CROSS-REFERENCE TO RELATED APPLICATIONS

The present application claims priority to U.S. Provisional Application No.

63/349,356; filed June 6, 2022, which is incorporated by reference herein in its entirety.

BACKGROUND OF THE INVENTION

Redshifting the solar spectrum entering a solar cell by creating near-infrared (NIR) photons from ultraviolet (UV) and blue photons via luminescence downconversion can increase the solar cell’s efficiency. Shifting the UV and blue spectrum to NIR reduces thermal losses and the recombination of electron-hole pairs generated from shallow light absorption near interfaces. Ytterbium (Yb) is a well-known luminophore for solar spectral shifting because the Yb ³⁺ emission via ² F _5/2 ² F _7/2 electronic transition at 1.24 eV is close to the bandgap of silicon (~ 1. 1 eV) and copper indium gallium diselenide (CIGS) (1.0-1.2 eV) (Gloeckler, M.; Sites, J. R. Band-gap grading in Cu(In,Ga)Se2 solar cells. J. Phys. Chem. Solids. 2005, 66, 1891-1894) Typically Yb is doped into a host, which absorbs in the UV and visible regions of the electromagnetic spectrum and transfers energy to the Yb ³⁺ , exciting it from the ² F _7/2 ground state to the ² F _5/2 state. The excited Yb ^3- emits NIR photons at ~1.24 eV upon relaxation. Thus, depositing a layer of a Yb-doped film with high photoluminescence quantum yield (PLQY) on top of a silicon solar cell can improve its solar cell efficiency by modifying the incident solar spectrum. Ideally, all blue (UV) photons are converted to NIR photons, and the maximum possible photoluminescence quantum yield (PLQY) is 100%.

However, there is another possibility. If the host bandgap is greater than twice the electronic transition, the energy transfer from the host to Yb ³⁺ can be via quantum cutting, a process wherein one UV-blue photon is converted to two NIR photons. In this case, PLQY can be >100% with a maximum of 200%. Indeed, Yb- doped CsPbX ₃ (X=C1, Br) have been shown to exhibit quantum cutting with PLQY as high as 190% (Pan, G. et al. Nano Lett. 2017, 17, 8005-8011; Milstein, T, et al. Nano Lett. 2018, 18, 3792-3799; Kroupa, D. M. et al. ACS Energy Lett. 2018, 3, 2390-2395). However, lead is toxic, and NIR PLQY from CsPbX ₃ decreases at high photon fluence. In the search for non-toxic alternatives to CsPbX ₃ , double and bismuth-based perovskites are emerging as promising hosts because they have high absorption coefficients and tunable bandgaps in the visible range (Creutz, S. E. et al., Chem. Mater. 2019, 31, 4685-4697; Creutz, S. E. et al., Nano Letters. 2018, 18 (2), 1118-1123). Specifically, Yb doping of Cs ₃ Bi ₂ Br ₉ , Cs ₂ AgInCl ₆ , and Cs ₂ AgBiBr ₆ (Tran, M.N. et al. J. Mater. Chem. A, 2021,9, 13026-13035; Lee. W. et al. J. Phys. Chem. C. 2019, 123 (4), 2665-2672; Schmitz, F. et al. J. Phys. Chem. Lett. 2020, 11, 8893-8900.) has been reported. Unfortunately, 28%, the highest PLQY from Yb-doped Cs ₂ AgBiBr ₆ thin film (Schmitz, F. et al. J. Phys. Chem. Lett. 2020, 11, 8893-8900), is still much lower than the minimum PLQY estimated to realize any increase in solar cell efficiencies (69% for typical Si solar cells and 67% for typical CIGS solar cells).

There remains a need in the art for downconverting and quantum cutting materials with high PLQY that, when placed on top of solar cells, can increase their power conversion efficiencies, and improve their lifetimes by reducing UV penetration into the solar cell and reducing solar cell heating. The present invention addresses this unmet need.

SUMMARY OF THE INVENTION

In one aspect, the present invention relates to a thin film comprising a Yb- doped double perovskite, wherein the double perovskite has the formula M ₂ AYb _x B(i- _X )X6; wherein each occurrence of M independently represents Cs or Rb; A represents Ag or Cu; B represents Bi, In, Sb, or Ga; x has a value between 0.01 and 0.20; and each X independently represents F, Cl, Br, or I. In one embodiment, the double perovskite has the formula M ₂ AYbxB(i.x)Cl(6-y)Br _y wherein y is an integer between 0 and 6. In one embodiment, the double perovskite has the formula Cs ₂ AgYbxBi(i- _X )Cl(6-y)Br _y . In one embodiment, the double perovskite has the formula Cs ₂ AgYb _x B(1- _X )Br6. In one embodiment, the value of x is between 0.05 and 0.10. In one embodiment, the value of x is between 0.06 and 0.09. In one embodiment, a photoluminescence quantum yield of the thin film is at least 45%.

In one aspect, the present invention relates to a solar cell comprising the Yb-doped double perovskite thin film. In one embodiment, the solar cell is selected from the group consisting of a silicon solar cell and a copper indium gallium selenide solar cell.

In one aspect, the present invention relates to a method of formulating a thin fdm, the method comprising the steps of: providing a substrate; providing a source of BX ₃ ; providing a source of YbX;; providing a source of AX; depositing BX ₃ , YbX ₃ , and AX on the substrate, wherein a ratio of molar flux of YbX3 to molar flux of BX ₃ is between 0.01 and 0.15; providing a source of MX; depositing MX on the substrate to provide a perovskite mixture; and annealing the mixture to provide a Yb-doped double perovskite thin film; wherein M represents Cs or Rb; A represents Ag, Cu, or Au; B represents Bi, In, Sb and Ga; and each X independently represents F, Cl, Br, or I. In one embodiment, the method further comprises the step of the step of ball milling at least one of BX ₃ , YbX ₃ , AX, and MX to produce a powder. In one embodiment, the step of depositing MX on the substrate further comprises the step of increasing the temperature of the substrate. In one embodiment, the temperature of the substrate is increased from about 30 °C to about 83 °C during the deposition of MX. In one embodiment, the ratio of molar flux of YbX ₃ to BX ₃ is between about 0.06 and 0.09.

In one embodiment, BX ₃ represents BiX ₃ ; AX represents AgX; MX represents CsX; and each X independently represents Br or Cl. In one embodiment, BX ₃ represents BiBr ₃ ; YbX ₃ represents YbBr ₃ ; AX represents AgBr; and MX represents CsBr. In one embodiment, the BiBr ₃ is deposited at an evaporation rate of about 1.5 A/s. The invention also relates to a thin film produced using these methods.

In one aspect, the present invention relates to a thin film comprising a Yb- doped double perovskite produced with a method comprising the steps of: providing a substrate; providing a source of BX ₃ ; providing a source of YbX ₃ ; providing a source of AX; depositing BX ₃ , YbX ₃ . and AX on the substrate, wherein a ratio of molar flux of YbX ₃ to molar flux of BX ₃ is between 0.01 and 0.15; providing a source of MX; depositing MX on the substrate to provide a perovskite mixture; and annealing the mixture to provide a Yb-doped double perovskite thin film; wherein M represents Cs or Rb; A represents Ag, Cu, or Au; B represents Bi, In, Sb and Ga; and each X independently represents F, Cl, Br, or I. In one embodiment, the ratio of molar flux of YbX ₃ to BX ₃ is between about 0.06 and 0.09. BRIEF DESCRIPTION OF THE DRAWINGS

The following detailed description of preferred embodiments of the invention will be better understood when read in conjunction with the appended drawings. For the purpose of illustrating the invention, there are shown in the drawings embodiments which are presently preferred. It should be understood, however, that the invention is not limited to the precise arrangements and instrumentalities of the embodiments shown in the drawings.

Figure l is a schematic diagram of an exemplary photovoltaic device.

Figure 2 depicts the structures of cubic CsPbBr ₃ (#221, Pm3m), cubic Cs2AgBiBr ₆ (#225, Fm 3 m), trigonal Cs3Bi ₂ Br ₉ (#164, P 3 ml), and Ytterbium ions substituting Ag ⁺ and Bi ³⁺ ions in the octahedra in Cs2AgBiBr ₆ .

Figure 3 is a plot of X-ray diffraction from Yb-doped Cs2AgBiBr ₆ films and simulated powder diffraction patterns of AgBr and Cs?AgBiBr ₃ , for comparison. All data and images are from films annealed at 300 °C for one hour.

Figure 4 is a plot of X-ray diffraction (XRD) patterns as-deposited Yb- doped Cs2AgBiBr ₆ films. XRD of films doped with 5-14% Yb show multiple weak peaks that do not belong to Cs2AgBiBr ₆ . Simulated (using VESTA software) XRD patterns of AgBr, Cs2AgBiBr6, and Cs3Bi2Br9 using CIF files are shown for comparison.

Figure 5 is a plot of Raman spectra of as-deposited Yb-doped Cs2AgBiBr ₆ films. Films doped with 5-14% Yb have Raman peaks at 197 and 170 cm' ¹ , which belong to an impurity phase, Cs3Bi2Br9. This impurity phase peaks match with the XRD of Cs3Bi2Br9.

Figure 6 depicts Scanning electron micrographs (SEMs) of Yb-doped (10% Yb) Cs2AgBiBr ₆ films annealed in a nitrogen-filled glove box under different conditions showing the effects of annealing temperature and annealing duration. The average composition of the larger crystal domains (e.g., large irregular domains in the first image and elongated crystals in the second and third images is 57.2% Br, 16.9% Ag, 21.4% Cs, 0.3% Yb, and 4.2% Bi, suggesting thatthe impurity is a Cs-Ag-Br phase, possibly CsAgBr2 (#63, Cmcm). The small Bi signal detected is likely coming from the Cs2AgBiBr ₆ surrounding or beneath these regions. Figure 7 is a plot of Raman scattering of Cs ₂ AgBiBm as-deposited films and films annealed at different temperatures. When annealed at 350 °C, CszAgBiBr ₆ decomposes, possibly to CssBiBr ₆ , as shown in the Raman spectra.

Figure 8 is a plot of XRD from Cs2AgBiBr ₆ as-deposited films and films annealed at different temperatures.

Figure 9 depicts SEM images of Cs2AgBiBr ₆ films annealed at different temperatures for 1 hour: as-deposited, 250 °C; 300 °C, and 350 °C. Films decompose when annealed at 350 °C.

Figure 10 depicts Raman spectra of undoped and Yb-doped CsiAgBiBr ₆ films. All data and images are from films annealed at 300 °C for one hour.

Figure 11 is a series of representative SEM images of Yb-doped (8%) Cs2AgBiBr ₆ films at different magnifications. All data and images are from films annealed at 300 °C for one hour.

Figure 12 is a series of SEM images Cs2AgBiBr ₆ films doped with 0-14% Yb annealed at 300 °C for 1 hour: 0% Yb, 3% Yb, 5% Yb, 8% Yb, 10% Yb, and 14% Yb.

Figure 13 depicts the absorbance and photoluminescence of Cs2AgBiBr& thin film annealed at 300 °C for 1 hour. The inset is the PL from an as-deposited Cs2AgBiBr ₆ thin film.

Figure 14 depicts XRD patterns from Cs2AgBiCl _y Br6- _y thin films, and simulated XRD patterns of Cs2AgBiC16 and Cs2AgBiBr ₆ . As dotted lines show, XRD peaks are shifted to higher 20 values with increasing amounts of chlorine, suggesting that two halide ions are mixed throughout the films. XRD Cs ₂ AgBiBr ₂ C1 ₄ contains impurity peaks of Cs ₂ AgBiC1 ₆ , suggesting phase segregation in this film after annealing.

Figure 15 is a plot of normalized absorbance from Cs2AgBiClyBr ₆ -y thin films thin film annealed at 300 °C for 1 hour. Thin-film interference fringes and background due to scattering were subtracted from the absorption spectra.

Figure 16 is a plot of normalized photoluminescence from Cs2AgBiClyBr ₆ - _y thin films thin film annealed at 300 °C for 1 hour. Thin-film interference fringes and background due to scattering were subtracted from the absorption spectra.

Figure 17 is a plot demonstrating that visible photoluminescence from Cs2AgBiBr ₆ thin film decreases after the film is annealed at 300 °C for one hour. Figure 18 is a depiction of energy transfer processes in Yb-doped

Cs ₂ AgBiBr ₆ .

Figure 19 is a plot of absorption and excitation (λ _em = 997 nm) spectra of Cs2AgBiBr ₆ thin fdm doped with 8% Yb and annealed at 300 °C for 1 hour. NIR PLQY at different excitation wavelengths is also plotted.

Figure 20 is a plot of orange and NIR PL from Cs ₂ AgBiBr ₆ thin film doped with 8% Yb and annealed at 300 °C for 1 hour.

Figure 21 is a plot of the orange emission from Cs ₂ AgBiBr ₆ thin films doped with 3-14% Yb. Emissions at all Yb levels are weak, as the noise suggests. However, the orange emission from 10% Yb-Cs ₂ AgBiBr ₆ film was particularly weak, and it is difficult to distinguish between the film’s PL and stray light, so its PL is not included here.

Figure 22 is a plot showing the dependence NIR PLQY of Cs2AgBiBr6 films doped with 3-14% Yb annealed under different conditions.

Figure 23 is a plot of the PLQY of Cs2AgBiBr ₆ films doped with 3 to 14% Yb and annealed at 300 °C for one hour. PL intensity of each film is scaled with PLQY.

Figure 24 depicts the long term stability of the PLQY of Cs2AgBiBr6 films doped with 8% Yb and annealed at 300 °C for one hour. The intensity is scaled with the corresponding quantum yield. PL was excited at 420 nm (10 nm bandwidth).

Figure 25 is a plot showing how external quantum efficiency of a typical CIGS solar cell changes after a layer of down-conversion materials is deposited on top of the solar cell. (Friedlmeier, T. M. et al., IEEE 42 ^nd Photovoltaic Specialist Conference. 2015, 1-3) For example, with 82.5% of PLQY, the solar cell’s efficiency can increase from 20.4 to 20.7% for CIGS.

Figure 26 is a plot showing how the external quantum efficiency of a typical silicon solar cell (Mazzarella, L. et al., Appl. Phys. Lett. 2015, 106, 023902) changes after a layer of down-conversion materials is deposited on top of the solar cell. For example, with 82.5% of PLQY, the solar cell’s efficiency can increase from 20.3 to 20.6% for Si.

Figure 27 shows how Ytterbium ions substitute Bi ³⁺ ions and Ag ⁺ ions in the octahedra in Cs2AgBiBr ₆ (#225, Fm3m). The alternating colors depict the alternating [AgBr ₆ ] ⁵ ' and [BiBr ₆ ] ³ ' octahedra. Figure 28 is a plot of NTR emission spectrum of a 8% Yb-doped Cs2 AgBiBr ₆ , thin fdm.

Figure 29 is a plot of temperature profdes during depositions and the corresponding photoluminescence quantum yield of the 8% Yb-doped Cs2AgBiBr ₆ thin films synthesized under each temperature condition.

Figure 30 depicts the XRD of as-deposited films before annealing without substrate temperature control (no temp control), with substrate temperature controlled at 30 °C and 48 °C and substrate temperature ramped to 70 °C during CsBr deposition. Simulated XRD patterns of the target product, Cs ₂ AgBiBr ₆ ,. the precursor, AgBr, a possible intermediate, CsAgBr ₂ , and a common bismuth perovskite phase, Cs ₃ Bi2Br9 are shown for comparison. The dashed lines mark the diffraction peaks of CsAgBr ₂ and AgBr, both of which are present in the as-deposited films regardless of the deposition temperature or the temperature control mode.

Figure 31 depicts X-ray diffraction patterns (shifted for clarity) of 8% Yb- doped Cs ₂ AgBiBr ₆ thin films deposited at different substrate temperatures (T _s ) and the simulated reference pattern of Cs2AgBiBr&.

Figure 32 depicts XRD pattern of Cs2AgBiBr ₆ films doped with 8% Yb deposited at 75 °C. Calculated XRD patterns for Cs2AgBiBr ₆ and Cs ₃ BiBr ₆ are shown for reference.

Figure 33 depicts SEM images of Cs2AgBiBr ₆ films doped with 8% Yb deposited at 75 °C. Impurity phases appear as rod-like morphologies. The EDS composition of these phases is 7.4% Bi, 20.6% Cs, 65.1% Br, 5.8% Ag, and 1.1% Yb. The ratio Cs:Bi:Br is 2.8: 1.0:8.8, close to that expected in Cs ₃ BiBr ₆ .

Figure 34 depicts SEM images of Cs2AgBiBr ₆ films doped with 8% Yb deposited in: no-temperature-control mode, constant-temperature mode at 30 °C, constanttemperature mode at 48 °C, constant-temperature mode at 75 °C, ramping-temperature mode to 70 °C and ramping-temperature mode to 83 °C. All films were annealed at 300 °C.

Figure 35 depicts PLQY of 8% Yb-doped Cs2AgBiBr ₆ films deposited under the ramping condition to 83 °C with varying BiBr ₃ evaporation rates. PLQYs are average values of multiple films synthesized under the same conditions. Figure 36 depicts initial PLQY of 8% Yb-doped CsiAgBiBr ₆ fdms deposited under the ramping condition to 83 °C with varying BiBr ₃ evaporation rates and after exposed to air for a few days. PLQYs are not average values and come from one set of fdms.

Figure 37 depicts XRD of 8% Yb-doped Cs ₂ AgBiBr ₆ fdms deposited under the ramping condition to 83 °C with varying BiBr ₃ evaporation rates.

Figure 38 is a plot of PLQY stability of a Cs2AgBiBr ₆ fdm doped with 8% Yb after removal from the glove box and exposure to air. The Cs2AgBiBr ₆ fdm was deposited in the ramping-temperature mode as the substrate temperature was increased from 30 °C to 83 °C during CsBr evaporation. The BiBr ₃ evaporation rate was 1.5 A/s. Day 0 is when the fdm was deposited, annealed, and removed from the glovebox.

Figure 39 depicts the excitation spectrum of 8% Yb-doped Cs2AgBiBr ₆ thin fdm synthesized with the BiBr ₃ evaporation rate of 1.5 A/s and the ramping temperature to 83 °C. PLQY corresponding to each excitation wavelength is included. PLQY measurements in this graph are measured after the fdm is exposed to air for 1 month.

Figure 40 depicts absorption and excitation spectra of 8% Yb-doped Cs2AgBiBr ₆ thin fdm synthesized with the BiBr ₃ evaporation rate of 1.5 A/s and the ramping temperature to 83 °C.

Figure 41 depicts SEM images of Cs2AgBiBr ₆ fdms doped with 8% Yb deposited using the temperature-ramping mode, where the substrate temperature was increased from 30 °C to 83 °C. The different images correspond to fdms deposited using varying BiBr ₃ evaporation rates. All fdms were annealed at 300 °C for one hour.

Figure 42 depicts XRD of 8%Yb-doped Cs2AgBiC14Br ₂ fdms deposited with stoichiometric or 50% excess BiC1 ₃ . Films are co-deposited or sequentially deposited, and the substrate temperature is ramped as described in the text during the depositions. XRD reference patterns of Cs2AgBiC16 and Cs2AgBiBr ₆ are included for comparison. NIR PLQY values of the fdms are also shown. XRDs are offset for clarity.

Figure 43 depicts SEM images of 8%Yb-doped Cs2AgBiC14Br ₂ fdms deposited with stoichiometric ratio and 50% excess BiCh fluxes. Films are co- or sequentially deposited, and the substrate temperature is ramped during the deposition as described in the text. Some SEMs show crystallites that appear different than the background films and are presumed to be impurity phases. Compositions of the impurity phases, measured using EDS, vary and differ from the homogeneous background film. The EDS shows the presence of Bi-deficient phases in small volume fractions not detectable using XRD. For example, the triangular crystal in Fig. 43d comprises 20.8% Cs, 17.3% Ag, 7.6% Bi, 34.8% Cl, 18.8% Br, and 0.6% Yb: Bi concentration is lower than that expected in stoichiometric Cs2AgBiC14Br2 (20% Cs, 10% Ag, 10% Bi, 40% Cl, and 20% Br).

Figure 44 depicts XRD patterns of Yb-doped Cs2AgBiC14Br2 films sequentially deposited with 50% excess BiC1 ₃ and different concentrations of Yb from 3 to 20%. NIR PLQY for each film is also included.

Figure 45 depicts SEM images of Yb-doped Cs2AgBiC14Br2 films sequentially deposited with 50% excess BiC13 and different concentrations of Yb from 3 to 20%.

Figure 46 depicts XRD patterns and Raman spectra of Yb-doped Cs2AgBiBr ₆ powders and after annealing. The reference patterns of Cs2AgBiBr ₆ is also included for comparison.

Figure 47 depicts the near-infrared emission curve of Yb-doped Cs2AgBiBr ₆ annealed powders with PLQY of 51%.

Figure 48 depicts XRD patterns and Raman spectra of Yb-doped Cs2AgBiBr ₆ as-deposited and annealed films. The reference XRD pattern of Cs2AgBiBr ₆ and AgBr are included. The reference Raman pattern of Cs2AgBiBr ₆ and Cs3Bi2Br9 are also included.

Figure 49 depicts the near-infrared emission curve of Yb-doped Cs2AgBiBr ₆ annealed film.

DETAILED DESCRIPTION

It is to be understood that the Figures and descriptions of the present invention have been simplified to illustrate elements that are relevant for a clear understanding of the present invention, while eliminating, for the purpose of clarity, many other elements found in photovoltaic devices. Those of ordinary skill in the art may recognize that other elements and/or steps are desirable and/or required in implementing the present invention. However, because such elements and steps are well known in the art and because they do not facilitate a better understanding of the present invention, a discussion of such elements and steps is not provided herein. The disclosure herein is directed to all such variations and modifications to such elements and methods known to those skilled in the art.

As used herein, each of the following terms has the meaning associated with it in this section. Unless defined otherwise, all technical and scientific terms used herein generally have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

The articles “a” and “an” are used herein to refer to one or to more than one (i.e. to at least one) of the grammatical object of the article. By way of example, “an element” means one element or more than one element.

As used herein, the term “about” will be understood by persons of ordinary skill in the art and will vary to some extent depending on the context in which it is used. As used herein when referring to a measurable value such as an amount, a temporal duration, and the like, the term “about” is meant to encompass variations of ±20% or ± 10%, more preferably ±5%, even more preferably ±1 %, and still more preferably ±0.1 % from the specified value, as such variations are appropriate to perform the disclosed methods.

As used herein, the term “substrate” refers to a structural surface beneath a layered material or coating (e.g., deposited material).

Throughout this disclosure, various aspects of the invention can be presented in a range format. It should be understood that the description in range format is merely for convenience and brevity and should not be construed as an inflexible limitation on the scope of the invention. Accordingly, the description of a range should be considered to have specifically disclosed all the possible sub-ranges as well as individual numerical values within that range. For example, description of a range such as from 1 to 6 should be considered to have specifically disclosed sub-ranges such as from 1 to 3, from 1 to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3 to 6 etc., as well as individual numbers within that range, for example, 1, 2, 2.7, 3, 4, 5, 5.3, and 6. This applies regardless of the breadth of the range. Description

The present invention is based in part on the unexpected discovery that Yb- doped double perovskite films can be used a down-converting coating on solar cells.

Methods of Formulating a Thin Film

In one aspect, the present invention relates to a method of formulating a Yb- doped double perovskite thin film, the method comprising the steps of providing a substrate; providing a source of BXi, providing a source of YbXi; providing a source of AX; depositing BXs, YbX ₃ , and AX on the substrate, wherein a ratio of molar flux of YbXi to molar flux of BX ₃ is between 0.01 and 0.15; providing a source of MX; depositing MX on the substrate to provide a perovskite mixture, and annealing the mixture to provide a Yb-doped double perovskite thin film; wherein M represents Cs or Rb; A represents Ag, Cu, or Au; B represents Bi, In, Sb and Ga; and each X independently represents F, Cl, Br, or I.

In one embodiment, B represents Bi. In one embodiment, M represents Cs. In one embodiment, A represents Ag. In one embodiment, each X represents Cl or Br.

In one embodiment, the rate of deposition of the various components can be tuned, which may affect resulting photoluminescence characteristics. In one embodiment, the temperature of the substrate and/or the temperature of the deposition apparatus may be tuned, which may affect the resulting photoluminescence characteristics. In one embodiment, the BX ₃ is deposited at a rate between 1.0 A/s and 2.0 A/s. In one embodiment, the BX ₃ is deposited at a rate between 1.0 A/s and 1.8 A/s. In one embodiment, BX ₃ is deposited at a rate of about 1.5 A/s. In one embodiment, BX ₃ represents BiBr ₃ .

In one embodiment, the method comprises the steps of providing a substrate, depositing BiBr ₃ , YbBr ₃ , and AgBr on the substrate; depositing CsBr on the substrate; and annealing the thin film.

As contemplated herein, the percent doping of Yb in the double perovskite is meant to indicate the ratio of the molar flux of the Yb-containing precursor YbX3 to that of the BX ₃ precursor. Thus, in one embodiment, Yb can be varied between 0% and 14% by controlling the YbBr ₃ evaporation rate. For example, in some embodiments, to deposit a 3% Yb-doped CsiAgBiBr ₆ , the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is set to 0.03. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is between 0.01 and 0.10. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is between 0.05 and 0.10. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is between 0.06 and 0.09. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is between 0.075 and 0.085. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.010. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.015. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.020. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.025. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.030. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.035. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.040. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.045. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.050. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.055. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.060. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.065. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.070. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.075. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.080. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.085. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.090. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.095. In one embodiment, the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is about 0.100.

In one embodiment, the BX ₃ , YbX ₃ , and AX are deposited simultaneously.

In one embodiment, the step of depositing BX ₃ , YbX3, and AgX comprises the step of evaporating sources BX ₃ , YbX ₃ , and AX in the presence of a substrate, wherein the temperature of the substrate is lower than the temperature of the sources of BX ₃ , YbX ₃ , and AX. Tn one embodiment, the temperature of the substrate is constant throughout the deposition process. In one embodiment, the temperature of the substrate increases throughout the deposition process. In one embodiment, the step of depositing MX on the substrate further comprises the step of increasing the temperature of the substrate. In one embodiment, the temperature of the substrate is increased from about 30 °C to about 83 °C over the course of the MX deposition.

In one embodiment, one or more of the precursors BX ₃ , YbX ₃ , AX, and MX are pre-milled to a homogenous powder prior to deposition. In one embodiment, the milling step reduces particle size and affords more efficient physical vapor deposition. Equipment that may be used for precursor milling includes but not limited to a ball mill, a roller mill, a hammer mill, and a jet mill.

In one embodiment, one or more of the precursors are pre-milled in a ball mill. Ball mills are used in the present invention to obtain a homogenous powder. In one embodiment, the ball mill comprises a fixed cylindrical vessel such as those known to the person of ordinary skill in the art. The axis of the cylinder can be both horizontal and have a small angle with the horizontal. In one embodiment, the ball mill is partially filled with balls. Abrasive media are made of ceramic or zirconia (beads between 3 mm to 10 mm). The inner surface of the cylinder is normally crossed out with an abrasion resistant material such as manganese steel. The ball mill rotates around a horizontal axis, partially filled with the material to be ground plus the abrasive medium, an internal cascade effect reduces the material to a fine powder.

In one embodiment, the centrifugal force in the ball mill is extremely high, resulting in very short grinding times. Ball mills have the advantage of powerful and fast crushing down to the submicron range, in addition the energy and speed are adjustable so that reproducible results are guaranteed. In one embodiment, the precursors are dry-milled. In one embodiment, the precursors are wet-milled (i.e., milled in the presence of water). In one embodiment, the precursors are milled in the presence of a suitable solvent for the desired powder properties. Thin Films

In one aspect, the present invention relates in part to thin films comprising a double perovskite and at least one dopant, wherein the at least one dopant comprises Yb; and wherein the double perovskite has the formula M2AYb _x B(i- _X )X6; wherein each M independently Cs or Rb; A represents a monovalent ion such as Ag, Cu, or Au; B represents a trivalent ion such as Bi, In, Sb and Ga; x has a value between 0.01 and 0.20; and each X independently represents F, Cl, Br, or I. In one embodiment, neither the double perovskite nor the dopant comprises lead.

In one embodiment, the double perovskite has the formula M2AYb _x B(i-x)Cl(6-y)Br _y wherein y is a number between 0 and 6. In one embodiment, the double perovskite has the formula Cs2AgYb _x Bi(i-x)Cl(6-y)Br _y . In one embodiment, the double perovskite has the formula Cs2AgYbxBi(i. _X )Br6. In one embodiment, the double perovskite has the formula Cs2AgBiC1 ₆ . In one embodiment, the double perovskite has the formula Cs2AgYbxBi(i- _X )Br6. In one embodiment, the double perovskite has the formula Cs2AgYbxB(i- _X )C14Br2.

In one embodiment, the thin film red-shifts incident UV and blue radiation to near-infrared radiation. In one embodiment, the photoluminescence quantum yield of the thin film is at least 45%.

In one embodiment, the Yb atoms in the double perovskite replace positions held by B atoms, such as Bi atoms. Thus, the Yb content is reported in percent of the B lattice positions displaced in the M2ABX6 structure, or M2AYb _x B(i- _X )X6. In one embodiment, x has a value between 0.01 and 0.15. In one embodiment, x has a value between 0.01 and 0.10. In one embodiment, x has a value between 0.05 and 0.10. In one embodiment, x has a value between 0.06 and 0.09. In one embodiment, x has a value of about 0.010. In one embodiment, x has a value of about 0.015. In one embodiment, x has a value of about 0.020. In one embodiment, x has a value of about 0.025. In one embodiment, x has a value of about 0.030. In one embodiment, x has a value of about 0.035. In one embodiment, x has a value of about 0.040. In one embodiment, x has a value of about 0.045. In one embodiment, x has a value of about 0.050. In one embodiment, x has a value of about 0.055. In one embodiment, x has a value of about 0.060. In one embodiment, x has a value of about 0.065. In one embodiment, x has a value of about 0.070. In one embodiment, x has a value of about 0 075. Tn one embodiment, x has a value of about 0.080. In one embodiment, x has a value of about 0.085. In one embodiment, x has a value of about 0.090. In one embodiment, x has a value of about 0.095. In one embodiment, x has a value of about 0.100.

In another aspect, the present invention relates to a thin film composition comprising a Yb-doped double perovskite, wherein the double perovskite has the formula M2ABX6; wherein each M independently Cs or Rb; A represents a monovalent ion such as Ag, Cu, or Au; B represents a trivalent ion such as Bi, In, Sb and Ga; and each X independently represents F, Cl, Br, or I; and wherein the composition further comprises a 1% to 15% of a Yb dopant. As contemplated herein, the percent doping of Yb (or, the Yb content) in the double perovskite is meant to indicate the ratio of the molar flux of the Yb-containing precursor YbX ₃ to that of the BX ₃ precursor during the synthesis of the thin film composition. Thus, in one embodiment, Yb can be varied between 0% and 15% by controlling the YbBr ₃ evaporation rate. For example, in some embodiments, to deposit a 3% Yb-doped Cs2AgBiBr ₆ , the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux is set to 0.03. In one embodiment, the Yb content is between 5% and 10%. In one embodiment, the Yb content is about 8%. In one embodiment, neither the double perovskite nor the dopant comprises lead.

In one embodiment, the double perovskite has the formula M2ABCl(6- _y )Br _y wherein y is a number between 0 and 6. In one embodiment, the double perovskite has the formula Cs2AgBiCl(6-y)Br _y . In one embodiment, the double perovskite has the formula Cs2AgBiBr ₆ . In one embodiment, the double perovskite has the formula Cs2AgBiC16. In one embodiment, the double perovskite has the formula Cs2AgBiChBr4. In one embodiment, the double perovskite has the formula Cs2AgBiC14Br2.

In one aspect, the present invention relates to a solar cell comprising a thin fdm disclosed herein. In a solar cell, the active layer converts photons (incident light) to excitons, which comprise an electron and a hole. The potential between the electrodes drives the electrons to the cathode and the holes to the anode, thereby generating an electric current. In one embodiment, the solar cell is a silicon solar cell. In one embodiment, the solar cell is a copper indium gallium selenide (CIGS) solar cell. Photovoltaic devices

The present invention relates in part to photovoltaic devices comprising a thin fdm of the present invention. Referring to Figure 1, exemplary photovoltaic device 100 is shown. In some embodiments, device 100 may include: down-converting thin fdm 110, first electrode 120, optional charge transporting layer 130, active layer 140, optional charge transporting layer 150, second electrode 160, and optional substrate 170. In some embodiments, the device may be encapsulated with glass from the top and the bottom, and the down-converting thin film 110 may be formed on the front or back of the top glass cover.

In some embodiments, first electrode 120 is a cathode, transporting layer 130 is an electron transporting layer, transporting layer 150 is a hole transporting layer, and second electrode 160 is an anode. In other embodiments, first electrode 120 is an anode, transporting layer 130 is a hole transporting layer, transporting layer 150 is an electron transporting layer, and second electrode 160 is a cathode.

First electrode 120 and second electrode 160 may comprise any material capable of conducting electrons. In one embodiment, the cathode is a low work function metal or metal alloy, including, for example, barium, calcium, magnesium, indium, aluminum, ytterbium, silver, a calcium: silver alloy, an aluminum: lithium alloy, or a magnesium: silver alloy. In some embodiments, first electrode 120 and second electrode 160 comprise gold, silver, fluorine tin oxide (FTO) or indium tin oxide (ITO), or conductive polymer layers. In some embodiments, either of first electrode 120 and second electrode 160, or both of first electrode 120 and second electrode 160, are reflective, transparent, semi-transparent or translucent.

In some embodiments, optional charge transporting layers 130 and 150 are independently an electron transporting layer and a hole transporting layer. In one embodiment, the electron transporting layer comprises a material capable of transporting electrons. In some embodiments, the hole transporting layer, when present, is in direct contact with the anode. In some embodiments, the electron transporting material, when present, is in direct contact with the cathode.

Exemplary electron transporting materials include, but are not limited to, semi-conductive metal oxides including oxides of titanium, tin, zinc, iron, tungsten, zirconium, hafnium, strontium, indium, cerium, yttrium, lanthanum, vanadium, caesium, niobium or tantalum, metal chelated oxinoid compounds, such as bis(2-methyl-8- quinolinolato)(para-phenyl-phenolato)aluminum(II) (BA1Q), tris(8- hydroxyquinolato)aluminum (Alq3), and tetrakis(8-hydroxyquinolato)-aluminum (ZrQ); azole compounds such as 2-(4-biphenylyl)-5-(4-t-butylphenyl)-l,3,4-oxadiazole (PB D), 3-(4-biphenylyl)-4-phenyl-5-(4-t-butylphenyl)-l,2,4-triazole (TAZ), and l,3,5-tri(phenyl- 2-benzimidazole)benzene (TPBI); quinoxaline derivatives such as 2,3-bis(4- fluorophenyl)quinoxaline; and phenanthroline derivatives such as 9,10- diphenylphenanthroline (DPA) and 2,9-dimethyl-4,7-diphenyl-l,10-phenanthroline (DDPA). In one embodiment, the electron transporting layer comprises TiCh, SnCh, Fe2O3, WO3, ZnO, Nb20s, SrTiCh, Ta20s, CS2O, zinc stannate, complex oxides such as barium titanate, binary and ternary iron oxides, or indium gallium zinc oxide (IGZO).

There is no particular limit to the composition of active layer 140. In one embodiment, the active layer comprises silicon. In one embodiment, the active layer comprises copper indium gallium selenide. In some embodiments, the active layer may include a stack of sublayers arranged for the purpose of absorbing different regions of the solar spectrum. In some embodiments, the active layer may include a stack of sublayers with different doping levels for promoting the separation of electrons and holes.

In certain embodiments, electrode 160 may be deposited on a substrate 170, which may be transparent, semi-transparent, translucent, or opaque. Substrate 170 may be rigid, for example quartz or glass, or may be a flexible polymeric substrate. Examples of flexible transparent semi-transparent or translucent substrates include, but are not limited to, polyimides, polytetrafluoroethylenes, polyethylene terephthalates, polyolefins such as polypropylene and polyethylene, polyamides, polyacrylonitrile and polyacrionitrile, polymethacrylonitrile, polystyrenes, polyvinyl chloride, and fluorinated polymers such as poly tetrafluoroethyl ene .

EXPERIMENTAL EXAMPLES

The invention is now described with reference to the following Examples. These Examples are provided for the purpose of illustration only, and the invention is not limited to these Examples, but rather encompasses all variations that are evident as a result of the teachings provided herein.

Example 1: High Photoluminescence Quantum Yield Near-Infrared Emission from a Lead-Free Ytterbium-Doped Double Perovskite

Yb-doped Cs2AgBiBr ₆ is a promising downconversion materials to redshift UV and blue photons to near-infrared. Cs2AgBiBr ₆ has a stable cubic structure at room temperature (#225, Fm3m, a=l 1.2499 A) (Gloeckler, M. Sites, J. R. J. Phys. Chem. Solids. 2005, 66, 1891-1894), and an indirect bandgap of ~2.2 eV. When excited with photons with energies greater than 2.2 eV, the bandgap energy, Yb-doped Cs2AgBiBr ₆ thin fdms synthesized via physical vapor deposition emit strong near-infrared luminescence centered at ~ 1.24 eV via the Yb ^{3+ 2} F5/2 —> ² F7/2 electronic transition. Robust, reproducible, and stable photoluminescence quantum yields (PLQY) as high as 82.5% are achieved with Cs2AgBiBr ₆ films doped with 8%Yb. Furthermore, PLQY remains high (>94% of initial value) after two months without encapsulation. This high PLQY indicates facile and efficient energy transfer from the perovskite host, Cs2AgBiBr ₆ , to Yb, making Cs2 AgBiBr ₆ the most promising lead-free downconversion material for solar spectrum shifting to increase solar cell efficiencies.

Herein, Yb-doped Cs2AgBiBr ₆ films with a maximum NIR PLQY of 82.5% and PLQY consistently in the 71-82.5% range for excitation energies above the bandgap (>2.2 eV) are disclosed, the highest values to date from a Yb-doped lead-free perovskite.

The double perovskite, Cs2AgBiBr ₆ , crystallizes in a stable cubic structure (#225, Fm3m, a=l 1.2499 A,) at room temperature. The structure can be viewed as a derivative of the widely studied CsPbBrs wherein the Pb ²⁺ in alternating [PB r6] ⁴ ' octahedra are replaced by Ag ⁺ or Bi ³⁺ , forming a 3D network of corn er- shari ng [AgBr ₆ ] ⁵ ' and [BiBr ₆ ] ³ ' octahedra which alternate periodically (Figure 2). When CsPbX ₃ is doped with Yb, Yb ³⁺ is thought to substitute Pb ²⁺ ions in the octahedra. Yb ³⁺ can replace Bi ³⁺ and Ag ⁺ ions when doped in Cs2AgBiBr ₆ , as shown in Figure 2. Yb ³⁺ substituting isovalent Bi ³⁺ would not need to create a vacancy or an antisite defect, whereas substituting Ag ⁺ has to be accompanied by charge compensating defects such as Ag vacancies (VAg) or Ag antisite (AgBi) defects. Physical vapor deposition (PVD), specifically evaporation, was employed to synthesize Yb-doped CsiAgBiBr ₆ thin fdms from CsBr, BiBr ₃ , AgBr, and YbBr ₃ , whose evaporation rates were measured using separate quartz crystal microbalances. Films were annealed post-deposition in a nitrogen-fdled glovebox at 250-350 °C. CsBr, BiBr ₃ , and AgBr evaporation rates and deposition durations were set to produce nominally stoichiometric Cs2AgBiBr ₆ . Yb doping was varied between 0% and 14% by controlling the YbBr ₃ evaporation rate. The Yb concentration is reported as a percent of Bi lattice positions in stoichiometric Cs2AgBiBr ₆ (i.e., to deposit a 3% Yb-doped Cs2AgBiBr ₆ the ratio of YbBr ₃ molar flux to BiBr ₃ molar flux was set to 0.03). X-ray diffraction patterns of Yb- doped fdms annealed at 300 °C for one hour are shown in Figure 3. All fdms crystallize with the Cs2AgBiBr ₆ cubic structure (#225, Fm3m). Annealing helps the precursors react completely and form the target Cs2AgBiBr ₆ structure. Otherwise, unreacted precursors (e.g., AgBr) and other impurity phases, such as Cs3Bi2Br9, are still present in the as- deposited fdms (Figure 4 and Figure 5). A small shift to higher 20 values was observed in the XRD patterns for Yb-doped fdms compared to the undoped Cs2AgBiBr ₆ fdm. Lattice parameters calculated from XRD data show a small unit cell contraction when Yb up to 10% is introduced to the perovskite structure (Table 1), which is reasonable because Yb ³⁺ ions (101 pm) have a smaller radius than both Bi ³⁺ (117 pm) and Ag ⁺ (129 pm) ions. Films with 10 and 14% Yb show XRD peaks from AgBr at 26.8° and 31.0°. One explanation is that Yb ³⁺ substitutes the Ag ⁺ ions, and the replaced silver forms AgBr as an impurity phase in the fdm. Another possibility is that adding Yb to Cs2AgBiBr ₆ destabilizes the structure and decomposes to AgBr and other impurity phases. Indeed, in addition to AgBr, Cs- AgBr ternary phases were detected in Cs2AgBiBr ₆ fdms when Yb concentration is 10% or greater (Figure 6). The AgBr and ternary Cs-Ag-Br phases are reported as commonly observed impurity phases during colloidal synthesis of Cs2AgBiBr ₆ . The Cs-Ag-Br impurity phases were more prevalent in as-deposited fdms and fdms annealed at 250 °C than in fdms annealed at 300 °C (Figure 6). The Cs2AgBiBr ₆ fdms decompose at 350 °C, (Figure 7, Figure 8, and Figure 9), making about 300 °C a suitable optimized annealing temperature.

Table 1. Lattice parameters (a) of undoped and Yb-doped Cs2AgBiBr ₆ fdms from XRD data.

Comparison of Raman scattering from undoped and Yb-doped films confirmed ytterbium incorporation. Raman spectrum of an undoped Cs2AgBiBr ₆ film consists of three peaks at 173, 130, and 69 cm' ¹ (Figure 10), which agrees well with the reported Raman spectra. Three vibrational modes, Ai _g , Eg, and T2 _g , either of [BiBr ₆ ] ³ ' or of [AgBr ₆ ] ⁵ ' octahedra, have been assigned to the peaks at 173, 130, and 69 cm' ¹ , respectively (Steele, J. A. et al., ACS Nano. 2018, 12 (8), 8081-8090). Doping Cs2AgBiBr ₆ with Yb shifts these peaks to higher wavenumbers: 69 to 80, 130 to 140, and 173 to 184 cm' ¹ at 14% Yb. Since Raman peaks shift to higher wavenumbers when the structure is compressed, the observed shifts in Yb-doped films confirm Yb ³⁺ ions do indeed substitute the octahedral cations in the perovskite structure creating compression strain on the unit cell. Raman spectra from the annealed films do not show any impurity phase peaks (Figure 10). In contrast, Raman spectra from the as-deposited films show peaks that can be assigned to Cs ₃ Bi2Br9 (Figure 5), suggesting that the impurity phases are present in the as-deposited films but are below the detection limit of Raman scattering in films annealed at 300 °C.

Scanning electron microscopy (SEM) images of annealed Yb-doped Cs2AgBiBr ₆ films show uniform films with grain sizes of a few hundred nanometers (Figure 11 and Figure 12). As-deposited films contain poorly defined small (<100 nm) Cs2AgBiBr6 grains and larger domains with different morphology and composition (Figure 6) consistent with unreacted precursors and impurity phases. The Cs2AgBiBr ₆ grains grow and become more well-defined while the unreacted precursors and impurity phases are converted to Cs- 2AgBiBr ₆ and disappear during the annealing step. Compositional analysis by EDS also indicates the presence of Yb in the films. For example, the composition of 8% Yb- Cs2AgBiBr ₆ film is 0.6% Yb, 9.2% Bi, 10.3% Ag, 18.5% Cs and 61.4% Br, close to the values expected from the precursor fluxes (0.8% Yb, 9.7% Bi, 9.7% Ag, 19.4% Cs and 60.5% Br). In summary, XRD, Raman, and SEM-EDS data show that up to 14% of Yb were incorporated into the Cs2AgBiBr ₆ fdm while the perovskite host’s cubic structure is still maintained.

Figure 13 shows the absorption and photoluminescence spectra of an undoped Cs2AgBiBr ₆ thin fdm. The absorption starts rising at 560 nm, suggesting a bandgap of ~ 2.2 eV, in the reported range of 1.8-2.3 eV for the indirect bandgap of this material (Kentsch, R. et al, J. Phys. Chem. C. 2018, 122, 25940-25947). The absorption peak at 435 nm matches reported data for nanocrystals (Creutz, S. E. et al, Nano Lett. 2018, 18, 1118-1123; Bekenstein, Y. et al, Nano Lett. 2018, 18, 3502-3508; Dey, A. et al, ACS Nano. 2020, 14, 5855-5861), and thin fdms (Wright, A. D. et al, J. Phys. Chem. Lett. 2021, 12, 3352-3360). In contrast, absorption measured on single crystals rises monotonically without any peaks (Steele, J. A. et al, ACS Nano. 2018, 12 (8), 8081-8090; Zelewski, S. J. et al, J. Mater. Chem. C. 2019, 7, 8350-8356; Slavney, A. H. et al, J. Am. Chem. Soc. 2016, 138, 2138-2141), likely a result of absorption saturation for thick samples. The absorption peak at 435 nm has been attributed to an exciton (Palummo, M. et al, ACS Energy Lett. 2020, 5, 457-463; Wright, A. D. et al, J. Phys. Chem. Lett. 2021, 12, 3352-3360; Yang, B. et al, Angew. Chem. Int. Ed. 2018, 57, 5359 -5363; Wu, C. et al, Adv. Sci. 2018, 5, 1700759), or the s-p transition on Bi (Igbari, F. et al, Nano Lett. 2019, 19, 2066-2073). The main objection to assigning this feature to an exciton has been the lack of a blue-shift in its wavelength as nanocrystal size is reduced (quantum confinement effect). However, it is not expected for Cs2AgBiBr ₃ > to exhibit absorption blue-shift due to quantum confinement because the exciton Bohr radius in Cs2AgBiBr ₆ is estimated between 0.3 to 0.5 A, smaller than one Cs2AgBiBr ₆ unit cell. Interestingly, Cs3Bi2Br9, with a comer-shared [BiBr ₆ ] ³ ' octahedra, also has an exciton absorption peak at 435 nm, associated with the localized exciton on [BiBr ₆ ] ³ ' octahedra (Tran, M. N. et al, J. Mater. Chem. A. 2021, 9, 13026-13035). The similarity between the two structures and their absorption features suggests that the Cs2AgBiBr ₆ absorption peak at 435 nm is also associated with localized excitons on the [BiBr ₆ ] ³ ' octahedra.

Photoluminescence (PL) from the undoped Cs2AgBiBr ₆ thin film is weak and comprises a broad emission centered around 630 nm (FWHM = 150 nm) (Figure 13). This broad emission has been observed in multiple studies (Steele, J. A. et al, ACS Nano. 2018, 12 (8), 8081-8090), but the origin is still under debate: it has been associated with bandedge transition, self-trapped excitons, and defect-related recombination. Cs2AgBiCl _y Br ₆ -y was deposited and the bandgap of Cs2AgBiBr ₆ shifted to higher energies by substituting bromine with chlorine to examine the origin of this orange PL. XRD patterns from Cs2AgBiCl _y Br6-y thin films indicate that the halides are mixed throughout the films (Figure 14). As shown in Figure 15, the blue absorption peak shifts from 435 nm for y = 0 to 393 nm for y = 4 as the bandgap increases with chlorine substitution. However, the emission does not shift and remains centered at -630 nm for all Cs2AgBiCl _y Br6- _y films (Figure 16). The intensity decreases with annealing (Figure 17). The lack of shift and decreasing intensity with annealing strongly suggests that the 630 nm emission originates from defects and is not due to a band-to-band transition (Figure 18). There is also a weak emission peak at 470 nm from the as-deposited undoped Cs2AgBiBr ₆ thin-film, but it disappears after annealing (Figure 13). The disappearance with annealing supports a defect origin. This Cs2AgBiBr ₆ emission peak at 470 nm was observed and attributed to a defect-related bound exciton. Cs3Bi2Bry thin film also has an emission peak at -470 nm, assigned to emission from excitons trapped on defects. The fact that both Cs2AgBiBr ₆ and Cs3Bi2Br9 thin films have an absorption peak at 435 nm and an emission peak at 470 nm indicates that the absorption and subsequent emission have a common origin and are associated with excitons forming via light absorption and then getting trapped and recombining on defects. The obvious candidate is a localized exciton on [BiBns] ³ ' octahedra, forming upon light absorption and becoming trapped on a Bi vacancy, VBI, before emission. Cation vacancies are a feature of the Cs3Bi2Br ₆ vacancy-ordered perovskite structure. In Cs2AgBiBr ₆ , vacancies are expected to be annealed since the perfect structure has an Ag or Bi cation in all octahedra.

Doping Yb into Cs2AgBiBr ₆ did not affect the optical absorption. Yb-doped Cs2AgBiBr ₆ thin films still have an absorption peak at 435 nm, with the onset at -560 nm (Figure 19), and the bandgap remains unchanged. The energy transfer from the Cs2 AgBi Br ₆ , host to the Yb ³⁺ is efficient, and all films doped with ytterbium emit strongly in the NIR (1.24 eV), in addition to the weak orange emission from the perovskite host (Figure 20 and Figure 21) The NIR emission peak is centered at 997 nm (FWHM = 40 nm), the expected Yb ^{3+ 2} FS/2^ ² F _7/2 electronic transition wavelength. Near-infrared PLQY depends strongly on the Yb concentration and annealing temperature. Films annealed at 300 °C have the highest PLQY, and their PLQYs are stable with time in ambient air (Figure 22 and Table 2). PLQY of fdms annealed at 250 °C decreases significantly after exposure to air (Table 2). Figure 22 shows that increasing the annealing time does not affect PLQY significantly; the difference in PLQY between films annealed for 1 hour and 2 hours is negligible for both annealing temperatures, 250 and 300 °C. As shown in Figure 22 and Figure 23 high near-infrared PLQY between 73.9 and 78.7% was achieved for films doped with 3 to 8% Yb (excitation wavelength was 420 nm with 10 nm bandwidth). Near-infrared emission decreases sharply when Yb concentration is higher than 8%, with PLQY reducing to 40.6% for films with 14% Yb (Figure 22). This decrease in PLQY coincides with impurity peaks appearing in XRD of the films with 10 and 14% Yb. Impurity phases and defect states introduce nonradiative relaxation pathways that compete with the energy transfer to Yb, decreasing the NIR emission. Films annealed at 300 °C are stable in the air, and PLQY does not change significantly with time: PLQY of Cs2AgBiBr ₆ film doped with 8% Yb retains 94% of the PLQY after two weeks and 91% after two months without any encapsulation (Figure 24).

Table 2. NIR PLQY of Cs2AgBiBr ₆ films doped with 3% Yb on the day of the deposition and after one day in the air. The excitation wavelength was 420 nm with 10 nm bandwidth. Dependence of one-day stability of PLQY is shown for different annealing times and temperatures.

Figure 19 shows the excitation spectrum of the Cs2AgBiBr6 film doped with 8% Yb and annealed at 300 °C for 1 hour (emission wavelength, λem= 997 nm). This film had the highest PLQY, 78.7%, when excited at 420 nm. Figure 19 also shows the absorption spectrum and PLQY measured at different excitation wavelengths. The excitation spectrum starts rising at 560 nm, the absorption onset, thus confirming that Yb ³⁺ ions are not excited by the incident photons but receive energy from the perovskite host, as illustrated in Figure 18. When a photon with an energy higher than the Cs2AgBiBr ₆ bandgap (~2.2 eV, or wavelengths lower than 560 nm) is absorbed by the perovskite, it creates an electron-hole pair with the electron being excited to the conduction band. When the electron recombines with the hole, it transfers energy to a nearby Yb ³⁺ ion. Electrons in Yb ³⁺ ions are excited from the ² F _7/2 to the ² F _5/2 state and then emit in the NIR region when they relax down to the ² F _7/2 state. Thus, a UV-visible photon is downconverted to a NIR photon via the energy transfer from Cs2AgBiBr ₆ to Yb ³⁺ ions (Figure 18). Interestingly, the excitation spectrum dips at 435 nm, where the exciton absorption peaks. However, PLQY remains high, 71.6%, when the film is excited at 440 nm. PLQY reaches the highest value of 82.5% when Cs2AgBiBr ₆ films doped with 8% Yb and annealed at 300 °C are excited at 360 nm. Depositing a layer of a downconversion material with 82.5% PLQY can increase a typical Si solar cell’s efficiency from 20.3 to 20.6%, and a CIGS solar cell’s efficiency from 20.4 to 20.7% (See Figure 25 and Figure 26). Further optimization and increases can lead to greater gains.

In summary, undoped and Yb-doped Cs2AgBiBr ₆ thin films via were synthesized via PVD. The thin films showed efficient energy transfer from the host to the Yb ³⁺ ions, which emits 1.24 eV NIR radiation with efficiencies as high as 82.5%. PLQY of Cs2AgBiBr ₆ films doped with 8% Yb and annealed post-deposition at 300 °C remains high even after two months in the air. The perovskite host, Cs2AgBiBr ₆ , has an exciton absorption centered at 435 nm, and the bandgap is estimated as 2.2 eV. The weak and broad emission at 630 nm is assigned to defect recombination. Using the record efficiencies reported here as a starting point, further improvements in synthesis conditions of this class of lead-free materials are likely to increase their PLQY towards 100%. This creates new prospects for modifying the solar spectrum by spectrum shifting to increase solar cell efficiencies.

Experimental Methods

Thin-film deposition Films were deposited in a glove-boxed physical vapor deposition (PVD) system (Angstrom Engineering) with six evaporation sources. The four precursors, BiBr ₃ (99%, Alfa Aesar), AgBr (99.9%, Beantown Chemical), CsBr (99.9%, Acros Organics), and YbBr ₃ (hydrate, 99.99%, Alfa Aesar), were loaded into separate RADAK sources and baked overnight at 60, 100, 100, and 110 °C, respectively. BiBr ₃ was loaded in quartz ampoules, while the others were loaded in alumina ampoules. The glass substrates (25 x 25 mm ² ) were cleaned by sonicating in a 1 :1 solution by volume of acetone (ACS Grade, VWR) and isopropanol (99.5%, VWR) for 30 minutes, dried in an oven, and cleaned with O ₂ plasma for 30 minutes using Expanded Plasma Cleaner PDC-001-HP (Harrick Plasma) before loading them onto the substrate holder. Each precursor’s evaporation rate was monitored during the deposition by separate quartz crystal microbalances (QCMs). The substrates’ temperatures were not controlled during the deposition. The tooling factors were determined by evaporating CsBr and BiBr ₃ separately and obtaining the film thickness from interference fringes in optical transmission. The tooling factors, the ratio of the deposition rate at the substrate to the deposition rate at the QCM expressed as %, were 39.7 and 42.9 for CsBr and BiBr ₃ , respectively. The tooling factor of CsBr was also used for YbBr ₃ and AgBr.

During the deposition of Cs2AgBiBr ₆ films, BiBr ₃ , AgBr, and CsBr were coevaporated onto glass substrates at 1.00, 0.37, and 1.21 A/s, respectively. With these deposition rates, the ratio of CsBr to BiBr ₃ to AgBr molar flux is 2: 1:1. The CsBr, BiBr ₃ , and AgBr source temperatures were the manipulated variables to keep the evaporation rates constant and were -555, 140, and 650 °C, respectively. The controller adjusts the temperature around these values to keep the deposition rate constant at the setpoints.

For Yb-doped Cs2AgBiBr ₆ films, three precursors, BiBr ₃ , AgBr, and YbBr ₃ , were co-deposited on glass substrates. CsBr was deposited on top of this BiBr ₃ -AgBr-YbBr ₃ layer. BiBr ₃ and AgBr evaporation rates were kept constant at 1.00 A/s and 0.37 A/s, respectively, while the YbBr ₃ evaporation rate was varied from 0.03 to 0.14 A/s to change the amount of Yb doping. The YbBr ₃ source temperature ranged from 627 to 688 °C, depending on the evaporation rate and precursor amount in the ampoule. The CsBr evaporation rate was 1.21 A/s. The system base pressure was 6>< 10' ⁷ Torr, while the chamber pressure rose to -10’ ⁶ Torr during the deposition. Each layer was deposited for 30 minutes, resulting in 490±10 nm thick films for CsiAgBiBr ₆ . The optical absorption spectrum was obtained from thinner films deposited for 15 minutes to avoid saturation. Films were annealed on a hot plate in the glovebox at different temperatures and for different durations.

For Cs2AgBiC16-yBr _y thin films, chloride salts were used along with bromide salts to achieve the target halide ratio: Cs2AgBiClBr ₅ films were deposited using AgCl, BiBr ₃ and CsBr; Cs2AgBiC1 ₃ Br ₃ films were deposited using AgCl, BiBr ₃ and CsCl; Cs2AgBiC14 Br ₂ films were deposited using AgCl, BiCh and CsBr. The evaporation rates of AgCl, BiBr ₃ , BiCh, CsBr and CsCl were 0.33, 1.00, 0.85, 1.21 and 1.07 A/s, respectively. Precursors were co-evaporated onto glass substrates for 30 minutes, and the films were then annealed at 300 °C for one hour.

Thin-film characterization

All films were characterized under ambient conditions at room temperature. Photoluminescence (PL) spectra were measured using a QuantaMaster-8075-21 (Horiba) spectrophotometer. Visible PL from all films was excited at 420 nm (5 nm bandwidth) with double monochromator filtered emission from a Xe-arc lamp and detected using a PMT detector. Near-infrared PL was excited at 420 nm for all films. For the film with the highest PLQY (8% Yb-Cs2AgBiBr ₆ ), the excitation wavelength was varied between 360 to 600 nm to examine the PLQY dependence on the excitation wavelength. NTR PL was detected using a liquid nitrogen-cooled InGaAs detector. PLQY was measured using an integrating sphere (Quanta-Phi, Horiba), and the lamp power was measured using Power Meter 843- R and 818-UV photodetector (Newport). X-ray diffraction (XRD) patterns from the films were recorded using a Bruker D8 Discover General Area Detector Diffraction System (GADDS) equipped with a Cu-Ka source. Raman spectra were acquired using a Thermo Scientific DXR Raman microscope. Thin films were excited with a 785 nm laser, and Raman scattering in the range of 50-1800 cm-1 was collected with a 50* Olympus objective, dispersed using a high resolution (2 cm-1) grating, and detected with a CCD detector. Films were examined using a Merlin field emission scanning electron microscope (Carl Zeiss, 5 kV, 110 pA). Their average composition over an area of approximately 10 pm ² was determined using energy-dispersive X-ray spectroscopy (Oxford Instruments EDS) and vendor-provided sensitivity factors. Optical absorptions of the fdms were recorded using an Agilent Cary 5000 UV-Vis-NIR in the 200-2000 nm range.

Calculation of the minimum near-infrared photoluminescence quantum yield needed to increase the efficiency of a solar cell with a given external quantum efficiency

Consider a solar cell with power conversion efficiency, 17 , external quantum efficiency,EQE(λ) , open-circuit voltage, V _oc and fill factor FF. When this solar cell is coated with a material that downconverts UV-blue photons to NIR photons, the downconversion (DC) material absorbs in the UV-blue region of the electromagnetic spectrum and reemits in the near-infrared (NIR) region. The new downconverted photon flux the solar cell receives is where 0(A) is the unconverted solar photon flux (e.g., AMI.5), A is the wavelength (nm), .4(A) is the film’s absorbance, is the NIR photoluminescence quantum yield, and /"(A) is the line shape of the NIR emission modeled as a Gaussian function whose center peak wavelength (A _o = 997 nm) and width are determined by fitting a typical experimental NIR emission spectrum. For Cs2AgBiBr ₆ , the wavelength corresponding to the perovskite bandgap, λ _g , _P , is 470 nm. The short circuit current created from this downconverted photon flux is given by where e is the electron charge and EQE(λ) is the solar cell’s external quantum efficiency. In these calculations, the EQE(λ) data for Si and CIGS was digitized and used. The calculations rely on the assumption that open-circuit voltage and fill factor, V _oc and FF remain the same and constant values are used. The new power conversion efficiency of the solar cell receiving the downconverted solar spectrum is Using these equations, the minimum r) _PL Q _Y required for 77* to exceed the original efficiency of the solar cell without the downconverting coating can be calculated. For the solar cells and V _0C , FF and EQE(λ) values from Friedlmeier, T. M. et al. (IEEE 42nd Photovoltaic Specialist Conference. 2015, 1-3) and from Mazzarella, L. et al. (Appl. Phys. Lett. 2015, 106, 023902), values of 69% and 67% are obtained for typical Si and CIGS solar cells, respectively.

Bohr radius calculation

Exciton Bohr radius is given by where e _r is the dielectric constant is the mass of an electron, a _B =O.O53 nm is the Hydrogen Bohr Radius and is the reduced mass, given by Both these estimates are much smaller than the unit cell dimensions.

Example 2: Reactive Physical Vapor Deposition of Yb-Doped Lead-Free Double Perovskite Cs2AgBiBr ₆ with 95% Photoluminescence Quantum Yield

Yb-doped Cs2AgBiBr ₆ is a promising lead-free halide double perovskite that can be used as a downconverting coating on silicon solar cells to redshift UV and blue photons to near-infrared where the quantum efficiencies are larger. Photoluminescence quantum yield (PLQY) of Yb-doped Cs2AgBiBr ₆ thin films synthesized via physical vapor deposition depends strongly on how the substrate temperature changes during deposition, which determines the amount of Bi incorporated into the film. Yb-doped Cs2AgBiBr ₆ films with PLQY as high as 95% were deposited with excess BiBr ₃ and by ramping substrate temperature during the deposition. Ramping the substrate temperature reduces BiBr ₃ loss from the film by promoting reactions that form Cs2AgBiBr ₆ . The films retain 93% of their initial PLQY values after one month. Cs2AgBiBr ₆ has a stable cubic structure at room temperature (#225, Fm 3 m, a=11.2499 A) (Gloeckler, M. Sites, J. R. J. Phys. Chem. Solids. 2005, 66, 1891-1894), and an indirect bandgap of ~2. 2 eV. Yb ³⁺ ions are hypothesized to replace Bi ³⁺ and Ag- ions in the 3D network of corner-sharing [AgBu,] ⁵ ' and [BiBr ₆ ] ³ ' octahedra which alternate periodically in the Cs2AgBiBr ₆ structure (Figure 27). An electron-hole pair is created when Cs2AgBiBr ₆ absorbs a photon with energy greater than its bandgap (-2.2 eV, or 560 nm). When the electron relaxes to the valence band edge, it simultaneously transfers the energy to a nearby Yb ³⁺ ion. in the crystal structure (which are determined by the synthesis conditions. The synthesis of Yb-doped Cs2AgBiBr ₆ thin films using physical vapor deposition (PVD) and the effects of Yb ³⁺ concentration and annealing conditions on the PLQY was discussed in Example 1, above. The dependence of PLQY on the substrate temperature and, in particular, the surprising dependence of PLQY on transient heating, is reported herein. Specifically, the highest PLQY values are obtained when the substrate temperature is ramped during the deposition, an effect that can be traced to the need to balance the reactions of the precursors to form the film and BiBr ₃ desorption, which both increase with increasing temperature but the former is desired but the latter is undesired.

Effect of substrate temperature on NIR PLQY

When Yb is doped into Cs2AgBiBr ₆ films, they emit strongly in the near-infrared region, as shown in Figure 28. The emission peak is centered at 997 nm with FWHM = 40.5 nm, the signature emission of the Yb ^{3+ 2} F5/2 —> ² F _7/2 electronic transition. Figure 29 shows the NIR PLQY from nominally stoichiometric Yb-doped (8% Yb) Cs2AgBiBr ₆ films deposited with no substrate temperature control, at constant substrate temperature (30, 48, and 75 °C) and with the substrate temperature ramped to 70 or 83 °C. The substrate temperature was not increased to values higher than 83 °C to prevent the re-evaporation and significant loss of BiBr ₃ . In the crucibles, BiBr ₃ starts evaporating at -65 °C at 0.1 A/s, and BiBr3 could re-evaporate from the substrates if the temperature ramp is too fast and the final value is too high. Figure 29 shows the substrate temperature as a function of time during these depositions and the corresponding PLQY of the resulting films after annealing in the nitrogen-filled glove box at 300 °C for one hour. When the temperature is kept constant at 30, 48, and 75 °C, the PLQYs are 47, 45, and 48%, respectively. When the temperature is not controlled, PLQY is 48%. The thermal sources heat the substrates radiatively when the films are deposited without controlling the temperature. However, the temperature increase is expected to be negligible due to the substrate stage’s large size and thermal inertia. Therefore, the substrate temperature remains close to 30 °C even when not controlled at 30 °C. Indeed, the PLQY, 48 %, is nearly the same as when the substrate temperature is controlled at 30 °C, 47%. Surprisingly, the factor that affects PLQY the most is the ramping of the temperature and not the temperature itself. For example, when the substrate temperature is kept constant at 75 °C throughout the process, PLQY is 48%. However, when the temperature is ramped to 70 or 83 °C, PLQY increases to 61-62% (Figure 29).

The importance of temperature ramping suggests that the PLQY may be determined by competing processes whose rates have different temperature dependencies. First, CsBr must react with the underlying AgBr-BiBr ₃ -YbBr ₃ layer to form Cs2AgBiBr ₆ . The precursors may not completely react at low temperatures, but their reaction rates increase with increasing temperature. Second, BiBr ₃ must remain in the film and not re-evaporate significantly to avoid the formation of Bi-deficient phases. However, some deposited BiBr ₃ can re-evaporate when the substrate temperature reaches and exceeds ~ 65 °C. Thus, there could be a competition between these two processes that determine the PLQY. At low temperatures, the precursors may not have reacted completely. At high temperatures, or if the temperature is increased too rapidly before BiBr ₃ is locked in as Cs2AgBiBr6, BiBr ₃ may evaporate, leading to Bi deficient phases in the films.

The reaction that forms CszAgBiBr ₆ , appears to proceed at temperatures as low as 30 °C but is not complete. The XRD from the as-deposited films at 30 °C show all the expected Cs2AgBiBr ₆ diffraction peaks (Figure 30). However, XRD also shows the presence of AgBr and CsAgBr ₂ , which indicates that not all BiBr ₃ has reacted. In fact, the XRD patterns from the as-deposited films all show the presence of AgBr and CsAgBr ₂ regardless of the deposition temperature, whether maintained constant throughout the deposition or ramped during CsBr deposition. Thus, some BiBr ₃ remains unreacted at all temperatures and may evaporate above 65 °C during deposition or annealing. This BiBr ₃ loss may lead to the formation of Bi-deficient phases such as CssBiBr ₆ during annealing. Indeed, XRD of the deposited films while keeping substrate temperature constant at 75 °C show Bi-deficient phases (Figure 31). Specifically, the XRD peak at 25.6° is attributed to CssBiBr ₆ , a commonly observed impurity phase when films are BiBr ₃ deficient (Figure 32). This indicates that significant amounts of BiBr ₃ must have re-evaporated at 75 °C. SEM images show these impurity phases as long rods (1-10 pm, see Figure 33), and EDS analysis shows that their elemental composition is close to that expected for CssBiBr ₆ (Figure 33). These Bi-deficient phases also absorb UV and blue light, but their PLQY is much lower, thus lowering the NIR PLQY of the films. In addition, their presence can also lead to nonradiative pathways such as recombination on defects at the grain boundaries between the desired and impurity phases. These nonradiative processes can compete with the energy transfer between the perovskite host and Yb ³⁺ , thus lowering the NIR PLQY.

When the substrate temperature is ramped up slowly during CsBr deposition, the reaction rate between the CsAgBr and BiBrs increases, and as they react to form Cs2AgBiBr ₆ film, BiBr ₃ is prevented from leaving. However, the impurity XRD peak at 25.6° is still present (Figure 31) in the film deposited with the substrate temperature ramped to 70 °C during CsBr deposition. The peak is weak, suggesting the impurity is only present in a small amount. The XRD peak at 25.6° is not present in the film deposited with the substrate temperature ramped to 83 °C during CsBr deposition. Thus, only small amounts of CssBiBr ₆ form in these films deposited in the ramping temperature mode. This suggests that ramping the temperature reduces the BiBr ₃ loss from the films.

SEM images of Cs2AgBiBr ₆ films doped with 8% Yb show grain-like features with fissures (Figure 34). These grain-like features are not well defined in films deposited in the no-temperature-control and the constant-temperature (30 and 48 °C) modes, i.e., at the lower temperature range investigated. When the films are deposited at higher temperatures, 75 °C, or while ramping the temperature to 70 and 83 °C during CsBr evaporation, the films consist of larger grain sizes (300-500 nm). These large grains show fissures that separate 50-100 nm smaller grain-like domains. The fissures are a result of film densification during annealing. However, the theoretical film contraction factor, 8, defined as the ratio of the sum of the individual unreacted precursor compound thicknesses to the thickness of the target material after the complete reaction, is 0.99. This calculation assumes a stoichiometric reaction and could be lower if, for instance, some BiBr ₃ leaves the film during the reaction. Thus, the appearance of the fissures supports that some BiBr ₃ leaves the film during the deposition and annealing. Seeing differences in morphology between the films in Figure 34 is surprising because all films are annealed at the same temperature, 300 °C, much higher than the deposition temperature. Therefore, one would expect the morphology to develop during the annealing and all films to be similar. However, this is not the case, indicating how the film forms during deposition at low temperatures (<83 °C) affect how the film morphology develops during annealing (at 300 °C). This is consistent with BiBr ₃ leaving the film during deposition and partial reaction completion during deposition. The remaining AgBr, CsAgBr ₂ , and BiBr ₃ (Figure 30) react during annealing, which may lead to different morphologies depending on the amount of BiBr ₃ in the film after deposition. Curiously, the morphology does not appear to affect PLQY strongly, as films with different morphologies, e.g., films deposited at 75 °C (Figure 34) and films deposited without temperature control (~ 30 °C) or at 30 °C have nearly the same PLQY.

Effect of BiBr ₃ evaporation rate on PLQY

Based on the dependence of PLQY on how the substrate temperature changes during deposition, the PLQY is correlated with the amount of Bi incorporated into the film. A series of depositions was conducted to test this hypothesis while increasing the BiBr ₃ evaporation rate from 1.0 to 1.8 A/s, with 1.0 A/s corresponding to stoichiometric Cs2AgBiBr ₆ . All films were deposited while ramping the substrate temperature to 83 °C during CsBr evaporation, and Yb doping was kept constant at 8%. Figure 35 and Figure 36 show the PLQY and XRD of films synthesized with different Bi Br ₃ evaporation rates. After the films are removed from the glovebox and exposed to air, PLQY initially increases (less than 5%) but remains stable after a few days in the air for up to 3 o days (Figure 35 and Figure 36). The PLQY improves from 61 to 78% as the BiBr ₃ rate increases from 1.0 to 1.5 A/s, suggesting that the excess BiBr ₃ compensates for the BiBr ₃ leaving and reduces the formation of impurity phases. XRD spectra of these films match with the reference pattern of Cs2AgBiBr ₆ and do not show the impurity peak at 25.6° (Figure 37). PLQY starts decreasing when the BiBr ₃ rate exceeds 1.5 A/s. Some Bi ³⁺ deficiency is needed to accommodate the Yb ³⁺ ion substitution into the BiBr6 ³ 'octahedra. Excess Bi ³⁺ ions compete with Yb ³⁺ ions for the octahedral positions, so the amount of Yb ³⁺ ions substitution may eventually decrease as more BiBr ₃ is added to the film. The changes in the Bi ³⁺ composition are small, but the film composition obtained from EDS measurements confirms the increase in bismuth concentration from 8.4% to 9.1%, with an evaporation rate from 1 0 to 1.8 A/s (Table 3). SEM images in Figure 41 show that the grain sizes increases and fissures disappear with increasing BiBrs flux to the surface. This confirms that the fissures in Figure 34 are related to the BiBr ₃ loss from the film.

Table 3. EDS data of a Cs2AgBiBr ₆ film doped with 8% Yb and deposited in the rampingtemperature mode as the substrate temperature was increased from 30 °C to 83 °C during CsBr evaporation. The stoichiometric composition is 20% Cs, 10% Ag, 10% Bi, 60% Br without taking into account the Yb, which is 8% of the octahedral Bi positions or 0.8%.

Excitation dependence of PLQY

The dependence of PLQY on the excitation wavelength for a film synthesized at an optimized BiBr ₃ evaporation rate was examined. The highest PLQY measured under 420 nm excitation is 84% and increases to 95% when the film is excited under 360 nm. The films retain 91% of their initial PLQY values after two months (Figure 38). Figure 39 shows the excitation spectrum and the PLQY dependence on the excitation wavelength of this film after 1 month. When the excitation energy is less than the CszAgBiBr ₆ bandgap (-560 nm, -2.2 eV), PLQY is 0%, indicating the perovskite host sensitizes the Yb ^3- ions. PLQY values are above 75% when the excitation wavelength is between 360 and 520 nm (2.4 eV). That PLQY is still >60% at 2.4 eV, less than twice the Yb ^{3+ 2} Fs/2 ² F _7/2 electronic transition (2. 5 eV) suggests that at least some of the perovskite- Yb ³⁺ energy transfer is not by quantum cutting. The dip at 440 nm in the excitation spectrum matches the absorption exciton peak in Figure 40, which suggests Yb ³⁺ ions receive energy from the perovskite host when the Csi AgBiBr ₆ excited electrons relax from the conduction band edge to the valence band.

Yb-doped Cs2AgBiBr ₆ thin fdms have been synthesized by sequential vapor deposition, and the dependence of its NIR emission on substrate temperature and BiBr ₃ evaporation rate was examined. PLQY of Yb-doped Cs2AgBiBr ₆ thin fdms depends strongly on the substrate temperature during the deposition and the BiBr ₃ evaporation rates. Three temperature modes were studied: no-temperature-control, constant-temperature, and ramping-temperature modes. Films synthesized using the ramping-temperature mode yielded the highest PLQY. The ramping-temperature mode allows precursors to react to lock in the volatile precursor BiBr ₃ as Cs2AgBiBr ₆ and prevent its re-evaporation. Evaporation of BiBr ₃ leads to bismuth-deficient phases such as CssBiBr ₆ , which introduces nonradiative relaxation pathways that compete with the energy transfer between the host Cs2AgBiBr ₆ and Yb ³⁺ ions. The BiBr ₃ evaporation rate was optimized to reduce the impurity formation, resulting in a PLQY of 95%. The film is stable under ambient conditions and retains 91% of its PLQY after one month.

Experimental Section

Thin-film deposition

Films were deposited in a glove-boxed physical vapor deposition system (Angstrom Engineering) CsBr (99.9%, Acros Organics), AgBr (99.9%, Beantown Chemical), and YbBr ₃ (hydrate, 99.99%, Alfa Aesar) were loaded into separate alumina crucibles and baked overnight at 100, 100, and 110 °C, respectively. BiBr ₃ (99%, Alfa Aesar) was loaded into a quartz crucible and baked at 60 °C overnight. The glass substrates (25 x 25 mm ² , Thin Film Devices) were sonicated in a 1: 1 solution by volume of acetone (ACS Grade, VWR) and isopropanol (99.5%, VWR) for 30 minutes, dried in an oven, and cleaned with O2 plasma for 30 minutes using Expanded Plasma Cleaner PDC-001-HP (Harrick Plasma). Each precursor’s evaporation rate was monitored by separate quartz crystal microbalances (QCMs). The tooling factors were determined and reported previously (Gloeckler, M. Sites, J. R. J. Phys. Chem. Solids. 2005, 66, 1891-1894).

To synthesize Cs2AgBiBr ₆ fdms doped with 8% Yb, BiBr ₃ , YbBr ₃ , and AgBr were co-deposited on glass substrates first, followed by CsBr. (Yb content is reported in percent of the Bi lattice positions in the Cs2 AgBiBr ₆ structure.) During all depositions, YbBr ₃ and AgBr evaporation rates were maintained at 0.08 A/s and 0.37 A/s, respectively. The effect of BiBr ₃ flux was explored by varying the evaporation rate from 1.00 to 1.80 A/s, with 1.0 A/s being the stoichiometric value for 8% Yb-doped Cs2AgBiBr ₆ and 1.1-1.8 A/s corresponding to 10-80% excess BiBr ₃ . The CsBr evaporation rate was 1.21 A/s. The CsBr, YbBr ₃ , and AgBr source temperatures were manipulated to control the evaporation rates but were approximately 550, 620, and 600 °C, respectively. The BiBr ₃ source temperature varied from 110 to 150 °C, depending on the evaporation rate and the starting amount of BiBr ₃ in the crucible. The effects of substrate temperature were examined in three modes: (1) without substrate temperature control (hereafter referred to as the no-temperature- control mode), (2) with substrate temperature maintained constant at a set value (hereafter referred to as the constant-temperature mode), and (3) with ramping the substrate temperature (hereafter referred to as the ramping-temperature mode). The heater was disabled for the no-temperature-control mode, and the substrates were heated naturally with the heat from thermal sources and evaporated materials. The substrates were kept at a constant temperature (30, 48, or 75 °C) throughout the depositions in the constanttemperature mode. For ramping temperature mode, the temperature was kept at 30 °C for the first layer (BiBr ₃ , AgBr, and YbBr ₃ ) and then ramped slowly to either 70 or 83 °C during the second layer (CsBr) deposition. The substrates were cooled to 30 °C immediately after the deposition. The system’s base pressure was 4 MO' ⁷ Torr, while the chamber pressure rose to ~10 ^-6 Torr during the deposition. Each layer was deposited for 30 minutes, resulting in 490±10 nm thick Yb-doped Cs2AgBiBr ₆ films. The films were annealed on a hot plate in the glovebox under nitrogen at 300 °C for one hour.

X-ray diffraction, scanning electron microscopy, UV-visible absorption spectroscopy, and PLQY measurements were conducted in the same manner as previously published (Tran, M. N. et al., J. Mater. Chem. A. 2021, 9, 13026-13035). NIR PL was excited at 420 nm for all films unless noted otherwise. In addition, the excitation wavelength dependence of the PLQY from the highest performing film was examined by varying the excitation between 360 to 600 nm.

Film Volume Shrinkage/Expansion Factor Calculations

When CsBr, AgBr, and BiBr ₃ react to form Cs2AgBiBr ₆ , the film will shrink because of the differences in the densities of these compounds. The shrinkage factor, 8 , is defined as the ratio of the sum of the individual unreacted precursor compound (i.e., CsBr, AgBr, and BiBr ₃ ) thicknesses to the thickness of the target material, Csi AgBi Br ₆ , after they have reacted completely via the reaction

2CsBr + AgBr + BiBr ₃ — * Cs ₂ AgBiBr ₆ .

The shrinkage factor is given by where the sum is over all the precursor compound (i.e., CsBr, AgBr, and BiBr ₃ ) deposition rates, Rj, and R _prO duct is the product film deposition rate. This factor depends only on the stoichiometric coefficients, v, densities, p and molecular weights, M, of the precursors and products. Using the values in Table 4, δ = 0.99.

Table 4. Table of densities and molecular weights for the precursors and the product.

Example 3; Strong near-infrared emissions from Yb-doped thin films

The double perovskite CsiAgBiCk-yBry bandgap can be tuned by varying y (0< y < 6). The bandgap determines the perovskite absorption, hence determining the light wavelength range is converted to NIR emission via downconversion. In this example, Yb is doped into in Cs2AgBiChBr2 thin films, which has a bandgap of 2.5 eV. The optimized synthesis conditions for Yb-doped Cs2AgBiBr6 thin films are used as the starting point and the experiments were designed to explore the dependence of PLQY on deposition and annealing conditions for Yb-doped Cs2AgBiC14Br2thin films.

Crystal structures and optical properties of Yb-doped Cs^AgBiChBr? films

Cs2AgBiC14Br2 films doped were deposited with 8% Yb using stoichiometric and 50% excess BiCh. Co-evaporation and sequential deposition was used for 8% Yb- Cs2AgBiC14Br2 films and the substrate temperature was ramped during the deposition. XRD patterns show that all Cs2AgBiC14Br2 films doped with 8% Yb and deposited with stoichiometric and excess BiCh form the target cubic structure (#225, Fm3m) (Figure

42). XRD peaks of Yb-doped Cs2AgBiC14Br2 films are shifted compared to the corresponding XRD peaks of Cs2AgBiCl6 and Cs2AgBiBr ₆ , and a single peak is observed at each expected diffraction suggesting that Cl" and Br" ions are well mixed throughout the films. The same XRD shift has been observed in Cs2AgBiC14Br2 powders. Alattice constant of 11.06 A can be extracted for Yb-doped Cs2AgBiC14Br2 films, which is within ±1% of the reported value in literatures (Gray, M. B. et al, J. Mater. Chem. C. 2019, 7 (31), 9686- 9689; and Dakshinamurthy, A. C. et al, J. Phys. Chem. C. 2022, 127 (3), 1588-1597). No impurity or phase segregation is detected from XRD data, suggesting Yb is doped successfully into Cs2AgBiC14Br2 structure. SEM images show that most of the films are made of uniform grains, which support the stability of Yb Cs2AgBiC14Br2 structure (Figure

43). SEM images of stoichiometric, co-deposited 8% Yb-doped Cs2AgBiC14Br2 films show scattered crystallites with different appearance and morphology than the background film suggesting the presence of some impurity phases scattered on the film surface (Figure 43). Composition analysis from EDS of these impurity phases suggests that they are Bi- deficient phases: they contain 7.6% Bi compared to the concentration expected from a stoichiometric film, 10%. Bi deficiency is expected since BiCh starts evaporating at 70°C, which is slightly higher than BiBr ₃ but still lower than the substrate temperature. EDS measurements from the homogeneous areas of all the films show the expected Cs2AgBiC14Br2 stoichiometry (Table 5). There is also no significant increase in Bi concentration in films deposited with excess BiCh. The excess BiCh leaves the film during the substrate heating and annealing.

Table 5. Composition of an 8% Yb-doped Cs2AgBiC14Br2 film deposited with stoichiometric or 50% excess BiCh.

Co-evaporated and sequentially deposited Yb-doped CsiAgBiChBr ₂ films with excess BiCh exhibit higher PLQY than those deposited using stoichiometric BiCh (Figure 42). This is consistent with BiCh leaving the films during deposition, which leads to the formations of Bi-deficient phase impurity that decreases NIR emission. Sequentially deposited films show higher PLQY than co-deposited films, suggesting that the reaction among precursors are different for two synthesis schemes. One explanation is that Yb ³⁺ ions substitute into the crystal structures more efficiently in the sequential deposition, which leads to higher PLQY. The highest PLQY, 45%, was obtained from 8% Yb-doped Cs2AgBiC14Br2 thin film, which is sequentially deposited with excess BiCh.

Next, using the sequential deposition scheme with excess BiCh, Yb concentration was varied from 3 to 20% to study the effects of Yb % doping on PLQY. XRD patterns show that all films crystallize with the target cubic structure Fm3m (Figure 44). No phase segregation and impurity phases were detected from XRD patterns. There is no clear XRD peak shift when Yb concentration increases. SEM images show phase-pure, homogeneous films for all Yb concentration (Figure 45). Films with 3% and 8% Yb show large and uniform grain size, approximately about 1 pm. Films with higher Yb concentrations contain more voids and smaller and non-uniform grain sizes, about 200 nm.

PLQY is highest for film doped with 8% Yb, which is the same optimal Yb concentration for Yb-doped Cs2AgBiBr ₆ films. PLQY decreases sharply from 45% to 14% when Yb concentration decreases from 8% to 3%, suggesting that not enough Yb ³⁺ ions are doped into the crystal structure for NIR emission. Increasing Yb concentration to above 8% also reduces PLQY, possibly due to quenching effect, in which the excitation energy transfers from one Yb ³⁺ to another and eventually to a defect where a nonradiative process quenches it.

In conclusion, the mix halide perovskite host Cs2AgBiChBr2 transfers energy efficiently to Yb, which leads to strong NIR emissions with quantum yield of 45%.

Experimental details Thin-film deposition

Films were deposited in a glove-boxed physical vapor deposition system (Angstrom Engineering). The glass substrates (15 x 15 mm ² , Thin Film Devices) were sonicated in a 1:1 solution by volume of acetone (ACS Grade, VWR) and isopropanol (99.5%, VWR) for 30 minutes, dried in an oven, and cleaned with 02 plasma for 30 minutes using Expanded Plasma Cleaner PDC-001 -HP (Harrick Plasma). Each precursor’s evaporation rate was monitored by separate quartz crystal microbalances (QCMs). During the deposition of Yb-doped Cs2AgBiC14Br2 films, precursors were either co-evaporated or sequentially deposited. For co-evaporation, four precursors, BiCh (99.9%, Alfa Aesar), AgCl (99.995%, Beantown Chemical), YbCh (hydrate, 99.9%, Alfa Aesar), and CsBr (99.9%, Acros Organics), were simultaneously deposited at 0.85, 0.33, 0.07, and 1.21 A/s, respectively. These evaporation rates correspond to stoichiometric Cs2AgBiC14Br2 doped with 8% Yb. The BiCh flux was increased to 1.3 A/s, corresponding to 50% excess BiCh. The substrate temperature was ramped from 30 °C to 83 °C during the deposition, which lasted 30 minutes.

For sequential deposition, three of the four precursors, BiCh, AgCl, and YbCh, were co-deposited on glass substrates, followed by CsBr deposition on top of this BiCh- AgCl-YbCh layer. The substrate temperature was kept at 30 °C during the first layer deposition but ramped to 83 °C during the second layer deposition. BiCh and AgCl evaporation rates were kept constant at 1.3 A/s (excess 50% BiCh) and 0.33 A/s, respectively, while the YbCh evaporation rate varied from 0.04 to 0.17 A/s. The CsBr evaporation rate was 1.21 A/s. stoichiometric Cs2AgBiC14Br2 films doped with 8% Yb were deposited using sequential deposition, where the BiCh evaporation rate was 0.85 A/s. Each layer was deposited for 30 minutes. All films were annealed under nitrogen at 300 °C for 1 hour. X-ray diffraction, scanning electron microscopy, UV-visible absorption spectroscopy, and PLQY measurements were conducted in the same manner as demonstrated in example 1 and 2.

Example 4; Physical vapor deposition of Yb-doped Cs2AgBiBr ₆ thin films from ball- milled powders

Yb-doped Cs2AgBiBr ₆ thin fdms can be formed via co-evaporating Yb-doped Cs2AgBiBr ₆ powders and YbBr ₃ . The powders were mechanochemically prepared via ball milling. Ball mill is a low-cost technique that can create high quality, homogeneous powders, which can be used in combination with physical vapor deposition to form thin films. In this example phase pure Yb-doped Cs2AgBiBr ₆ powders are synthesized with quantum yield up to 51%, and the powders were co-evaporated with YbBr ₃ to yield thin films that also exhibits strong near-infrared emission.

Yb-doped Cs2AgBiBr ₆ powders

The mixture powders with the stoichiometric ratio of 8% Yb-doped Cs2AgBiBr6 was ball milled for four hours and form dark red powders. After annealing at 300 °C for 1 hour in the furnace, the powders turn light orange. X-ray diffraction and Raman spectra of the powders confirms the formation of Cs2AgBiB re cubic structure (#225, Fm3m) (Fig. 46 a, b). X-ray diffraction peaks of the annealed powders are sharp, well-defined (average FWHM = 0.4°), and match well with the reference pattern simulated from the cubic Cs2AgBiBr ₆ structure. No extra XRD peaks were observed from common impurity phases such as Cs3Bi2Br9, Cs2AgBr ₃ and AgBr. The sharp and well-matched XRD pattern suggests that the anneal powders consist of crystalline, phase pure Cs2AgBiBr ₆ . Raman spectrum of the annealed powders exhibit the same pattern as undoped Csz AgBiBr ₆ thin film, with three signature peaks at 75, 136 and 179 cm' ¹ . These peaks are assigned to the vibration modes T _2g , Eg and Ai _g of the [BiBr ₆ ] ³ ' and [AgBr ₆ ] ⁵ ' octahedral, respectively. These peaks are shifted to higher wavenumbers compared to the undoped Cs2AgBiBr ₆ thin film (Fig. 46b), indicating the incorporation of Yb into the perovskite structure. Yb ³⁺ replaces Bi ³⁺ and Ag ⁺ ions when doped in Cs2AgBiBr ₆ , as shown in Figure 2. Since Yb ³⁺ ions have smaller radius than Bi ³⁺ and Ag ⁺ ions, the crystal structure is compressed with Yb doping, which shifts Raman peaks to higher wavenumbers. The as-mixed Yb-doped Cs2 AgBiBr ₆ , powders also show XRD and Raman patterns that belong to Cs2 AgBiBr ₆ , structure, but the peaks are much broader than those of annealed powders (Fig. 46 a, b). This indicating that the as-mixed powders are less crystalline than the annealed ones, suggesting that post annealing is important to obtain high, quality crystalline powders.

The annealed Yb-doped Cs2AgBiBr ₆ powders exhibit strong near-infrared emission when excited with ultraviolet light (X _cx = 360 nm). The PL peak is centered at 998 nm (FWHM = 41 nm), which is emitted via Yb ³⁺ electronic transition ² F5/2— > ² F _7/2 (Fig. 47). The quantum yield of annealed powders is 51%, indicating efficient energy transfer between the perovskite host Cs2AgBiBr ₆ to Yb ³⁺ ions and strong near-infrared emission from the activated Yb ³⁺ ions.

Yb-doped Cs2AgBiBr6thin films

The annealed powders were evaporated simultaneously with YbBrs to yield Yb- doped Cs2AgBiBr ₆ . The deposited film was annealed at 300 °C for one hour, which is the optimized annealing condition, as discussed in Example 1. Both as-deposited and annealed films show the XRD pattern corresponding to the Cs2AgBiBr ₆ target perovskite structure (Figure 48a). The as-deposited film contains some broad and asymmetric XRD peaks at 26.77 and 31.25°, which can be attributed to impurity phases AgBr. This suggest that while the majority of Cs2AgBiBr6 powders evaporated and form Cs2AgBiBr ₆ thin film, a small portion decomposes to AgBr and and other phase impurities at high evaporating temperatures. This is also observed by Pantaler, M. et al. MRS Advances. 2018 3, 1819— 1823 and Fan, P. et al. Nanomaterials 2019, 9 (12), 1760. The impurity peak at 31.25° disappears for the annealed film while the peak at 26.77° becomes more resolved, which indicates that the impurity phases can react during annealing to form Cs2AgBiBr ₆ . The Raman spectrum of the as-deposited film shows the three Cs2AgBiBr ₆ peaks at 180, 139 and 75 cm' ¹ , and a shoulder peak at 192 cm' ¹ , which can be matched with Cs3Bi2Br9 (Figure 48b). This shoulder peak is not present in the Raman spectrum of annealed film, which consists only signature Cs2AgBiBr ₆ peaks at 180, 140 and 75 cm' ¹ . Raman data agrees well with XRD patterns that a small amount of Cs2AgBiBr ₆ powders decomposes to AgBr and Cs3Bi2Br<) at high evaporating temperatures, but the impurities react during annealing to form Cs2AgBiBr ₆ . These Raman peaks are also shifted to higher wavenumbers compared to undoped fdm, suggesting the incorporation of Yb into the perovskite crystal structure. The annealed Yb-doped CszAgBiBr ₆ fdm shows strong near-infrared emission centered at 997 nm and the emission curve obtains the same shape as the annealed powder (Figure 49).

Experimental details

Powder preparation

All precursors were baked under vacuum to remove moisture before ball milling. Four precursors include AgBr (0.5089 g, 99.9%, Beantown Chemical), CsBr (1.1721 g, 99.9%, Acros Organics), BiBr ₃ (1.2190 g, 99%, Alfa Aesar) and YbBr ₃ (0.1173 g, 99.99%, Alfa Aesar) were added to a 50ml a zirconia milling jar. Zirconia balls of different sizes (50.5132 g in total) were also added to the jar. The powder mixture was ball milled using MSE Supplies Vertical High Energy Planetary Ball Mill inside a glovebox for 4 hours in alternating directions. Each direction was run for 30 minutes, rest for one minute and switched to the other direction. The rotation speed is about 700 RPM. After ball milling, the powders were transferred to an alumina boat and annealed inside a furnace (Thermolyne Benchtop Furnace) at 300 °C for 1 hour.

Thin-film deposition

Films were deposited in a glove-boxed physical vapor deposition system (Angstrom Engineering). The glass substrates (25 x 25 mm ² , Thin Film Devices) were sonicated in a 1:1 solution by volume of acetone (ACS Grade, VWR) and isopropanol (99.5%, VWR) for 30 minutes, dried in an oven, and cleaned with 02 plasma for 30 minutes using Expanded Plasma Cleaner PDC-001-HP (Harrick Plasma). The substrate temperature was kept at 30 °C. The annealed Yb-doped Cs2AgBiBr ₆ powders and YbBr ₃ powders were loaded in alumina crucibles. Two powders were deposited simultaneously for 20 minutes. The evaporation rate of YbBr ₃ was maintained at 0.05 A/s with the corresponding temperature of 600 °C. The evaporation rate of Yb-doped CszAgBiBr ₆ , powders vary from 2-4 A/s, with the corresponding temperature of 400-600 °C. The final film thickness calculated from precursors fluxes is about 220 nm.

Powers and thin film characterizations X-ray diffraction, scanning electron microscopy, Raman, and PLQY measurements were conducted in the same manner as demonstrated in Example 1 and 2. NIR PL was excited at 360 nm. The disclosures of each and every patent, patent application, and publication cited herein are hereby incorporated herein by reference in their entirety. While this invention has been disclosed with reference to specific embodiments, it is apparent that other embodiments and variations of this invention may be devised by others skilled in the art without departing from the true spirit and scope of the invention. The appended claims are intended to be construed to include all such embodiments and equivalent variations.