Halide Perovskites for Selective Ultraviolet-Harvesting Transparent Photovoltaics

Abstract

•Halide perovskites are demonstrated in UV-selective absorbing transparent solar cells•Compositions are ideally tuned to band gaps to enable the highest PCE at the highest AVT•Transparent devices show good stability under ambient conditions and UV light Transparent photovoltaic (TPV) technologies offer an exciting approach to produce smart windows on buildings, vehicles, mobile electronics, and greenhouses. In this work, we develop a new approach to TPVs based on halide perovskites where the band gap is precisely tuned with a range of compositions to selectively harvest only UV photons with tailored band gaps between 410 and 440 nm. We develop a processing method to overcome the challenges that exist to control the morphology of these less soluble compositions. Full optimization of these perovskite cells could quickly yield theoretical TPVs with PCEs up to 7% with >99% visible transparency. Such devices would rival state-of-the-art TPVs that selectively harvest near-infrared light while also providing a route to higher-efficiency multi-junction TPVs. Moreover, these initial demonstrations are sufficient to power smart windows and many emerging applications today while providing the high level of aesthetics needed for actual adoption. Halide perovskite materials have emerged as a potential Si replacement with excellent photovoltaic properties. There has been growing interest in applying halide perovskites to semitransparent and spatially segmented transparent photovoltaics (TPVs) to enable a greater range of deployment routes. However, the continuous-band absorption of these semiconductors prevents near-infrared selective harvesting typically targeted for TPVs with the highest efficiency and transparency needed to meet the aesthetic demands of many potential applications. In this work, we demonstrate TPVs based on perovskite semiconductors where the band gap is sensitively tuned with a range of compositions to selectively harvest only ultraviolet photons with band gaps between 410 and 440 nm. This approach offers theoretical efficiencies up to 7% with >99% visible transparency when precisely targeting band gaps around 435 nm. Practical optimization of these perovskite cells could quickly yield TPVs with power conversion efficiencies rivaling state-of-the-art near-infrared harvesting TPVs today while also providing a route to higher-efficiency multi-junction TPVs. Halide perovskite materials have emerged as a potential Si replacement with excellent photovoltaic properties. There has been growing interest in applying halide perovskites to semitransparent and spatially segmented transparent photovoltaics (TPVs) to enable a greater range of deployment routes. However, the continuous-band absorption of these semiconductors prevents near-infrared selective harvesting typically targeted for TPVs with the highest efficiency and transparency needed to meet the aesthetic demands of many potential applications. In this work, we demonstrate TPVs based on perovskite semiconductors where the band gap is sensitively tuned with a range of compositions to selectively harvest only ultraviolet photons with band gaps between 410 and 440 nm. This approach offers theoretical efficiencies up to 7% with >99% visible transparency when precisely targeting band gaps around 435 nm. Practical optimization of these perovskite cells could quickly yield TPVs with power conversion efficiencies rivaling state-of-the-art near-infrared harvesting TPVs today while also providing a route to higher-efficiency multi-junction TPVs. Transparent photovoltaic (TPV) technologies offer an effective approach to produce smart windows on buildings, vehicles, and greenhouses. TPVs can both regulate the transmission of solar heat and provide electricity generation by photoelectron conversion of the invisible part of the solar spectrum.1Lunt R.R. Theoretical limits for visibly transparent photovoltaics.Appl. Phys. Lett. 2012; 101: 043902Crossref Scopus (122) Google Scholar, 2Lunt R.R. Bulovic V. Transparent, near-infrared organic photovoltaic solar cells for window and energy-scavenging applications.Appl. Phys. Lett. 2011; 98: 113305Crossref Scopus (262) Google Scholar, 3Traverse C.J. Pandey R. Barr M.C. Lunt R.R. Emergence of highly transparent photovoltaics for distributed applications.Nat. Energy. 2017; 2: 849-860Crossref Scopus (392) Google Scholar Often, the most critical parameter in determining the adoptability of TPVs is the overall average visible transparency (AVT), since aesthetic demands place strict requirements for applications in mobile electronics and windows.3Traverse C.J. Pandey R. Barr M.C. Lunt R.R. Emergence of highly transparent photovoltaics for distributed applications.Nat. Energy. 2017; 2: 849-860Crossref Scopus (392) Google Scholar To achieve the highest combination of transparency and efficiency requires selectively harvesting all of the invisible parts of the solar spectrum, including the near-infrared (NIR) and ultraviolet (UV), resulting in efficiency limits up to 20.1% and record efficiencies now around 5.0% for AVTs >50%.1Lunt R.R. Theoretical limits for visibly transparent photovoltaics.Appl. Phys. Lett. 2012; 101: 043902Crossref Scopus (122) Google Scholar, 3Traverse C.J. Pandey R. Barr M.C. Lunt R.R. Emergence of highly transparent photovoltaics for distributed applications.Nat. Energy. 2017; 2: 849-860Crossref Scopus (392) Google Scholar While there is substantially less overall solar photon flux in the UV, efficiencies up to 7% are theoretically achievable.1Lunt R.R. Theoretical limits for visibly transparent photovoltaics.Appl. Phys. Lett. 2012; 101: 043902Crossref Scopus (122) Google Scholar This is because the approach of selectively harvesting UV-only photons can allow very high photovoltages (>1.5 V) from the high photon energies. This also enables integration of more traditional semiconductors, such as GaN, ZnO, and NiO. However, integration of these materials would require additional band gap tuning to achieve a precise band gap at 435 nm.4Strite S. Morkoç H. GaN, AlN, and InN: a review.J. Vac. Sci. Technol. B Nanotechnol. Microelectron. 1992; 10: 1237-1266Crossref Google Scholar, 5Özgür Ü. Alivov Y.I. Liu C. Teke A. Reshchikov M.A. Doğan S. Avrutin V. Cho S.J. Morkoç H. A comprehensive review of ZnO materials and devices.J. Appl. Phys. 2005; 98: 041301Crossref Scopus (10043) Google Scholar, 6Sawatzky G.A. Allen J.W. Magnitude and origin of the band gap in NiO.Phys. Rev. Lett. 1984; 53: 2339-2342Crossref Scopus (760) Google Scholar Utilizing band gaps >440 nm would result in sharp drops in color-rending index and transparency unsuitable for many applications, while band gaps much less than 430 nm drop the performance and photocurrent potential significantly. To date, only a few UV-harvesting TPV devices have been reported with less-ideal band gaps.7Davy N.C. Sezen-Edmonds M. Gao J. Lin X. Liu A. Yao N. Kahn A. Loo Y.-L. Pairing of near-ultraviolet solar cells with electrochromic windows for smart management of the solar spectrum.Nat. Energy. 2017; 2: 17104Crossref Scopus (102) Google Scholar, 8Karsthof R. Räcke P. von Wenckstern H. Grundmann M. Semi-transparent NiO/ZnO UV photovoltaic cells.Phys. Status Solidi A. 2016; 213: 30-37Crossref Scopus (64) Google Scholar Karsthof et al.8Karsthof R. Räcke P. von Wenckstern H. Grundmann M. Semi-transparent NiO/ZnO UV photovoltaic cells.Phys. Status Solidi A. 2016; 213: 30-37Crossref Scopus (64) Google Scholar reported a semitransparent NiO/ZnO UV photovoltaic cell with a power conversion efficiency (PCE) of 0.1% that was limited by a band gap <380 nm. More recently, UV-selective organic TPVs have also been demonstrated by Loo and co-workers with a PCE of up to 1.5% with an AVT of 60% and a visible band gap in the active layer of about 2.3 eV, which was capable of providing more than enough power for a smart window.7Davy N.C. Sezen-Edmonds M. Gao J. Lin X. Liu A. Yao N. Kahn A. Loo Y.-L. Pairing of near-ultraviolet solar cells with electrochromic windows for smart management of the solar spectrum.Nat. Energy. 2017; 2: 17104Crossref Scopus (102) Google Scholar Chu and co-workers reported a UV- and near-infrared protective semitransparent smart windows based on metal halide solar cells.9Boopathi K.M. Hanmandlu C. Singh A. Chen Y.-F. Lai C.S. Chu C.W. UV- and NIR-protective semitransparent smart windows based on metal halide solar cells.ACS Appl. Energy Mater. 2018; 1: 632-637Crossref Scopus (18) Google Scholar The semitransparent device resulted in an efficiency of up to 0.75% with an AVT of 49%. These studies are encouraging and demonstrate the potential for TPV applications. Because this UV-harvesting approach can enable the use of continuous-band absorbing semiconductors, it also offers a great untapped opportunity for halide perovskites, which are now being considered for replacements for Si, CdTe, and GaAs. Halide perovskite materials have emerged as a strong light-harvesting candidate with excellent photovoltaic properties,10Burschka J. Pellet N. Moon S.-J. Humphry-Baker R. Gao P. Nazeeruddin M.K. Gratzel M. Sequential deposition as a route to high-performance perovskite-sensitized solar cells.Nature. 2013; 499: 316-319Crossref PubMed Scopus (7829) Google Scholar, 11Liu M. Johnston M.B. Snaith H.J. Efficient planar heterojunction perovskite solar cells by vapour deposition.Nature. 2013; 501: 395-398Crossref PubMed Scopus (6571) Google Scholar, 12Topolovsek P. Lamberti F. Gatti T. Cito A. Ball J.M. Menna E. Gadermaier C. Petrozza A. Functionalization of transparent conductive oxide electrode for TiO2-free perovskite solar cells.J. Mater. Chem. A. 2017; 5: 11882-11893Crossref Google Scholar high quantum efficiencies above 90%, and a certified efficiency up to 22.7%.13National Renewable Energy Laboratory. (2018). Best research-cell efficiencies. https://www.nrel.gov/pv/assets/images/efficiency-chart.png.Google Scholar Perovskite materials are also among the cheapest light-absorption semiconductor materials14Cai M. Wu Y. Chen H. Yang X. Qiang Y. Han L. Cost-performance analysis of perovskite solar modules.Adv. Sci. 2017; 4: 1600269Crossref Scopus (297) Google Scholar, 15Liu D. Kelly T.L. Perovskite solar cells with a planar heterojunction structure prepared using room-temperature solution processing techniques.Nat. Photonics. 2014; 8: 133-138Crossref Scopus (2291) Google Scholar, 16Snaith H.J. Perovskites: the emergence of a new era for low-cost, high-efficiency solar cells.J. Phys. Chem. Lett. 2013; 4: 3623-3630Crossref Scopus (2308) Google Scholar with readily tunable band gap through compositional control.17Kojima A. Teshima K. Shirai Y. Miyasaka T. Organometal halide perovskites as visible-light sensitizers for photovoltaic cells.J. Am. Chem. Soc. 2009; 131: 6050-6051Crossref PubMed Scopus (15167) Google Scholar, 18Noh J.H. Im S.H. Heo J.H. Mandal T.N. Seok S.I. Chemical management for colorful, efficient, and stable inorganic–organic hybrid nanostructured solar cells.Nano Lett. 2013; 13: 1764-1769Crossref PubMed Scopus (3752) Google Scholar In this work, we exploit halide perovskite tunability as a platform for creating UV-harvesting TPVs with ideal band gaps and visible absorption cutoffs that enable high AVT and color rendering index (CRI). We demonstrate TPV devices from various halide perovskite compositions with a PCE up to 0.52% simultaneously with an AVT of 73% and a CRI over 93. Given that these cells are limited by quantum efficiencies of 20%–30%, efficiencies approaching 4% should be achievable since halide perovskite cells now readily achieve quantum efficiencies above 90%. This study shows the exciting potential of new halide perovskite compositions for application in highly transparent photovoltaics and TPV multi-junctions. Methylammonium (MA) lead halide has become a champion perovskite semiconductor that has been broadly investigated in the past decade.10Burschka J. Pellet N. Moon S.-J. Humphry-Baker R. Gao P. Nazeeruddin M.K. Gratzel M. Sequential deposition as a route to high-performance perovskite-sensitized solar cells.Nature. 2013; 499: 316-319Crossref PubMed Scopus (7829) Google Scholar, 11Liu M. Johnston M.B. Snaith H.J. Efficient planar heterojunction perovskite solar cells by vapour deposition.Nature. 2013; 501: 395-398Crossref PubMed Scopus (6571) Google Scholar, 17Kojima A. Teshima K. Shirai Y. Miyasaka T. Organometal halide perovskites as visible-light sensitizers for photovoltaic cells.J. Am. Chem. Soc. 2009; 131: 6050-6051Crossref PubMed Scopus (15167) Google Scholar, 19Zhou H. Chen Q. Li G. Luo S. Song T.-b. Duan H.-S. Hong Z. You J. Liu Y. Yang Y. Interface engineering of highly efficient perovskite solar cells.Science. 2014; 345: 542Crossref PubMed Scopus (5546) Google Scholar, 20Stranks S.D. Eperon G.E. Grancini G. Menelaou C. Alcocer M.J.P. Leijtens T. Herz L.M. Petrozza A. Snaith H.J. Electron-hole diffusion lengths exceeding 1 micrometer in an organometal trihalide perovskite absorber.Science. 2013; 342: 341Crossref PubMed Scopus (7756) Google Scholar, 21Xing G. Mathews N. Sun S. Lim S.S. Lam Y.M. Grätzel M. Mhaisalkar S. Sum T.C. Long-range balanced electron- and hole-transport lengths in organic-inorganic CH3NH3PbI3.Science. 2013; 342: 344Crossref PubMed Scopus (5487) Google Scholar While other compositions are being actively investigated,22Comin R. Walters G. Thibau E.S. Voznyy O. Lu Z.-H. Sargent E.H. Structural, optical, and electronic studies of wide-bandgap lead halide perovskites.J. Mater. Chem. C. 2015; 3: 8839-8843Crossref Google Scholar little work has been done to optimize devices with band gaps near the visible absorption edge. Figure 1A shows the absorption spectra of MA-based perovskite films with various halide substitutions (Cl, Br, I). Among these perovskites, the iodide (MAPbI3) and bromide (MAPbBr3) perovskites have strong visible absorption and thus have been used in traditional opaque solar cells as the light absorber.17Kojima A. Teshima K. Shirai Y. Miyasaka T. Organometal halide perovskites as visible-light sensitizers for photovoltaic cells.J. Am. Chem. Soc. 2009; 131: 6050-6051Crossref PubMed Scopus (15167) Google Scholar, 23Heo J.H. Song D.H. Im S.H. Planar CH3NH3PbBr3 hybrid solar cells with 10.4% power conversion efficiency, fabricated by controlled crystallization in the spin-coating process.Adv. Mater. 2014; 26: 8179-8183Crossref PubMed Scopus (430) Google Scholar, 24Jeng J.-Y. Chiang Y.-F. Lee M.-H. Peng S.-R. Guo T.-F. Chen P. Wen T.-C. CH3NH3PbI3 perovskite/fullerene planar-heterojunction hybrid solar cells.Adv. Mater. 2013; 25: 3727-3732Crossref PubMed Scopus (1271) Google Scholar, 25Heo J.H. Im S.H. Noh J.H. Mandal T.N. Lim C.-S. Chang J.A. Lee Y.H. Kim H.-j. Sarkar A. Nazeeruddin M.K. et al.Efficient inorganic–organic hybrid heterojunction solar cells containing perovskite compound and polymeric hole conductors.Nat. Photonics. 2013; 7: 486Crossref Scopus (2201) Google Scholar Although there have been increasing efforts to prepare semitransparent solar cells by MAPbI3, the device AVT and CRI are generally low due to the direct tradeoff in efficiency and transparency.26Fu F. Feurer T. Jäger T. Avancini E. Bissig B. Yoon S. Buecheler S. Tiwari A.N. Low-temperature-processed efficient semi-transparent planar perovskite solar cells for bifacial and tandem applications.Nat. Commun. 2015; 6: 8932Crossref PubMed Scopus (372) Google Scholar, 27Bush K.A. Bailie C.D. Chen Y. Bowring A.R. Wang W. Ma W. Leijtens T. Moghadam F. McGehee M.D. Thermal and environmental stability of semi-transparent perovskite solar cells for tandems enabled by a solution-processed nanoparticle buffer layer and sputtered ITO electrode.Adv. Mater. 2016; 28: 3937-3943Crossref PubMed Scopus (369) Google Scholar, 28Eperon G.E. Burlakov V.M. Goriely A. Snaith H.J. Neutral color semitransparent microstructured perovskite solar cells.ACS Nano. 2014; 8: 591-598Crossref PubMed Scopus (386) Google Scholar, 29Jung J.W. Chueh C.-C. Jen A.K.Y. High-performance semitransparent perovskite solar cells with 10% power conversion efficiency and 25% average visible transmittance based on transparent CuSCN as the hole-transporting material.Adv. Energy Mater. 2015; 5: 1500486Crossref Scopus (212) Google Scholar, 30Eperon G.E. Bryant D. Troughton J. Stranks S.D. Johnston M.B. Watson T. Worsley D.A. Snaith H.J. Efficient, semitransparent neutral-colored solar cells based on microstructured formamidinium lead trihalide perovskite.J. Phys. Chem. Lett. 2015; 6: 129-138Crossref PubMed Scopus (162) Google Scholar, 31Della Gaspera E. Peng Y. Hou Q. Spiccia L. Bach U. Jasieniak J.J. Cheng Y.-B. Ultra-thin high efficiency semitransparent perovskite solar cells.Nano Energy. 2015; 13: 249-257Crossref Scopus (274) Google Scholar, 32Kwon H.-C. Kim A. Lee H. Lee D. Jeong S. Moon J. Parallelized nanopillar perovskites for semitransparent solar cells using an anodized aluminum oxide scaffold.Adv. Energy Mater. 2016; 6: 1601055Crossref Scopus (82) Google Scholar, 33Hanmandlu C. Chen C.-Y. Boopathi K.M. Lin H.-W. Lai C.-S. Chu C.-W. Bifacial perovskite solar cells featuring semitransparent electrodes.ACS Appl. Mater. Interfaces. 2017; 9: 32635-32642Crossref PubMed Scopus (43) Google Scholar To date, the highest reported AVT for MAPbI3-based TPVs is only 46% (for reference, the highest-AVT organic TPV is 66%).32Kwon H.-C. Kim A. Lee H. Lee D. Jeong S. Moon J. Parallelized nanopillar perovskites for semitransparent solar cells using an anodized aluminum oxide scaffold.Adv. Energy Mater. 2016; 6: 1601055Crossref Scopus (82) Google Scholar, 34Véron A.C. Zhang H. Linden A. Nüesch F. Heier J. Hany R. Geiger T. NIR-absorbing heptamethine dyes with tailor-made counterions for application in light to energy conversion.Org. Lett. 2014; 16: 1044-1047Crossref PubMed Scopus (56) Google Scholar By comparison, MAPbCl3 has an absorption cutoff at 410 nm, thereby enabling AVTs over 90%.1Lunt R.R. Theoretical limits for visibly transparent photovoltaics.Appl. Phys. Lett. 2012; 101: 043902Crossref Scopus (122) Google Scholar Previous theoretical work has shown that the ideal UV-only TPV band gap is between 430 and 440 nm.1Lunt R.R. Theoretical limits for visibly transparent photovoltaics.Appl. Phys. Lett. 2012; 101: 043902Crossref Scopus (122) Google Scholar To enhance the light-absorption range while avoiding influence on the visible light transmission, we have developed devices with doped compositions of MAPbCl3-xBrx to sensitively position absorption cutoffs between 410 and 440 nm. Bromide substitution of x = 20% results in the absorption of mixed halide perovskite film that extends ideally to 440 nm.22Comin R. Walters G. Thibau E.S. Voznyy O. Lu Z.-H. Sargent E.H. Structural, optical, and electronic studies of wide-bandgap lead halide perovskites.J. Mater. Chem. C. 2015; 3: 8839-8843Crossref Google Scholar Accordingly, UV-harvesting TPVs are developed with MAPbCl3 and MAPbCl2.4Br0.6 perovskite absorbers with the inverted device structure (Figure 1B). The UV-harvesting perovskite films are prepared by solution deposition via spin coating. Due to the fast crystallization rates and the limited solubility of MAPbCl3-based precursors in typical solvents, greater attention to the processing is required to form smooth perovskite layers with minimal haze and roughness to prevent shorting (Figures 2 and S1). To overcome this challenge, a vacuum-assisted solution deposition process is combined with methylamine gas post-treatment to obtain high-quality films that are highly transparent (Figures S2 and S3).35Liu D. Traverse C.J. Chen P. Elinski M. Yang C. Wang L. Young M. Lunt R.R. Aqueous-containing precursor solutions for efficient perovskite solar cells.Adv. Sci. 2018; 5: 1700484Crossref Scopus (51) Google Scholar, 36Zhou Z. Wang Z. Zhou Y. Pang S. Wang D. Xu H. Liu Z. Padture N.P. Cui G. Methylamine-gas-induced defect-healing behavior of CH3NH3PbI3 thin films for perovskite solar Cells.Angew. Chem. Int. Ed. 2015; 54: 9705-9709Crossref PubMed Scopus (355) Google Scholar, 37Li C. Pang S. Xu H. Cui G. Gas based synthesis and healing process toward upscaling of perovskite solar cells: progress and perspective.Sol. RRL. 2017; 1: 1700076Crossref Scopus (38) Google Scholar, 38Zhou Y. Padture N.P. Gas-induced formation/transformation of organic–inorganic halide perovskites.ACS Energy Lett. 2017; 2: 2166-2176Crossref Scopus (47) Google Scholar Individually, these methods (along with other standard methods) lead to rough, incomplete, hazy, and highly scattering films.39Zheng E. Yuh B. Tosado G.A. Yu Q. Solution-processed visible-blind UV-A photodetectors based on CH3NH3PbCl3 perovskite thin films.J. Mater. Chem. C. 2017; 5: 3796-3806Crossref Google Scholar, 40Kumawat N.K. Dey A. Kumar A. Gopinathan S.P. Narasimhan K.L. Kabra D. Band gap tuning of CH3NH3Pb(Br1–xClx)3 hybrid perovskite for blue electroluminescence.ACS Appl. Mater. Interfaces. 2015; 7: 13119-13124Crossref PubMed Scopus (296) Google Scholar, 41Adinolfi V. Ouellette O. Saidaminov M.I. Walters G. Abdelhady A.L. Bakr O.M. Sargent E.H. Fast and sensitive solution-processed visible-blind perovskite UV photodetectors.Adv. Mater. 2016; 28: 7264-7268Crossref PubMed Scopus (203) Google Scholar However, as shown in scanning electron microscopy images, MAPbCl3 and MAPbCl2.4Br0.6 films grown by this combined method show that both the pure chloride and bromide-doped perovskite films are smooth and uniform (Figures 3 and S4). Although the films do not exhibit grains as large as iodide perovskite films, strong diffraction peaks observed in the X-ray diffraction (XRD) patterns indicate good crystallization of the perovskite films with a single preferred orientation (Figure 3C). The diffraction peaks at 15.6°, 31.5°, and 48.0° correspond to the MAPbCl3 cubic phase indexed as the (100), (200), and (300), respectively. With 20% bromide doping, the diffraction peaks are shifted to slightly lower diffraction angles (15.4°, 31.1°, and 47.4°), which still are consistent with a cubic phase and also consistent with other literature reports.22Comin R. Walters G. Thibau E.S. Voznyy O. Lu Z.-H. Sargent E.H. Structural, optical, and electronic studies of wide-bandgap lead halide perovskites.J. Mater. Chem. C. 2015; 3: 8839-8843Crossref Google Scholar, 42Liu Y. Yang Z. Cui D. Ren X. Sun J. Liu X. Zhang J. Wei Q. Fan H. Yu F. et al.Two-inch-sized perovskite CH3NH3PbX3 (X = Cl, Br, I) crystals: growth and characterization.Adv. Mater. 2015; 27: 5176-5183Crossref PubMed Scopus (819) Google Scholar This shift indicates a larger lattice constant (a = 5.737 ± 0.008 Å) resulting from the larger Br atoms, compared with the pure MAPbCl3 (a = 5.677 ± 0.002 Å).Figure 3UV-Harvesting Perovskite Film PropertiesShow full caption(A and B) Surface morphology of smooth UV-harvesting perovskite films for (A) MAPbCl3 and (B) MAPbCl2.4Br0.6. The scale bar represents 1 μm.(C) XRD patterns of UV-harvesting perovskite films.(D) Photographs of halide perovskite films.View Large Image Figure ViewerDownload Hi-res image Download (PPT) (A and B) Surface morphology of smooth UV-harvesting perovskite films for (A) MAPbCl3 and (B) MAPbCl2.4Br0.6. The scale bar represents 1 μm. (C) XRD patterns of UV-harvesting perovskite films. (D) Photographs of halide perovskite films. Photographs of various compositions of halide perovskite films are shown in Figure 3D. The mixed halide perovskite film of MAPbCl2.4Br0.6 shows a light-yellow color due to the absorption cutoff at the edge of visible light (the visible range is determined as 435–670 nm for optimal theoretical TPVs with a corresponding visible transmission >99.5% and CRI >95),1Lunt R.R. Theoretical limits for visibly transparent photovoltaics.Appl. Phys. Lett. 2012; 101: 043902Crossref Scopus (122) Google Scholar, 3Traverse C.J. Pandey R. Barr M.C. Lunt R.R. Emergence of highly transparent photovoltaics for distributed applications.Nat. Energy. 2017; 2: 849-860Crossref Scopus (392) Google Scholar whereas the MAPbCl3, MAPbBr3, and MAPbI3 appear clear, orange, and dark brown, respectively. UV-harvesting TPVs were prepared with the architecture shown in Figure 1B. A thin (∼5 nm) poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT) layer was spin coated on indium tin oxide (ITO) substrate to act as the hole extraction layer. After the perovskite film was grown on the PEDOT layer, a thin C60 layer (20 nm) was deposited as the electron extraction layer.35Liu D. Traverse C.J. Chen P. Elinski M. Yang C. Wang L. Young M. Lunt R.R. Aqueous-containing precursor solutions for efficient perovskite solar cells.Adv. Sci. 2018; 5: 1700484Crossref Scopus (51) Google Scholar, 43Liu D. Wang Q. Traverse C.J. Yang C. Young M. Kuttipillai P.S. Lunt S.Y. Hamann T.W. Lunt R.R. Impact of ultrathin C60 on perovskite photovoltaic devices.ACS Nano. 2018; 12: 876-883Crossref PubMed Scopus (69) Google Scholar Bathocuproine (BCP) was then deposited on the fullerene followed by the top transparent electrode composed of Ag (5 nm)/Alq3 (60 nm), where the Alq3 acts as an optical layer to enhance the electrode visible transparency.34Véron A.C. Zhang H. Linden A. Nüesch F. Heier J. Hany R. Geiger T. NIR-absorbing heptamethine dyes with tailor-made counterions for application in light to energy conversion.Org. Lett. 2014; 16: 1044-1047Crossref PubMed Scopus (56) Google Scholar, 44Zhang H. Wicht G. Gretener C. Nagel M. Nüesch F. Romanyuk Y. Tisserant J.-N. Hany R. Semitransparent organic photovoltaics using a near-infrared absorbing cyanine dye.Sol. Energy Mater. Sol. Cells. 2013; 118: 157-164Crossref Scopus (38) Google Scholar The current–voltage (J-V) curves of TPV devices are shown in Figure 4A and Figure 4B shows the histogram of device PCEs. The champion pure chloride perovskite TPV device results in a PCE of 0.33%, with a JSC of 0.67 mA cm−2, a VOC of 1.18 V, and a fill factor (FF) of 41.1%. In contrast, the best mixed halide perovskite TPV cell exhibits a PCE of 0.52%, with a JSC of 0.92 ± 0.10 mA cm−2, a VOC of 1.26 V, and an FF of 44.9%. The JSC value is within error of the integrated external quantum efficiency (EQE) spectrum value of 0.83 mA cm−2. Due to the broader absorption range of MAPbCl2.4Br0.6 than MAPbCl3, it is expected that the JSC of MAPbCl2.4Br0.6 device will be higher than the MAPbCl3 device. Notably, the VOC and FF of the MAPbCl3 device is also lower than the MAPbCl2.4Br0.6 device. We infer that the greater difficulty in processing pin-hole-free MAPbCl3 films increases the charge recombination and results in the low photovoltaic parameters of the MAPbCl3 device (Figure 3A). The 50 separate MAPbCl3 devices (black) and 86 separate MAPbCl2.4Br0.6 devices result in an average PCE of 0.24% ± 0.04% and 0.42% ± 0.06%, respectively (Table 1). The low deviations of PCE value indicate the high reproducibility of these TPV devices.Table 1Summary of Device Parameters for Selective UV-Harvesting TPV DevicesDeviceVOC (V)JSC (mA cm−2)FF (%)PCE (%)Band gap (eV)AVT (%)CRIMAPbCl3 (best)1.11 ± 0.11 (1.18)0.61 ± 0.9 (0.67)35.8 ± 3.7 (41.1)0.24 ± 0.04 (0.33)3.0472.194.4MAPbCl2.4Br0.6 (best)1.13 ± 0.09 (1.26)0.85 ± 0.9 (0.92)43.5 ± 3.1 (44.9)0.42 ± 0.06 (0.52)2.8373.093.8The band gap was measured from optical absorption of halide perovskite films (Figure S10). Open table in a new tab The band gap was measured from optical absorption of halide perovskite films (Figure S10). In addition, device J–V curves were recorded under forward and reverse bias and were found to exhibit negligible hysteresis (Figure S5). Since the solar orientation varies with time and location, the light irradiation angle is an important factor for the performance of solar cells.9Boopathi K.M. Hanmandlu C. Singh A. Chen Y.-F. Lai C.S. Chu C.W. UV- and NIR-protective semitransparent smart windows based on metal halide solar cells.ACS Appl. Energy Mater. 2018; 1: 632-637Crossref Scopus (18) Google Scholar, 45Ding Y. Young M. Zhao Y. Traverse C. Benard A. Lunt R.R. Influence of photovoltaic angle-dependence on overall power output for fixed building integrated configurations.Sol. Energy Mater. Sol. Cells. 2015; 132: 523-527Crossref Scopus (12) Google Scholar, 46Young M. Traverse C.J. Pandey R. Barr M.C. Lunt R.R. Angle dependence of transparent photovoltaics in conventional and optically inverted configurations.Appl. Phys. Lett. 2013; 103: 133304Crossref Scopus (16) Google Scholar Thus, we also conducted angle-dependent tests of the UV-harvesting TPVs (Figure S6 and Table S1). Under normal incidence, the device shows the highest photocurrent and PCE as expected. As the irradiation angle decreases, the device JSC also gradually decreases. This decrease is in good agreement with the cosine law (Table S1) describing the reduction of the power density of a beam spreading over a larger area. Additionally, the performance of the TPVs was tested under illumination from the front and rear transparent electrode (Figure S7).9Boopathi K.M. Hanmandlu C. Singh A. Chen Y.-F. Lai C.S. Chu C.W. UV- and NIR-protective semitransparent smart windows based on metal halide solar cells.ACS Appl. Energy Mater. 2018; 1: 632-637Crossref Scopus (18) Google Scholar Since the Ag/Alq3 electrode can reflect and absorb parts of the UV spectrum, the photocurrent when illuminated from the top Ag/Alq3 side is slightly lower than the ITO side. The UV-harvesting TPVs generally exhibit good air stability. Figure S8 shows that the device can maintain ∼92% of the original efficiency when stored in ambient atmosphere. Preliminary UV light resistance was also evaluated. After 12 hr of continuous UV light irradiation, the device performance did not show any obvious change (Figure S9). These preliminary results are encouraging and lifetime test will be detailed in future work. The optical properties of UV-harvesting TPV devices were systematically investigated. Provided in Figure 5 is the absorption (A), reflection (R), transmission (T), and EQE spectra for each complete device. To satisfy the photon balance consistency check, we show that EQE + R + T < 100% everywhere in the spectrum.3Traverse C.J. Pandey R. Barr M.C. Lunt R.R. Emergence of highly transparent photovoltaics for distributed applications.Nat. Energy. 2017; 2: 849-860Crossref Scopus (392) Google Scholar As with the MAPbCl3 and MAPbCl2.4Br0.6 perovskite films, the corresponding TPV devices exhibit sharp absorption cutoffs in the UV while efficiently transmitting the visible light. Both UV-harvesting TPV devices exhibit exceptionally high AVTs of 72.1% and 73.0% for MAPbCl3 and MAPbCl2.4Br0.6, respectively (Table 1). This work is the first report for an AVT over 70% of any type of transparent photovoltaic cells with transparent active layers (Table S2). Only a few spatially segmented photovoltaic modules and transparent luminescent solar concentrators are reported to have AVTs over 70%, but the latter contains no transparent electrodes over the active area.3Traverse C.J. Pandey R. Barr M.C. Lunt R.R. Emergence of highly transparent photovoltaics for distributed applications.Nat. Energy. 2017; 2: 849-860Crossref Scopus (392) Google Scholar, 47Biancardo M. Taira K. Kogo N. Kikuchi H. Kumagai N. Kuratani N. Inagawa I. Imoto S. Nakata J. Characterization of microspherical semi-transparent solar cells and modules.Sol. Energy. 2007; 81: 711-716Crossref Scopus (43) Google Scholar, 48Yoon J. Baca A.J. Park S.-I. Elvikis P. Geddes Iii J.B. Li L. Kim R.H. Xiao J. Wang S. Kim T.-H. et al.Ultrathin silicon solar microcells for semitransparent, mechanically flexible and microconcentrator module designs.Nat. Mater. 2008; 7: 907Crossref PubMed Scopus (529) Google Scholar Shown in Figures 5C and 5D are photographs of complete perovskite TPV devices. While the individual films show a very slight yellow color, the addition of the top transparent cathode creates optical interference that results in a more neutral color rendering for the complete device. This can be seen in the reduction of the transmission curve and increase in the reflection curve past 650 nm. Nonetheless, because there is very little absorption of the active layer in any part of the visible spectrum, both devices show exceptionally high CRI of 94.4 and 93.8 for MAPbCl3 and MAPbCl2.4Br0.6, respectively (Table 1), referenced against the AM1.5 G solar spectrum (not an arbitrarily chosen correlated color temperature). CRI is another important factor for describing transparent devices, which quantitatively describes the color quality for reproducing the incident light source through a transparent medium and also utilized in the lighting industry. The high CRI value indicates the device has a high aesthetic quality and will have little impact on the color perception when placed over windows or displays. We note that it is a significant challenge to develop these large-band gap compositions in highly transparent perovskite TPVs as many of the standard fabrication techniques utilized for other low-band gap Br and I-based perovskite compositions generally lead to poor-quality films that are hazy and filled with pinholes when there is a high Cl content needed to get to large enough band gaps. The fabrication methods tested are summarized in Figures S11–S16. While hazy films are suitable for some photodetector and opaque applications, haziness is one of the most detrimental characteristics for integration in windows and displays, as the “view” is quickly compromised even at 10%–20% haziness. Moreover, this can significantly impact the yield and scalability by resulting in a large number of shorting pathways that was indeed observed experimentally for most of the processing techniques tested. By combining the vacuum-assist method with an MA gas treatment, this enabled smooth films with the highest device transparency and the best PCE. Although the anti-solvent techniques can also yield highly transparent TPV devices (Table S3) and can produce high device efficiency for traditional MAPbI3-xClx solar cells (Figure S17),49Jeon N.J. Noh J.H. Kim Y.C. Yang W.S. Ryu S. Seok S.I. Solvent engineering for high-performance inorganic–organic hybrid perovskite solar cells.Nat. Mater. 2014; 13: 897Crossref PubMed Scopus (5280) Google Scholar the efficiency of TPV devices prepared by the anti-solvent method was still low. We note that the AVT in our data is not correlated to the band gaps of the perovskite. One would typically expect the larger band gap to have a higher AVT. However, because both band gaps are larger than 2.8 eV (wavelength is smaller than 435 nm), the band gap does not play a dominant role since this is past the key visible cutoff. If the band gap was smaller than 2.8 eV (wavelength is larger than 435 nm), AVT would be strongly correlated with the band gap. Rather, the small differences in the AVT between the two perovskite compositions stem from the slight changes in the refractive index that sensitively impact the overall reflection and the parasitic absorption in the transparent electrodes. In this case, while the reflection spectrum is slightly larger for MAPbCl2.4Br0.6, this is more than compensated by lower parasitic absorption in the electrodes that leads to a slightly higher overall AVT.2Lunt R.R. Bulovic V. Transparent, near-infrared organic photovoltaic solar cells for window and energy-scavenging applications.Appl. Phys. Lett. 2011; 98: 113305Crossref Scopus (262) Google Scholar Moving forward, the full optimization of these devices with Cl-based perovskite compositions will require additional consideration of precursor development and processing strategies to improve the film formation quality. Despite enabling UV-harvesting TPV devices from halide perovskite semiconductors with nearly ideal optical properties (absorption cutoffs, color rendering, and transparencies), these first demonstrations are still limited by modest quantum efficiencies. This likely stems from the low solubility of chloride perovskite materials (limited to 30 wt % in dimethylformamide [DMF]) and dimethyl sulfoxide (DMSO), which leads to thin perovskite films (100–150 nm) that can harvest only up to 60%–70% of the UV light (Figures 5 and S4). This further indicates that the internal quantum efficiencies (IQE) are lower than 40%, which likely could be enhanced by increasing the film quality or crystalline grain size despite already having a strongly preferred crystalline alignment. Moreover, in Figures 5A and 5B, we see that the device reflection is greater than 10% in the range of 600–700-nm wavelengths, which reduces the transparency unnecessarily. Overall, with enhancement in the processing, the quantum efficiency could increase practically from 30% to 90%, FF from 45% to 70%, and VOC from 1.26 V to 2.0 V,50Lunt R.R. Osedach T.P. Brown P.R. Rowehl J.A. Bulović V. Practical roadmap and limits to nanostructured photovoltaics.Adv. Mater. 2011; 23: 5712-5727Crossref PubMed Scopus (160) Google Scholar it is then possible to achieve over 4% efficiency with this hybrid perovskite composition and approach the theoretical efficiency limits of 7% with the transparency nearing that of glass alone (See Table 2). This demonstrates that highly transparent TPV devices have great potential for smart windows and other applications in buildings and smart devices.33Hanmandlu C. Chen C.-Y. Boopathi K.M. Lin H.-W. Lai C.-S. Chu C.-W. Bifacial perovskite solar cells featuring semitransparent electrodes.ACS Appl. Mater. Interfaces. 2017; 9: 32635-32642Crossref PubMed Scopus (43) Google Scholar, 51Barile C.J. Slotcavage D.J. Hou J. Strand M.T. Hernandez T.S. McGehee M.D. Dynamic windows with neutral color, high contrast, and excellent durability using reversible metal electrodeposition.Joule. 2017; 1: 133-145Abstract Full Text Full Text PDF Scopus (142) Google Scholar, 52Li D.H.W. Lam T.N.T. Chan W.W.H. Mak A.H.L. Energy and cost analysis of semi-transparent photovoltaic in office buildings.Appl. Energy. 2009; 86: 722-729Crossref Scopus (174) Google Scholar, 53Murray J. Ma D. Munday J.N. Electrically controllable light trapping for self-powered switchable solar windows.ACS Photonics. 2017; 4: 1-7Crossref Scopus (66) Google ScholarTable 2Theoretical Efficiency Limits of UV-Only Harvesting TPV DevicesAbsorption Cutoff (nm)VOC (V)JSC (mA cm−2)CRIPCE (%)4002.691.4499.83.664302.492.5696.86.434402.433.1392.97.184502.373.6786.88.22Calculated based on the method from Lunt.1Lunt R.R. Theoretical limits for visibly transparent photovoltaics.Appl. Phys. Lett. 2012; 101: 043902Crossref Scopus (122) Google Scholar Open table in a new tab Calculated based on the method from Lunt.1Lunt R.R. Theoretical limits for visibly transparent photovoltaics.Appl. Phys. Lett. 2012; 101: 043902Crossref Scopus (122) Google Scholar In summary, we have demonstrated UV-harvesting TPV devices based on the platform of halide perovskite light absorbers. This study indicates the TPV devices can selectively harvest UV light with ideally positioned absorption cutoffs to minimize visual impact while providing a pathway to maximizing performance. Based on mixed composition halide perovskites, we show TPV devices with ideally tuned band gap (around 435–440 nm) that exhibit up to 0.52% PCE with AVT of 73.0% and a CRI over 93. The high AVT of 73% makes these UV-harvesting TPV devices one of the highest-performing highest-AVT photovoltaic cells reported to date. This study reveals that the exceptional potential of perovskite materials can effectively translate to TPV devices that will see rapid gains as quantum efficiencies are improved. Ultimately, these types of technologies will be an important compliment to other near-infrared harvesting TPVs to rapidly reach both single-junction and multi-junction TPV efficiency limits.