Thin film photovoltaic devices using zintl-phosphide

Abstract

A methodology for performing high throughput, computational screening for materials usable as thin film absorbers is disclosed. The method includes selecting a search set of inorganic compounds, and sequentially filtering the search set with a set of filters. The filters exclude materials on the basis of optical and electrical properties such as band gap, carrier effective mass and the presence of low-formation-energy, deep defects that can act as strong nonradiative recombination centers. The application of this search methodology suggests certain classes of materials usable as absorbers in thin film photovoltaic cells.

Description

THIN FILM PHOTOVOLTAIC DEVICES USING ZINTL-PHOSPHIDECROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application is related to, claims priority to, and incorporates herein by reference for all purposes U.S. Provisional Patent Application No. 63 / 688,132, filed August 28, 2024.STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH

[0002] This invention was made with Government support under contract DE-SC0023509 awarded by the U.S. Department of Energy. The Government has certain rights in the invention.BACKGROUND OF THE INVENTION

[0003] This invention relates to materials and structures for thin film photovoltaic cells, and specifically, to materials for an absorptive layer of thin film photovoltaic cells.

[0004] Photovoltaic cells are and will continue to be a vital component of solar energy generation infrastructure in the U.S. and globally. Currently about 4% of U.S. energy is generated from photovoltaic cells, however, this number is expected to increase as concerns about climate change and carbon emission drive the transition toward more renewable, carbon-free energy sources.

[0005] A p-n junction type photovoltaic cell (PV cell) typically comprises of two layers of semiconductor material, an “n type” layer and a “p type” layer. An n-type layer is conventionally created by introducing impurities (i.e., a dopant) to an intrinsic semiconductor material. This creates excess electrons in the material’s valence band, and the material acquires a negative charge. A p-type layer is conventionally created by introducing a dopant that creates excess positive charges (“holes”) in the material’ s valence band, and the material acquires a positive charge. These materials lack an electron relative to the intrinsic semiconductor material that defines the overall matrix. Generally speaking, an n-type material is a material that has free electrons in its valence band, and a p-type material is a material that has free “holes”, i.e., an absence of electrons equivalent to free positive charges in its valence band.1QB\98203989.1

[0006] In a typical p-type PV cell, a thin n-type layer is provided atop a thicker p-type layer creating a p-n junction. Light passes through the top surface of the device and n-doped layer to the p-doped layer. Light is then absorbed in the p-doped layer (called the “absorber” layer), and the absorption process energizes electrons in the p-layer such that they are energetically promoted from their valence band to the conduction band where they are free to flow into the n-doped layer. Positively charged holes move from the n-doped layer into the p-doped layer. The p-n junction permits negative charge flow from the p layer to n layer, but the device acts as a diode to prevent back flow of negative charge from the n to the p layer. Thus, exposure to light creates a voltage between the n and p layers, and placing a load between the n-type and p-type layers causes current to flow through the load. The band gap (energy difference between the valence and conduction bands) of the layers may be tuned by selecting doping concentration, dopants and physical properties of the device to optimize absorption based on the wavelength range of light incident on the device.

[0007] In addition to the semiconductor layers, the top side of the device will typically include a transparent electrode (e.g., of indium-tin-oxide) through which current is extracted from the device, and a protective first layer (typically glass carrying a thin-film antireflective coating). The bottom side of the p-layer, the absorber layer, will typically carry a non-transparent electrode (e.g., a copper electrode), and there may be a substrate layer below the electrode (e.g., glass, sapphire, or the like). Individual solar cells will typically be connected in series to form “modules” having some usable DC voltage (e.g., 12 or 24V). Modules may then be connected in parallel into an “array” which may be provided on one physical solar panel.

[0008] Other PV module architectures are also possible, such as p-i-n modules, where the p and n layers are separated by an undoped intrinsic semiconductor layer. Currently, more than 90% of photovoltaic cells in use are bulk semiconductor devices where the n and p layers are formed from amorphous, polycrystalline or single crystal silicon. These materials have various tradeoffs in terms of cost, optical efficiency and electrical efficiency.

[0009] Thin film photovoltaic cells have recently attracted interest. In contrast to bulk semiconductor devices, a thin film photovoltaic cell has n and p type layers that range from a few nanometers to a few microns thick, as compared to bulk semiconductor devices where the silicon2QB\98203989.1layers can be hundreds of microns thick. Thin film PV cells are also potentially cheap to fabricate, and may be applied on many different substrates, including plastic. One oft-cited application for thin film PV cells is to deposit large arrays of devices on residential or office building window glass, where they would absorb a portion of incident light for power generation. Additionally, thin-film solar cells have manufacturing advantages in that they can require smaller upfront capital investment, less energy input, and produce fewer greenhouse gas emissions to produce than today’s dominant silicon-based technology. Thin-film technologies can also offer compatibility with flexible substrates and band-gap tuning opportunities, leading to building-integrated power generation or tandem devices in combination with silicon PV. Thin-film photovoltaics are also promising because their manufacturing can require smaller upfront capital investment, less energy input, and produce fewer greenhouse gas emissions than today’s dominant silicon-based technology.

[0010] Among the conventional materials used as solar cell absorbers, CdTe and Cu(In,Ga)Se2 (CIGS) have achieved high power conversion efficiency of over 22% and have been commercialized for almost two decades. These conventional materials have drawbacks, however. CdTe and CIGS rely on critical elements, such as tellurium and indium, raising concerns for being able to use devices based on such materials at scale, for example, in multi -terawatt scale deployment. Over the last decade, earth-abundant thin-film materials such as Cu2ZnSn(S,Se)4 (CZTS) and, in particular, lead halide perovskites (e g., MAPbb, where MA = CHsNH+) have emerged. However, CZTS lags behind commercial thin-film technologies in efficiency, while halide perovskites suffer from long-term stability issues.

[0011] It is clear that large-scale deployment of thin-film PV demands high-efficiency, stable, and low- cost solar absorbers other than those based on conventional materials. For over half a century, only a limited number of materials have been explored for thin-film PV applications, as conventional materials discovery is an intrinsically slow process. Most of the materials research efforts in the thin-film PV field have been devoted to modifying known semiconductors (cation substitution in CZTS, mixed-cation / mixed-halide perovskites, etc.), but the recent advent of halide perovskites demonstrates that an unexpected new family of PV materials can still be found and revolutionize the field.3QB\98203989.1

[0012] Accordingly, there is a need for new materials for use as thin-film PV cell absorbers which are environmentally stable, based on abundant materials, and have high conversion efficiency. Additionally, there is a need for streamlined search methods to discover new materials potentially suitable for use as thin film PV absorbers.SUMMARY OF THE INVENTION

[0013] In one embodiment, a methodology is presented for discovering materials suitable for use as thin-fdm PV cell absorbers. The methodology includes selecting an initial search set of materials, which preferably is a search set of inorganic materials likely to have good environmental stability parameters. That initial search set is then subjected to hierarchical filtering and exclusion according to a number of criteria. A first criterion excludes materials likely to be thermodynamically unstable, and therefore unsuitable for devices used in air, exposed to water and used at typical environmental temperatures likely to be experienced in a PV installation. A second criterion excludes materials from the filtered set on the basis of bulk electrical and optical properties such as band gap and carrier effective mass. A third criteria excludes materials from the filtered set on the basis of defect properties, for example, whether the properties support defect- assisted nonradiative carrier recombination above some threshold, which tends to decrease conversion efficiency. A fourth criteria based on a computation of solar cell efficiency is performed. These criteria may, in alternative embodiments may be reordered, and some may be omitted, however, the hierarchical filtering approach set forth above is preferred. Materials selected according to the aforementioned process may then be synthesized and experimentally verified.

[0014] In another aspect, materials suitable for use as thin film PV absorbers for use as p-type or intrinsic layers are disclosed, as well as PV devices fabricated with layers having such materials. In particular, after applying the hierarchical filtering methodology described above, BaCd2P2 was determined to be a long carrier lifetime and stable solar absorber. In another aspect, a family of materials is disclosed that are suitable for use as thin-fdm PV absorbers, those materials combining a suitable band gap, small effective masses, and promising defect properties that will not cause strong nonradiative carrier recombination.4QB\98203989.1

[0015] In yet another aspect, thin film PV devices are disclosed that incorporate the discovered materials herein disclosed.

[0016] In one embodiment, the invention includes an absorber layer in a photovoltaic cell comprising a compound of the form AMzPnz, where A= Ba, Sr, Ca, Mg, M= Mg, Cd, Zn, and Pn= Bi, As, Sb, P, N). This absorber layer may comprise BaCd2P2. The absorber layer may be used as the p-layer in a p-n junction PV device (in which case it may be doped) or as the intrinsic layer. Accordingly, PV devices such as p-n junction or p-i-n junction devices including the aforementioned absorber layer are also within the scope of the invention. These devices are preferably thin film devices.

[0017] In another embodiment, a method is provided for selecting materials usable as solar absorbers. The method includes the steps of identifying an initial material search set and sequentially subjecting the initial material search set to filters, the application of which excludes materials on the basis of thermodynamic stability, electronic properties, defect properties and optical conversion efficiency. Exclusion on the basis of electronic properties may include exclusion on the basis of band gap and carrier effective mass. Exclusion on the basis of defect properties may include exclusion on the basis of the presence of low-formation-energy, deep defects that can act as strong nonradiative recombination centers. The initial search set may be limited to inorganic materials. The method may include further exclusion of materials on the basis of environmental properties, such as by excluding materials that are unstable in air or water, or at temperatures expected under typical operating conditions.

[0018] The use of BaCd2?2 in photovoltaics has certain advantages. For example, BaCd2?2 contains prevalent elements that are widely available. The material is environmentally stable, and as the experimental and computational results disclosed herein demonstrate, devices fabricated with the material are highly efficient.

[0019] The invention and the technical environment are explained in greater detail below with reference to the appended drawings. It should be noted that the invention is not intended to be limited by the embodiments which are set out. In particular, unless otherwise explicitly set out, it is also possible to extract partial aspects of the content and features, including those of the claims,5QB\98203989.1and to combine them with other components and findings or features from the present description or other claims.

[0020] These and other features, aspects, and advantages of various embodiments of the present invention will become better understood with regard to the following description, appended claims, and accompanying figures.BRIEF DESCRIPTION OF THE DRAWINGS

[0021] FIG. 1 schematically illustrates a thin film photovoltaic cell, which may incorporate layers comprising inventive materials described in this disclosure.

[0022] FIG. 2 illustrates steps of a hierarchical search methodology usable for the discovery of thin film PV absorber materials.

[0023] FIG. 3 is a plot of various materials identified according to inventive screening methods suitable for use as solar absorbers.

[0024] FIG. 4 shows the crystal structure for BaCd2P2 and illustrates the material’s computed bulk and optical properties relative to other accepted absorber materials.

[0025] FIG. 5 illustrates defect properties for BaCd2P2.

[0026] FIG. 6 illustrates certain physical and electronic properties of a synthesized sample ofBaCd2P2.

[0027] FIG. 7 illustrates data regarding empirical characterization of BaCd2P2 samples.

[0028] FIG. 8 illustrates a matrix of HSE+SOC computed direct band gaps for P-3ml ANEP materials, where A= Ba, Sr, Ca, Mg; M = Zn, Cd, Mg; and Pn = N, P, As, Sb, Bi.

[0029] FIG. 9 illustrates a matrix of HSE+SOC computed indirect band gaps for P-3mlAM2Pn2 materials, where A= Ba, Sr, Ca, Mg; M = Zn, Cd, Mg; and Pn = N, P, As, Sb, Bi. Together with FIG. 8, the calculations show the band gaps associated with compounds which may be suitable for use as solar absorbers.6QB\98203989.1

[0030] FIG. 10 illustrates a generalized crystal structure for a family of compounds likely usable as solar absorbers.DETAILED DESCRIPTION OF THE INVENTION

[0031] Before the present invention is described in further detail, it is to be understood that the invention is not limited to the particular embodiments described. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting. The scope of the present invention will be limited only by the claims. As used herein, the singular forms "a", "an", and "the" include plural embodiments unless the context clearly dictates otherwise.

[0032] It should be apparent to those skilled in the art that many additional modifications beside those already described are possible without departing from the inventive concepts. In interpreting this disclosure, all terms should be interpreted in the broadest possible manner consistent with the context. Variations of the term "comprising", "including", or "having" should be interpreted as referring to elements, components, or steps in a non-exclusive manner, so the referenced elements, components, or steps may be combined with other elements, components, or steps that are not expressly referenced. Embodiments referenced as "comprising", "including", or "having" certain elements are also contemplated as "consisting essentially of' and "consisting of' those elements, unless the context clearly dictates otherwise. As used herein, "consisting essentially of or "comprising substantially" means that specific further components can be present, namely those not materially affecting the essential characteristics of the compound or composition. It should be appreciated that aspects of the disclosure that are described with respect to a system are applicable to the methods, and vice versa, unless the context explicitly dictates otherwise.

[0033] Described below is a method for conducting high-throughput (HT) computational screening and experimental verification for identifying suitable materials for PV, thin-film absorbers. HT computational screening can significantly accelerate materials discovery. Although earlier HT searches for solar absorbers focused on bulk material properties, such as band gap, effective mass, and optical absorption coefficient, the method to be described also incorporates complex properties related to intrinsic defects. Considering defect-assisted nonradiative carrier7QB\98203989.1recombination is advantageous in the search for new high-performance solar absorbers. Indeed, designing and discovering materials with “defect tolerance” and long carrier lifetime has become an important target for the PV community. Yet, only a few HT studies have included defect properties in the screening process, and their focus so far was on copper-based inorganic materials and perovskite compounds. Moreover, prior computational predictions have not yet led to the experimental realization of a proposed solar absorber.

[0034] The search methodology was used to identify and verify BaCd2P2 as a long carrier lifetime and stable solar absorber. An exemplary methodology is a hierarchical computational material screening process using one or more criteria, followed synthesis and validation of candidate materials. A search method begins by initially selecting a material search set. In one search, the search set was chosen to be 40,000 inorganic materials selected an initial search space. One example of an initial search space is the Materials Project database, most of which are registered in the Inorganic Crystal Structure Database (ICSD). Other structural databases can be used as an initial search space, for example OQMD (https: / / oqmd.org / ), AFLOW (https: / / aflowlib.org / ), Gnome (https: / / www.gnome.org / ), and Alexandria (https: / / alexandria.icams.rub.de / ). Functional criteria are then applied to this initial search set to sequentially fdter for candidate materials. Depending on the exact criteria used, the materials resulting from this search combine a suitable band gap, small effective mass, and promising defect properties that will not cause strong nonradiative carrier recombination.

[0035] Applying the search methodology described above yields candidate materials for thin fdm PV absorbers. One particularly suitable candidate uncovered according to this search methodology is Zintl-phosphide BaCd2?2, which demonstrates computed nonradiative recombination rates that are better than or comparable with those in high-efficiency solar absorbers such as the halide perovskites. The suitability of BaCd2P2 was verified by experimentally synthesize and characterize the material, showing that it is highly stable in air and water. Bright photoluminescence (PL) and time- resolved microwave conductivity (TRMC) measurements confirm the computed direct band gap of 1.45 eV and present a promising long carrier lifetime of up to 30 ns. All of these results indicate that BaCd2P2 is a promising high-performance solar cell absorber with the potential to open a new avenue in PV for an entire family of Zintl AM2X2 solar absorbers (where A and M are +2 ions and X is a pnictogen).8QB\98203989.1

[0036] Referring now to FIG. 1, there is shown a schematic diagram of the layer structure of a p-type, p-n junction thin film photovoltaic cell 100 according to an inventive embodiment. Referring to the layers from top to bottom, a top layer 105 is a transparent cover layer. In conventional bulk solar cells, this layer is typically glass, but in thin film devices, this layer may be a polymeric layer chosen for its optical (i.e., transparent in the light waveband of interest, not birefringent) as well as mechanical qualities. In certain embodiments, this top protective layer is a transparent metal oxide or nitride layer. The top protective layer 105 may carry or may itself be an anti-reflective layer, depending on its thickness. The protective top layer should be mechanically and environmentally durable and stable, and should resist damage from weathering and cleaning.

[0037] Beneath the protective layer 105 is a top electrode layer 110. Top electrode layer 110 may be a transparent conductor such as indium tin oxide (ITO). In certain embodiments, a conductive top layer (i.e., a transparent conductive oxide or TCO material) may provide both protective and conductive functions. In a p-type device such as the one pictured, an n-type semiconductor layer 115 sets below the top electrode. This layer is sometimes referred to as the window layer or the emitter layer in a p-type device. Because the n-type layer will typically have a relatively high index of refraction at the working wavelengths of the device, it is typical to incorporate an anti -refl ection coating (not illustrated) in layer 115, which is situated on top of n- type layer 115. Beneath the n-type layer 115 is the p-type layer 120. This layer is referred to as the absorber, and the materials discussed below are usable for this layer. The absorber layer absorbs photons, resulting in promotion of electrons from its valence band to its conduction band, where they cross the p-n junction, as shown. Positive holes migrate across the junction in the opposite direction. This creates a voltage between the p-type layer and the n-type layer.

[0038] Below the p-type layer is a conductive electrode 125, which may be metal. The structure of the device 100 may be built on a substrate 130, which may be any material suitable for fabrication of the subsequent upper layers by conventional fabrication process (e.g., CVD).

[0039] Referring now to FIG. 2, there is shown a selection methodology for determining suitable materials to be used as thin-fdm solar absorbers according to inventive embodiment. In a9QB\98203989.1first step 205, a set of materials is identified as a search set. The initial search set can be defined by those materials for which a DFT electronic structure and effective mass data are available.

[0040] After an initial search set of materials is identified, a set of filtering criteria is applied to sequentially winnow the number of candidate materials. While the filtering steps of FIG. 2 (e.g., 210, 215, 220 and 225) are presented in a particular order, which is preferred, this order is not limiting. For example, in the example implementation of the screening method described below, materials were first screened on the basis of band gap and carrier mass, and screening on the basis of stability was reserved until conversion efficiency and cost could be evaluated. Thus, the presented filtering steps may be applied in any order, certain steps may be omitted, and other steps may be added (e.g., screening on the basis of material rarity, availability, toxicity, ease or difficulty of fabrication and synthesis, etc.) without deviating from the invention.

[0041] Referring still to FIG. 2, once the search set is selected it is filtered for materials having suitable thermodynamic stability (step 210). This means, inter alia, the materials identified at this stage are stable at expected operating temperatures, pressures, exposures to moisture and atmosphere, and other environmental conditions expected under expected operating conditions. Certain thresholds may be applied to various physical and environmental parameters to exclude materials in the search set based on instability at expected environmental and operating conditions. The result of this process is a set of materials that are acceptably stable under expected operating and environmental conditions. In particular, materials whose energy above hull is above 0.1 eV / atom may be excluded, as these materials are unlikely to be thermodynamically stable. Exemplary operating conditions for outdoor use are between 10 °C and 60 °C.

[0042] In a next step 215, the stable materials are further filtered according to electronic properties. Here, criteria and optionally thresholding is again applied to select materials usable as PV absorbers under the expected operating conditions. Exemplary electronic properties for which the materials are filtered include band gap and carrier effective mass. The desired band gap will be driven by the expected optical operating spectrum.

[0043] In some embodiments, materials can be retained based on having a band gap between 0.3 and 2.0 eV, as determined by DFT (e.g., using GGA-PBE). This range is selected considering that a solar absorber should have a suitable band gap of ~1.5 eV according to the detailed-balance10QB\98203989.1(Shockley-Queisser) model and that (semilocal) DFT severely underestimates the band gap of semiconductors.

[0044] Some of the materials falling within the above band gap range also have an indirect band gap. For targeting thin-film solar absorbers, indirect band gap materials can be excluded where the direct band gap is more than about 0.15 eV (e.g., in the range of 0.13- to 0.17 eV) higher than the indirect band gap. The band gaps can be refined using the HSE hybrid functional (with 25% screened exchange). Materials identified this way having a direct band gap within the range of about 1 eV to about 2.5 eV and an indirect band gap within 0.15 eV from that (i.e., indirect band gap is less than 0.15 eV lower than the direct band gap) can be retained for further filtering steps.

[0045] Materials can be filtered out depending on electron and hole effective masses that are larger than 1.0 mo and 10 mo (where mo is the free electron mass), respectively. The effective masses can be conductivity effective masses which can be derived from a uniform, Fourier interpolated k-point grid of electron band structure and Boltzmann transport calculations within the constant relaxation time approximation as reported by Ricci, F. et al., “An ab initio electronic transport database for inorganic materials.” Sci. Data 4, 2017, 170085.10.1038 / sdata.2017.85. The effective masses computed in this way take into account nonparabolicity of bands and multiple pockets with close energy in different part of the Brillouin zone.

[0046] In a next step 220, the remaining materials are filtered for defect properties. Intrinsic point defects may be computed, for example using the PyCDT package. Variables such as vacancies, cation-cation antisites, interstitials and cation-anion antisites may be considered in the computation. In certain examples, the defect calculations may be performed at the DFT-PBE level. At this level of theory, truly deep defect levels may be predicted to be shallow, due to the DFT band-gap problem. Defect levels calculated in this way may be corrected by aligning the PBE and HSE band edges. In consideration of nonradiative carrier recombination via deep levels, the nonradiative carrier lifetime may be computed assuming a constant cross section such as 1013cm2for all defects. Materials having a nonradiative carrier lifetime shorter than 1 ns can be removed from consideration as this qualitatively means that a material has low-formation-energy, deeplevel defects. Defects for the remaining materials may be computed at a full HSE level, which may include structural relaxations, as the screening involving PBE defect calculations with shifted band11QB\98203989.1edges can overestimate carrier lifetime, because some truly deep defects may remain predicted as shallow. If more computational resources are available, a direct screening at the HSE levels for defects could be used as well (without an intermediate PBE calculation). Approaches combining PBE with a so-called single-shot HSE approach can also be used for this screening.

[0047] In a next step 225, the remaining materials are filtered for conversion efficiency and optionally for other properties such as material rarity and cost,

[0048] The theoretical solar-cell power conversion efficiencies of remaining candidates may be predicted using a nonradiative detailed balance model. This model accounts for defect-assisted nonradiative carrier recombination. The optical absorption spectra and intrinsic defect properties of the material are important for the prediction of conversion efficiencies of a material. In some embodiments, full HSE defect formation energies and charge transition levels may be used. Combinations of vacancies, cation-cation antisites, interstitials and cation-anion antisites (e.g., vacancies and cation-cation antisites) may be considered. A constant capture cross section can be assumed for all defects. If not known, the absorption spectra can be computed, for example in the independent-particle approximation and at the DFT-PBE level, with a rigid shift of the absorption edge to match the HSE band gap.

[0049] For DFT calculations concerning the electronic properties and the defect properties, the VASP code and its standard 5.2 PBE projector-augmented wave (PAW) pseudopotentials may be used as recommended by the then Materials Project with a plane- wave energy cutoff of 520 eV. For the defect calculations, the supercell size may be chosen such that each supercell lattice dimension is greater than 10 A, and a f-only k-point may be used.

[0050] Materials cost may be estimated via the volume cost, based on the cost of the elements and material density, which may be obtained from the Materials Project or derived from Vegard’s law. Supply risks for the elements may be assessed in terms of Herfindahl-Hirschman index (HHI) for geographical concentration of production and reserve.

[0051] It will be recognized that filtering materials on the basis of the criteria presented requires estimating the various material properties, and then comparing the estimates to predetermined thresholds. For example, filtering on the basis of band gap and carrier effective12QB\98203989.1mass require modeling the properties of materials in the set and comparing the properties that result from the models with threshold values to select acceptable band gaps and carrier effective masses.

[0052] In one example, applicants chose a search set based on the Materials Project database, which includes full data on material crystal structure, electronic band structure and carrier effective masses. 39,659 materials within the Materials Project’s database were obtained without any limitations on chemistry, such as the elements included or thermodynamic stability. A tiered computational screening approach using high throughput (HT) computing resources was used, as reflected in FIG. 2, to search for solar cell absorbers among these materials. When computing relevant electronic and defect properties using HT screening it was found to be advantageous to combine calculations at different levels of theory from (semilocal) density functional theory (DFT) to more accurate, but computationally expensive, Heyd-Scuseria-Ernzerhof (HSE) hybrid functional. One exemplary method of conducting screening for electronic properties (i.e., step 215) included first computing DFT band gaps and effective masses, followed by a refinement of the band gap estimate using the HSE hybrid functional, and then defect computations combining DFT and HSE.

[0053] To screen on the basis of defect properties (i.e., step 220), applicants performed full HSE defect calculations for 19 candidate materials that had been determined to have acceptable band gap and effective mass characteristics according to the prior screening step. On the basis of this defect calculation, the solar-cell efficiency of devices using those materials was computed using an extended detailed-balance (i.e., Shockley-Queisser) model, which accounts for defect- assisted nonradiative carrier recombination. At this screening stage, applicants examined only vacancies and cation-cation antisites without considering interstitials and cation-anion antisites. For nonradiative recombination, applicants assume a constant carrier capture cross-section for all defects. This effectively penalizes materials with low-formation-energy, deep defects. However, this model can discriminate between high- performance solar absorbers and materials that are unlikely to realize high efficiency because of their defect properties.

[0054] The result of these steps (computing band gap, carrier effective maps, defect impacts on efficiency, and modeling conversion efficiency) resulted in the identification of 19 materials, which are shown in FIG. 3. Also shown in FIG. 3 are 12 established solar absorbers in use13QB\98203989.1currently for comparison. These materials are plotted with their theoretical power conversion efficiency and estimated material cost.

[0055] The materials shown in FIG. 3 are also categorized and may be screened in terms of their thermodynamic stability under expected operating and environmental conditions. Given that stability issues are hampering the deployment of halide perovskite solar cells, it is advantageous to take into account stability in selecting candidate materials. To provide a measure of stability enabling comparison of materials, the thermodynamic stability (energy above hull) of a material was computed versus competing phases from DFT. While somewhat helpful, this is not an ideal indicator of stability in air and with respect to moisture. In fact, most non-oxide materials show thermodynamic instability in contact with oxygen. Kinetic effects and surface passivation are likely to be key factors for their stability but are currently poorly assessed by theory. One way to assess the stability of the candidate materials is reflected in the reporting in FIG. 3, is to classify materials according to experimental reports on these materials, including the work reporting their synthesis and structural determination. Direct experimental information on air / moisture stability can be found for most of the candidate materials. Interestingly, the reported air / moisture stabilities agree with the solid-state- chemistry experience of the applicants as to broad trends: the higher the alkali and alkaline-earth metal per formula unit, the less stable to air / moisture the material will likely be. For example, according to this experience, KZnP (33%) should be less stable than BaCd2P2 (20%), which is in line with the experiment. BaLi As, BaNaP, KSb, and Na4CdP2, which lack direct experimental information on air / moisture stability, are expected to be highly sensitive to air / moisture, given their high alkali and alkaline-earth metal per formula unit.

[0056] As can be seen in FIG. 1, the most promising candidates when combining power conversion efficiency, material cost, and stability are BaCd2?2 and BaZrSa. While BaZrSs perovskite has recently been studied for thin-film PV applications, BaCd2?2 is classified as a Zintl compound and has received little attention in the PV field. Although BaCd2?2 has recently been shown to possess a suitable band gap and small effective masses, the disclosed HT screening method, which adds the consideration of defect-assisted nonradiative recombination, makes it stand out from the vast number of materials with suitable band gaps. Other candidates, such as the spinel ScCd2S4 or I SnTes, have cost and abundance issues. They are still of scientific interest, and substitution strategies could be considered to reduce or even eliminate the use of the critical14QB\98203989.1elements (Sc and Te). Other interesting candidates, such as KZnP, Na2KSb, SrCuN, and CaaAsN, are excluded due to stability concerns, confirmed by their experimental synthesis reports. Based on these considerations, applicants have identified BaCd2P2 as the most promising solar absorber candidate. The following disclosure provides a more detailed study of the optoelectronic and defect properties of BaCd2P2.BaCdzP2 computed bulk optoelectronic properties

[0057] BaCd2?2 crystallizes in the 733mlCaA12Si2 Zintl-type structure, which is illustrated at FIG. 4a. The crystal structure consists of alternating layers of tetrahedrally coordinated Cd and octahedrally coordinated Ba. This structure type has attracted substantial interest for thermoelectric applications, e.g., the antimonides MgsSb2 and CaMg2Sb2. FIG. 4b shows the BaCd2P2 band structure calculated with HSE. The direct band gap value of 1.45 eV is ideal for singlejunction solar cells according to the detailed-balance model. There is a competing indirect gap that should not be detrimental to PV performance and, instead, has been suggested to be beneficial to carrier lifetime. The band structure exhibits dispersive valence and conduction bands. As a result, the effective masses are low for both electrons (0.11-0.74 mo) and holes (0.44-0.63 o). Further carrier mobility calculations including phonon scattering give a room-temperature mobility of 111-916 cmWs for electrons and 123-242 cm2 / Vs for holes, depending on the crystallographic direction. The computed mobility is for a perfect single crystal of BaCd2P2 and provides an upper bound for the mobilities in a polycrystalline fdm with impurities. Although lower than in III-V semiconductors such as GaAs, the mobilities are comparable with those in many other solar absorbers such as CdTe, CZTS, and MAPI3. Besides, as shown in FIG. 4c, the computed optical absorption coefficient of BaCd2P2 is over 104cm'1in the visible light range and is on par with those of known thin-film solar absorbers.BaCdiP2 defect charge transition levels and dopability

[0058] Applicants identify BaCd2P2 as a material of interest, not only for its band gap, carrier transport, and optical absorption but also for the absence of low-formation-energy, deep defects that can act as strong nonradiative recombination centers. In the initial screening, only limited types of intrinsic point defects were considered. What follows is a comprehensive first-principles15QB\98203989.1study of intrinsic point defects in BaCd2P2. The electrical behavior of the defects is assessed based on their calculated formation energies and charge transition levels. The defect formation energy depends on the elemental chemical potentials that are related to synthesis conditions. The allowed chemical potentials of Ba, Cd, and P satisfy the stability condition of BaCd2?2 and are bound by the formation of different secondary phases in the Ba-Cd-P system.

[0059] FIGs. 5A and 5B plot the defect formation energies as a function of Fermi level under P-poor and P-rich conditions, respectively. These figures show that vacancies are the dominant defect species in BaCd2P2: FB.I and Fea are the dominant acceptors; Fp is the dominant donor. When ionized, the vacancies can have low formation energies for Fermi-level positions close to the respective band edges. Fortunately, they are all shallow defects and can therefore be ruled out as nonradiative recombination centers. Early on in the study of lead halide perovskites, it was hypothesized that the long carrier lifetime and defect tolerance of MAPI3 result from the shallow nature of vacancy defects. Although for MAPI3 this picture has been challenged by recent studies, BaCd2P2 clearly shows shallow vacancies for both cations and anions. The formation of shallow vacancies has been linked to the electronic structure of the material and, more specifically, to the presence of an antibonding upper valence band and a bonding lower conduction band. It has been argued that this type of bonding pushes up the valence-band maximum (VBM) and pushes down the conduction-band minimum (CBM), making the band edges close to the vacancy dangling-bond states and hence leading to shallow vacancy defect levels. This defect-tolerant electronic structure concept applies partially to BaCd2P2; that is, for BaCd2P2, the upper valence band consists of antibonding states between Cd 4t7 and P 3 , while the lower con- duction band arises mostly from the antibonding interactions between Cd 5s and P 3s and 3p). In this regard, BaCd2?2 is similar to MAPI3, where both the up- per valence band and lower conduction band are antibonding (Pb 6.s- 1 5p and Pb 6p- 1 5p, respectively).

[0060] Some of the antisite defects in BaCd2?2 can also be low in formation energy, notably for Bac and Cdna. Given the large difference in ionic radii between Cd2+and Ba2+(1.09 versus 1.49 A°), one would expect that it should be energetically unfavorable to form Bacd than Cdna. However, Applicants’ analysis has shown no clear correlation between ionic size difference and antisite formation energy. Rather, previous studies have found that elements with similar preferred oxidation states tend not to form antisites with deep levels. This is the case here for Ba occupying16QB\98203989.1the site of the smaller Cd but, surprisingly, not for Cd replacing the larger Ba. The Bacd introduces no defect level in the band gap of BaCd2P2, as expected. By contrast, Cdna is a deep double acceptor. This may be understood from the larger antisite space of Cdna, in which not only Cd2but also Cd+and Cd° with bigger ionic sizes can form. Both the Pea and Cdp antisites are amphoteric (i.e., exhibiting both positive and negative charge states across the band gap) and have deep levels in the band gap. The amphoteric behavior of these two antisites may be linked to the oxidation behavior of P, which can act both as a cation and anion and has oxidation states spanning from + 5 to - 3. The interstitials (P;, Cd / , and especially Ba.) have overall high formation energies and should not be a concern for nonradiative carrier recombination. Based on these results, it may be advisable to assess quantitatively the effectiveness of Cdna, Pea, and Cdp as nonradiative recombination centers. It is noted that that cation-anion antisites have also been found to form deep defects in other phosphide solar absorbers, such as InP and Zmd .

[0061] The defect formation energies plotted FIGs. 5A and 5B provide important insights into doping in BaCd2P2. Irrespective of the elemental chemical potentials, BaCd2P2 clearly will not be highly doped, either p-type or n-type, by the intrinsic point defects. The equilibrium Fermi-level position, which can be estimated to be near the crossing point between the formation-energy curves for the dominant acceptor and donor species (such as Fed and Fp in FIG. 5A), never gets close to the band edges, implying very low free carrier concentration. Extrinsic acceptor doping could potentially make BaCd2?2 p-type doped, especially under the P-rich condition, where the Fp donor and other donor defects have overall high formation energies, meaning weak compensation (FIG. 5BB). In contrast, n-type doping in BaCd2P2 is likely to be challenging due to strong compensation from Fha and Fed, whose formation energies can drop to nearly or even below zero for Fermi-level positions close to the CBM. This doping asymmetry agrees with the relatively high absolute energy of the valence-band edge of BaCd2?2. Thus, it is clear that BaCd2?2 is a natively semi-insulating material with high potential of realizing good p-type doping. It could be used as an intrinsic layer in a p-i-n junction or as a p-type layer in a p-n junction.Deep defects and nonradiative recombination in BaCdiPi

[0062] Strong nonradiative carrier recombination requires deep defects with a high enough concentration but also large carrier capture coefficients. As can be seen in FIGs. 5A-5B, the deep17QB\98203989.1defects with low formation energies are CdBa, Ped, and Cdp. The capture coefficients can be calculated from first principles, considering nonradiative transition between the deep defect level and an electron at the CBM or a hole at the VBM via multiphonon emission. Applicants calculated capture coefficients for CdBa, Ped, and Cdp and determine nonradiative carrier recombination rates in BaCd2P2. Here it is assumed that intrinsic BaCd2?2 is being used as an absorber layer in a p-i-n device. The calculation is described in reference to the remaining sub figures of FIG. 5.

[0063] The CdBa gives rise to a deep (0 / -) level at O.eV above the valence (FIGs. 5A-5B). However, the configuration coordinate diagram for CdBa (0 / -) transition indicates that the electron capture barrier by Cd° and the hole capture barrier by Cd Ba are both large. Therefore, without explicitly computing the capture coefficients, CdBa should not be a strong recombination center despite being a deep-level defect. The Ped has a (+ / -) transition level near the mid gap. The direct transition between - 1 and + 1 states will not be efficient due to the low probability of a defect capturing two electrons (or two holes) at once. The recombination process of Ped can be treated by including the unstable neutral charge state (q = 0) and through two steps involving the (0 / -) and (0 / +) transition levels, as previously done for the iodine interstitial (L) in MAPI3. This amounts to four capture processes and coefficients in total: CO and C+ for electron ( / / ) and CO and C- for hole (p) capture, as shown in FIG. 5C. The Pea (0 / -) and (0 / +) levels are also very deep and locate 0.71 and 0.78 eV above the VBM, respectively. FIGs. 5D and 5E show the configuration coordinate diagrams for the Ped (0 / -) and (0 / +) transitions, respectively. All four capture processes have small energy barriers, which qualitatively indicates that the capture processes by Ped can be easy. Quantitatively, for Ped, our calculations — including electron-phonon coupling — yield a total capture coefficient (Got) of 2:77310- 7 cm3 / s at 300, which is significant but comparable with that of the most active recombination center (L) in MAPI3. The local atomic structures of P°, P’, and P+plotted in FIG. 5C highlight large displacements of the phosphorus antisite as a result of the charge transitions: the electron (hole) capture by P° to a downward (upward) displacement of the phosphorus antisite, and P'Cd (P+Cd) shows a trigonal planar (trigonal pyramidal) P4 configuration with three equidistant P-P bonds. The Cdp antisite gives rise to a deep (0 / 2 +) level in the band gap. Compared with Ped, Cdp presents a negligibly small Got at 300 K, mainly due to large hole capture.

[0064] The above results indicate that Ped is the most efficient nonradiative recombination18QB\98203989.1center in BaCrhPz, followed by Cdp, and that Cdiia is inactive. In the following, the nonradiative recombination rates in intrinsic BaCd2P2 at 300 K are shown. The nonradiative recombination rate (denoted as A) is given by A = Ctot 3 Nd, where Got and Nd are the total capture coefficient and concentration of a recombination center, respectively. FIG. 5F plots the nonradiative recombination rate A in intrinsic BaCd2P2 as a function of the elemental chemical potentials that control defect formation energy and concentration. The recombination rate is larger under P-rich conditions, which favor the formation of Ped. Remarkably, the recombination rate is at most ~105s'1, which is orders of magnitude smaller than that of MAPI3 (~10 s, computed at a similar level of theory). This is because the Ped concentration is rather low (on the order of 1011cm'3or less) in intrinsic BaCd2P2. Indeed, the lowest formation energy of Ped, even under the extreme P-rich condition, is higher than 1.0 eV (see FIG. 5B), which is still larger than the highest formation energy of the L in MAPI3 under the iodine-rich condition. It should be noted that that the Got of Ped is comparable with that of L in MAPbh. Using the calculated A rates, the nonradiative lifetime (= 1=^4) in intrinsic BaCd2?2 is on the order of at least 10 ms, indicating the potential for a remarkably long bulk carrier lifetime.

[0065] If BaCd2?2 is used as a p-type absorber layer in a p-n junction, the carrier capture behavior changes. In p-type BaCd2P2, the electron capture coefficients will dominate the total capture coefficient. As a consequence, not only Ped but also Cdp will act as nonradiative centers. In addition, the Ped and Cdp concentrations will be orders of magnitude higher when the Fermi level gets closer to the valence band (FIGs. 5A and 5B). Assuming that the Fermi level is 0.2 eV above the VBM, the nonradiative recombination rates are ~106— 108s'1. Such values are still on par with those in standard inorganic p-type solar absorbers CdTe, CIGS, and CZTS, where the computed A rates of ~107s'1have been typically reported.

[0066] Thus far, it has been described how BaCd2?2 is particular well suited for use as an absorber in a p-i-n junction PV cell and as a p-type absorber in a p-n junction PV cell. Thus far, this has been determined by the property computations described above in reference to FIGs. 4 and 5. What follows is a discussion of the experimental verification of these conclusions. Specifically, BaCd2?2 was synthesized and the resulting material was tested to confirm the computation-based conclusions discussed above.19QB\98203989.1Synthesis and stability of BaCdzPz

[0067] The computational filtering steps described above illustrated in connection with FIG. 2 (205-210), i.e., filtering the material search set for band gap, optical absorption, carrier mobilities, intrinsic point defects, and nonradiative carrier recombination at deep defects, computing theoretical efficiency, etc., highlight BaCd2P as a high-performance inorganic solar absorber made of low-cost elements. While BaCd2?2 has been reported in the literature, thus far no but no experimental characterization of the optoelectronic properties of this material had been made. To verify the computational analysis, BaCd2?2 powder samples were synthesized by solid- state reaction of the elemental Ba, Cd, and P in stoichiometric ratio in sealed silica ampoules. Certain physical and electronic properties of the synthesized sample are reflected in the plots of FIG. 6. FIG. 6A plots the powder X-ray diffraction (PXRD) pattern of the BaCd2?2 sample, which shows no coexistence of secondary phases and confirms the previously reported crystal structure. The crystal lattice parameters obtained from PXRD Rietveld refinement agree well with previous measurements and our calculations.

[0068] As highlighted by halide perovskites, whose applications are limited by their stability air- and moisture-sensitivity has become an important concern for the technological deployment of new solar cell absorbers. BaCd2P2 has the advantage over these materials in that it is highly stable in ambient air (for more than 6 months at room temperature) and that immersing a sample in water for 12 h causes no appreciable changes in the PXRD pattern. Additionally, BaCd2?2 is found to be resistant to 2.5 M KOH solution for at least 72 h, probably due to passivation of the surface with hydroxide, as indicated by the appearance of a small PXRD peak at 29° (20) corresponding to Cd(OH)2 (FIG. 6B). Only strong acids, such as HC1, concentrated HNO3, and aqua regia, can dissolve BaCd2P2. The PXRD pattern of BaCd2?2 after heating in the open air at 300oC shows no significant change in crystal structure (FIG. 5B). PXRD taken after a flow test in N2 gas reveals the sample to be stable even at 600 °C. In a vacuum (i.e., in an evacuated and sealed ampoule), annealing at temperatures of up to 1,100 °C results in no changes, attested to by the similar PXRD patterns. During thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC) in an ambient atmosphere, BaCd2P2 does not show any appreciable change until 425 °C, after which the material starts to oxidize (FIG. 6C). Morphology and overall stoichiometry of the synthesized BaCd2P2 are examined using scanning electron microscopy (SEM) and electron-20QB\98203989.1dispersive X-ray spectroscopy (EDS) (FIG. 4D). The EDS analysis reveals a nearly stoichiometric BaCd2P2 sample and, importantly, no impurity phases in the sample. Overall, all these results demonstrate a very stable BaCd2P2 material for which oxidation is probably kinetically limited and which should not require complex encapsulation and cumbersome processing.PL and carrier lifetime in BaCdzPz powder

[0069] Optoelectronic characterization of the BaCd Pz powder samples was also performed. Certain optoelectrical properties of the synthesized sample are reflected in the plots of FIG. 7. PL spectra were collected using a micro-PL setup in air, which allows the targeting of individual grains to conduct a statistical analysis of the PL intensity. FIG. 7A shows a pronounced PL peak at 1.46 eV (847 nm) at 298 K, in very good agreement with the HSE band gap (1.45 eV) as well as the step in the transmittance spectra of BaCdzlE (shown in the same figure). This suggests that the PL peak at 1.46 eV is due to band-to-band radiative recombination. A weaker PL peak at 1.27 eV (980 nm) is also observed, probably due to a shallow defect level close to the band edge. Notably, the intensity of the PL peaks only slightly decreases when the sample is heated up to 368 K and is fully recovered upon cooling back to 298K (FIG. 7A). The robust and strong PL spectra confirm that BaCd2P2 is also highly stable in terms of its optoelectronic properties in the entire temperature range of 298-323 K for typical solar cell operation.

[0070] To assess the PL efficiency more quantitatively and benchmark it versus a well-studied solar absorber material, the PL spectra of BaCdzP and GaAs powders under identical excitation conditions are compared in FIG. 7B. The GaAs sample was obtained by pulverizing a piece of prime-grade single-crystal GaAs (001) wafer into a fine powder. GaAs is chosen because it is a well-established, high-efficiency solar absorber and has a similar band gap. The BaCdzPz and GaAs powder samples have similar particle sizes, so the expected result is comparable degrees of surface recombination and multiple scattering optical losses in both samples. Remarkably, FIG. 7B shows that the PL peak intensity of the unoptimized BaCd Pz sample is already in the same order as that of the GaAs sample (i.e., only ~3 times weaker). This is also clearly seen in FIG. 7B inset, which shows the statistical histograms for the 100 random spots on each sample. The result is encouraging because BaCdzP , as a new solar absorber material, lacks the level of purity of the prime GaAs wafer.21QB\98203989.1

[0071] For a more quantitative study of carrier transport and excited-state dynamics in BaCd2P2, time-resolved and steady-state microwave conductivity (TRMC and SSMC) measurements on BaCd2?2 powder were performed. Microwave conductivity measurements employ a microwave field to probe the conductive and dielectric properties of semiconductor materials without the need for electrical contacts, making them particularly suitable for characterizing powder samples of emerging materials before high-quality thin films are available. The results of this analysis are presented in FIG 7, specifically the material’s, equilibrium (i.e., dark) properties, its transient carrier dynamics (FIGs. 7C-D), and its photo- conductivity action spectrum. Some of these dates are presented in the supplemental materials attached in Appendix 1. FIG. 7C shows photoconductivity transients expressed as a product of charge carrier yield and mobility sum for three representative laser fluences. The solid blue lines show a fit to the data employing three exponential decay components from which the population-weighted average carrier lifetime is calculated. As seen in FIG. 7D, the carrier lifetime is within 10-30 ns, depending on fluence; lower fluences result in longer lifetimes. An up to 30-ns photoconductivity lifetime is promising for unoptimized BaCd2P2 powders because the surface recombination effects are likely at play. For comparison, such a carrier lifetime already exceeds those measured for CdTe (before 2014) and CZTS (present) thin films. Even lead halide perovskites, which are well known for their long carrier lifetimes, started with carrier lifetimes of a few to a few hundred ns measured in thin films. As discussed above, carrier lifetime is strongly correlated with the efficiency potential and prospects of a new solar cell absorber.

[0072] The yield-mobility product extracted from TRMC provides a weighted sum of the electron and hole mobilities, assuming that every absorbed photon creates an unbound electronhole pair (4 = 1). Here, its value is calculated using the sum of the pre-exponential factors from the fits in FIG. 7C so as to account for the recombination that occurs within the instrument response function. These values are also fluence dependent. Interestingly, the yield-mobility product increases with fluence, from 1 cm2 / Vs at the lowest fluence to 4 cm2 / Vs at the highest fluence. Similar behavior is observed in steady-state photomodulation, where photoconductivity rises superlinearly (DsfFl :5) with light flux (F). Under most circumstances, the steady-state photoconductivity is expected to be either linear (first-order recombination) or sublinear (DsfF0:5 for bimolecular recombination). Both observations suggest that the present mobility and lifetime results are limited by a significant population of trap sites: as the injection density in- creases and22QB\98203989.1traps become filled, the average carrier mobility increases. This observation could be associated with the high surface area of the powder under investigation (FIG. 6D - inset). The steady-state photoconductivity action spectra and the PL spectra (FIG. 7A) are consistent with this surface trap hypothesis. The former exhibits a non-zero tail to the red of the evident absorption onset near 855 nm (i.e., 1.45 eV), and the latter shows a broad second emission peak in the same region.

[0073] Finally, it should be noted that the BaCd2?2 powder measured by TRMC has an XRD coherence length of only -100 nm, which likely suggests that the microwave frequency-dependent mobilities reported above are severely limited by scattering effects due to crystalline grain boundaries or extended structural defects. Indeed, using a confinement length of 100 nm and an observed 9.6-GHz mobility of 4 cm2 / Vs yields an intrinsic mobility of -100 cm2 / Vs, in agreement with the computed mobilities. It is worth noting that in the early-stage development of halide perovskite solar cells, the carrier mobilities in MAPb were also within the order of magnitude measured for BaCd2?2 but increased with improvements in film quality.

[0074] The computational and experimental results described above indicate that BaCd2?2 is a promising solar absorber material for single-junction solar cells. It combines an optimal band gap (1.45 eV) with good transport properties and its measured carrier lifetime on un- optimized powder is high (up to 30 ns). Critical to large-scale PV deployment, BaCd2P2 does not contain any critical elements and shows high stability in air and water. Even using non-optimized samples, BaCd2P2 already compares favorably to other thin-fdm absorber materials. No established thin- film absorbers possess BaCd2P25s combination of long carrier lifetime, earth-abundance (outperforming CdTe and CIGS), and high stability (outperforming the halide perovskites). While the toxicity of Cd is a legitimate concern, as for Pb in the halide perovskites, element toxicity has so far not stopped the development of either CdTe or halide perovskite solar cells. From previous studies of CdTe solar cells, it has become clear that the toxicity risks can be properly managed by PV module end-of-life recapture and recycling of Cd. Additionally, several other Zintl phosphides crystallize in the same structure as BaCd2P2, including the less-toxic CaZn?P2 and SrZn2?2. In view of the exceptional properties exhibited by BaCd2P2, these Zn-based compounds are of interest for future studies and could directly eliminate the toxicity concerns.

[0075] A recent report by Quadir et al regarding CaZn2P2 demonstrate time-resolved23QB\98203989.1microwave conductivity measurements indicate a photoexcited carrier lifetime of up to ~30 ns at low laser fluence for a freshly prepared fdm. Calculations show the absence of low-formation- energy, deep defects in CaZn2?2. Additionally, Quadir et al demonstrates that crystalline thin fdms of CaZn2?2 can be synthesized at low temperature using a reactive sputtering approach. (Quadir, S. et al., Adv. Energy Mater. 2024, 14, 2402640)

[0076] Additional reports confirm BaCd2P2 has remarkable properties. Hautzinger et al report the synthesis of crystalline nanoparticles of BaCd2?2 and demonstrates these particles exhibit long- lived photoexcited carriers (~160 ns average weighting), as determined by time-resolved PL spectroscopy. Hautzinger also reports fabrication of a BaCd2?2 thin film by a solid state ligand exchange protocol. (Hautzinger, M. et al., ACS Nano, 2025, 19, 12345-12353) Kassa et al report growth of BaCd2P2 crystals having surface areas of ~ 1 mm2were grown using Sn flux. The carrier lifetime was estimated using photoconductive and dark currents, and found to be ~ 300 ns at a laser power of 1.5 kW / m2. (Kassa, G. et al., “BaCd2P2: a defect-resistant “GaAs” https: / / doi.org / 10.48550 / arXiv.2506.20829) Pike et al report a study of the ANLP (A= Ca, Sr, Ba, Yb, Mg; M = Zn, Cd, Mg; and Pn = N, P, As, Sb, Bi) family of Zintl phases. Specifically included are the synthesis of BaCd2P2, CaCd2P2, SrCd2P2, and BaCd2As2 powders and CaZn2?2 and SrZn2P2 thin films and photoluminescence spectroscopy studies to measure the direct band gaps of these materials, as well as computational evaluation of numerous other members of the AM2Pn2 family. (Pike, A, et al., “Map of the Zintl AM2Pn2 Compounds: Influence of Chemistry on Stability and Electronic Structure”, Chemistry of Materials, 2025, 37(13), 4684-4694 DOI: 10.1021 / acs.chemmater.5c00353)

[0077] The properties of BaCd2?2 discussed above demonstrate that the material is suitable for use as an absorber in thin film solar devices, including single junction p-n and p-i-n type devices. Thankfully, all of the constituent elements of BaCd2P2 are already utilized in commercial optoelectronic thin-film technologies. While single crystal films of the material are preferred, in view of the long carrier lifetime, even at the powder level, polycrystalline BaCd2P2 films may be sufficient for high solar cell performance. As far as devices are concerned, based on the band gap and predicted dopability of BaCd2P2, it is advisable to target a single-junction solar cell adopting either a p-n or p-i-n junction device architecture, with BaCd2?2 used as a p-type or intrinsic absorber layer, respectively. Such devices would use suitable n-type buffer layers (for p-n junction)24QB\98203989.1and hole- and electron-selective layers (for p-i-n junction).

[0078] Additionally, the work described above suggests that the Zintl AM2X2 pnictide family offers band-gap-tuning opportunities through chemical substitution, which would open the door to the use of these Zintl-based materials for tandem cells. FIGs. 8 and 9 plot direct and minimum bandgaps for various AAhPm compounds, where A= Ba, Sr, Ca, Mg, M= Mg, Cd, Zn, and Pn= Bi, As, Sb, P, N). As can be seen, these materials demonstrate band gaps that range from very small (i.e., almost metallic) to well above what would be suitable for use as PV absorbers, but include attractive candidate materials within the ~1 5ev range. Solar absorbers using any of these materials, as well as the HT, stable materials shown in FIG. 3 are within the scope of the invention.

[0079] More generally, a compound of formula AM2X2 crystallizing in the crystal structure depicted in Fig. 10 in the space group P-3ml with A sites (large atoms; octahedral), M sites (medium atoms; tetrahedral) and X sites (small atoms). The A sites can be occupied by a single or mixture of alkali-earth and related elements (e.g., Mg, Ca, Ba, Sr, Eu, ...), M can be occupied by a single or mixture of 2+ elements such as Cd, Zn or Mg (and others), and X can be occupied by a single or mixture of 3- anions such as N, P, Sb, As and Bi (and others). It is anticipated that alloys of different combinations of materials, which each material being a compound having the immediately aforementioned form will have qualities that will produce suitable solar absorbers. Alloys of different such materials may be advantageous as a means to tune the band gap of the absorber layer.

[0080] Any other compound including quaternary and quinaries derivating from this crystal structure including the ones deviating from the A=+2, M=+2 and X=-3 trend including for instance, quaternaries with A occupied by a +1 element (e.g., Na or Rb) and M occupied by a +3 element and +2 element (e.g., Zn and Al) and X a -3 anion may also be used. Such compounds are expected to have similar properties to the materials described above.

[0081] In light of the principles and example embodiments described and illustrated herein, it will be recognized that the example embodiments can be modified in arrangement and detail without departing from such principles. Also, the foregoing discussion has focused on particular embodiments, but other configurations are also contemplated. In particular, even though expressions such as "in one embodiment", "in another embodiment", "in an embodiment", or the25QB\98203989.1like are used herein, these phrases are meant to generally reference embodiment possibilities, and are not intended to limit the invention to particular embodiment configurations. As used herein, these terms may reference the same or different embodiments that are combinable into other embodiments. As a rule, any embodiment referenced herein is freely combinable with any one or more of the other embodiments referenced herein, and any number of features of different embodiments are combinable with one another.

[0082] Although the invention has been described in considerable detail with reference to certain embodiments, one skilled in the art will appreciate that the present invention can be used in alternative embodiments to those described, which have been presented for purposes of illustration and not of limitation. Therefore, the scope of the appended claims should not be limited to the description of the embodiments contained herein.26QB\98203989.1

Claims

CLAIMSWhat is claimed is:

1. An absorber layer in a photovoltaic cell comprising a compound of the form AM2Pn2, where A= Ba, Sr, Ca, Mg, M= Mg, Cd, Zn, and Pn= Bi, As, Sb, P, N).

2. The absorber layer of claim 1, comprising BaCd2P2.

3. A photovoltaic device comprising an absorber layer comprising a compound of the form AM2Pn2, where A= Ba, Sr, Ca, Mg, M= Mg, Cd, Zn, and Pn= Bi, As, Sb, P, N).

4. The photovoltaic device of claim 3, wherein the absorber layer comprises BaCd2P2.

5. The device of claim 4, wherein the absorber layer has been doped such that it is a p-type material.

6. The device of claim 5, wherein the device includes a p-n junction comprising an emitter layer of n-type material adjacent to the absorber layer.

7. The device of claim 6, further comprising a transparent top electrode adjacent to the emitter layer and a bottom electrode adjacent to the absorber layer.

8. The device of claim 4, wherein the device comprises an n-type, an intrinsic and a p-type layer, and wherein the intrinsic layer comprises the absorber layer.

9. The device of claim 8, wherein the BaCd2P2 material is undoped.

10. The device of claim 4, wherein the absorber layer comprises polycrystalline BaCd2P2.

11. The device of claim 4 wherein the material of the absorber layer has an energy gap between its valence and conduction band in the range of 1 to 2 eV.27QB\98203989.

112. The device of claim 4, wherein the device is a thin film solar cell.

13. A method of identifying materials suitable for use as solar cell absorber layers comprising: identifying an initial material search set; sequentially subjecting the initial material search set to filters, the application of which excludes materials on the basis of thermodynamic stability, electronic properties, defect properties and optical conversion efficiency.

14. The method of claim 13, wherein exclusion on the basis of electronic properties comprises exclusion on the basis of band gap and carrier effective mass.

15. The method of claim 13, wherein exclusion on the basis of defect properties comprises exclusion on the basis of the presence of low-formation-energy, deep defects that can act as strong nonradiative recombination centers.

16. The method of claim 13, wherein the initial material search set includes only inorganic materials.

17. The method of claim 13, wherein the filters further exclude materials on the basis of environmental properties.

18. The method of claim 17, wherein exclusion on the basis of environmental properties includes exclusion of materials that are unstable in air or as a result of exposure to water.

19. An absorber layer in a photovoltaic cell comprising at least one compound having a crystalline structure in the space group P-3ml, where the A sites are occupied by a single or mixture of alkali-earth elements, M sites are occupied by a single or mixture of 2+ elements and X sites are occupied by a single or mixture of 3- anions.28QB\98203989.

120. The absorber layer of claim 19 comprising an alloy of two or more compounds having a crystalline structure in the space group P-3ml, where the A sites are occupied by a single or mixture of alkali-earth elements, M sites are occupied by a single or mixture of 2+ elements and X sites are occupied by a single or mixture of 3- anions.29QB\98203989.1

Images