. Heske Band CdS∕Cu(In,Ga)S2 cells.Appl. Lett. 2005; 86: 062109Crossref (77) extent aforementioned issues critically choice layer. Cu(In,Ga)S2/CdS forms energetically non-ideal consequently output Voc.26Hengel I. Neisser Current transport CuInS2:Ga/Cds/Zno cells.Thin 2000; 361–362: 458-462Crossref (59) 27Klenk Characterisation modelling 2001; 387: 135-140Crossref (264) 28Unold Sieber Ellmer reactive magnetron sputtering.Appl. 2006; 88: 213502Crossref (63) energetic position conduction minimum CdS (negative band-offset ΔEc) causes cliff-like electronic-band barrier interface.13Merdes Scholar,29Merdes Sáez-Araoz Ennaoui Recombination mechanisms highly Zn(S,O)/Cu(In,Ga)S2 based 2009; Scholar,30Konovalov Material requirements CIS 2004; 451–452: 413-419Crossref (60) alloying upshift edge. alteration wide-band-gap Cu(In,Ga)S2.31Wilhelm H.-W. Scheer influence thickness.J. 109: 084514Crossref (48) 32Neisser Hengel Matthes T.W. Álvarez-García Pérez-Rodríguez Romano-Rodríguez M.-C. Effect incorporation sequentially absorbers.Sol. 67: 97-104Crossref (38) 33Gloeckler Sites J.R. Efficiency limitations 480–481: 241-245Crossref (236) cells, deduced temperature-dependent (EA) were always band-gap energy, except al.7Kim Scholar,34Merdes Abou-Ras CdS/Cu(In,Ga)S2 efficiencies reaching 12.9% process.Prog. 2013; 21: 88-93Crossref (52) 35Klenk Bakehe Kaigawa Reiß Optimising cells—design analysis.Thin 424-429Crossref (36) 36Riedel Riediger Ohland Keller Knipper Parisi Photoelectric obtained processing different temperatures.Sol. 270-273Crossref 37Scheer Luck photovoltaics.Sol. 77: 777-784Crossref (64) 38Scheer Walter Fearheiley M.L. Lewerenz 10.2% efficiency.Appl. 1993; 63: 3294-3296Crossref (313) 39Marsen Wilhelm Steinkopf Klemz Unold copper-deficiency multi-stage co-evaporated layers 519: 7224-7227Crossref (9) attempt configuration cliff spike type (positive ΔEc), edge such have explored.6Hiroi Scholar,40Ennaoui Bär Kropp Highly-efficient mini-modules Zn(S,O) alternative bath 14: 499-511Crossref (79) allows light absorption. spike-like Zn(O,S), EA Cu(In,Ga)S2.29Merdes These results suggest prone recombination. Furthermore, provides promising avenue issue Finally, back-contact accounts additional realize related selenide tackled gradient, providing fewer minority carriers near back contact.41Gabor A.M. Tuttle Bode M.H. Franz Tennant A.L. Contreras M.A. Noufi Jensen D.G. Hermann Band-gap engineering Cu(In,Ga) Se2 films (In,Ga)2Se3 precursors.Sol. 1996; 41–42 (247–260, 247–260)Crossref (198) 42Feurer Bissig Weiss T.P. Carron Avancini Löckinger Buecheler Tiwari A.N. Single-graded CIGS narrow bandgap cells.Sci. Technol. Adv. 19: 263-270Crossref (34) 43Schock Thin heterostructures.in: Chalcogenide Wiley-VCH Verlag, 2011: 9-127Google challenge suppress performance. challenges by; Cu-composition especially exploring regime, engineering, (3) grading prevent backside exploit above-mentioned factors adopting multistage co-evaporation prepare deposition control helps attaining profile.44Chirilă Pianezzi Bloesch Gretener Uhl A.R. Fella Kranz Perrenoud Seyrling al.Highly Cu(In,Ga)Se2 flexible polymer films.Nat. 857-861Crossref (721) Scholar,45Gabor Albin D.S. High-efficiency CuInxGa1−xSe2 made (Inx,Ga1−x)2Se3 precursor films.Appl. 1994; 65: 198-200Crossref (558) methodically investigate effect varying ratios bulk. As indicator representative determine room-temperature (PL) spectroscopy compare CGI.46Wurfel potential radiation.J. C: State 15: 3967-3985Crossref (702) Scholar,47Gütay Bauer G.H. Local fluctuations properties sub-micron resolved towards “real life” conditions.Thin 517: 2222-2225Crossref (37) (SQ) lead QFLS, excess carrier loss. respect indicates interface.48Babbe Choubrac optical diode ideality factor enables fast screening semiconductors cells.Sol. RRL. 1800248Crossref (17) Scholar,49Kirchartz Stolterfoht Photoluminescence-based halide perovskites photovoltaics.Adv. 1904134Crossref (129) increases become Cu-poor, keeping CGI ratio above 0.9, along accompanied transient (TRPL) analysis. < 0.9 manuscript. assess impact characteristics cells. Cu(In,Ga)S2/Zn(O,S) very indicating also confirmed saturation current extrapolating absorber. Remarkably, Zn(O,S)/ composition. result, 388 conducted 0.93 1.29 having [Ga]/[Ga+In] (GGI) 0.12–0.18. single-stage processes, Figure 1 (refer experimental procedures S1 details). first structural subsequently terms nature particular focus correlate layers. explore differences absolute calibrated time-resolved capacitance measurements. high-quality interface-passivated 1A shows microstructural features 1-stage, 2-stage, 3-stage processed ratios, seen cross-sectional scanning electron (SE) imaging (see top-view SEM image S2). 1-stage stoichiometric (1.02) exhibits columnar grains rough surface, previous observation Ga-free CuInS2.5Lomuscio Grain boundaries provide shunting paths deleterious performance.5Lomuscio Scholar,51Schwarz On chemistry films.Nano 76: 105081Crossref (8) 2-stage (1.29) sample’s SE micrograph acquired CuSx phase, exists absorbers.52Fiechter Tomm Kanis Kautek W. homogeneity region, modes chalcopyrite-type CuInS2.phys. stat. sol. (b). 245: 1761-1771Crossref (20) During second stage beyond point, assists sulfur incorporation.53Kaigawa Improved Cu(In,Ga)S2.Thin 2002; 415: 266-271Crossref (100) It evident images compact smooth morphology. adjusting duration final (third) films, size appears diminished. Notably, reveal smaller Mo Yet, seems favorable larger films.5Lomuscio To probe orientation θ-2θ X-ray