Energy Landscape and Electronic Properties of Yttrium Oxysulfide (Y2O2S) Dragana Jordanov, Jelena Zagorac, Johann Christian Schön, Klaus Doll, Dejan Zagorac, Albena Paskaleva Abstract This study investigates the electronic properties and crystalline phases of yttrium oxysulfide (Y2O2S) using theoretical computational methods. A crystal structure prediction was employed to explore the compound’s energy landscape. A global optimization using empirical potentials was performed, followed by a local optimization via ab initio calculations using the generalized gradient approximation (GGA), the local density approximation (LDA), and the hybrid B3LYP functional. The computational results confirm the stability of the alpha phase, whose structure aligns with the experimentally observed trigonal modification. Furthermore, four previously unknown metastable polymorphic phases, β-, γ-, δ-, and ε-Y2O2S structures, were identified. An examination of the electronic structure reveals that Y2O2S, a wide-bandgap semiconductor, possesses an indirect energy gap. The results obtained show good agreement with the available literature. A further analysis of the structure–property relationship of the predicted Y2O2S polymorphs indicates the possibility of an indirect-to-direct bandgap transition, which could have significant applications in optoelectronics. 1. Introduction Materials containing rare-earth elements (REEs) exhibit chemical and physical properties of great interest to modern industrial processes with applications in many fields, including solar energy, wind turbines, batteries for electric vehicles and mobile phones, and color display devices. These materials are realized as glasses, metal alloys, or ceramics, where the REE is present as a constituent material, such as in the ceramic yttrium oxide [,], and/or as a dopant. The core motivation for selecting specific dopants in Y2O2S relies on a strategic balance of structural symmetry compatibility, low phonon energy exploitation, a wide bandgap (WBG), and intentional defect engineering. In this study, the focus is on yttrium oxysulfide (Y2O2S), which has received considerable attention as a phosphorescent host material and is widely used in the luminescent materials industry [,]. When doped with Co2+, Mn2+, and/or rare-earth cations, such as europium, terbium, or erbium, in various concentrations, yttrium oxysulfide exhibits luminescence and can be used as a cathodoluminescent material [,,,]. Due to the special electronic structure of Eu3+, it has been a preferred dopant: Eu-activated Y2O2S emits bright red light under cathode-ray excitation, and it has been widely used as an important material for color display devices [,]. In particular, when co-doped with Mg2+ and Ti4+, it is employed as a long-afterglow (persistent luminescence) material with excellent stability and high-quality performance [,,,,]. For example, the Y2O2S:Eu,Mg,Ti nanotube arrays, synthesized via the sol–gel template method, are used in biochemistry and medical devices due to their highly ordered, integrated structure []. Nevertheless, the luminescence properties of doped Y2O2S can be improved. While most of the effort tends to be invested in optimizing the choice of dopants and their concentrations [], the host material Y2O2S has not received much attention, and there have been only a few studies dealing with the optical properties of the pure bulk Y2O2S. Yttrium oxysulfide can be prepared as phosphor nanoparticles with infrared-to-green and blue upconversion emission [], doped yttrium oxysulfide nanostructures [], or quantum structures []. We note that an important feature of rare-earth oxysulfides is that those materials are wide-gap (4.6–4.8 eV) semiconductors []. The calculations for crystalline Y2O2S show that the material has an indirect bandgap and that the top of the valence bands exhibits an anisotropic character, i.e., an anisotropic mass of the hole in reciprocal space []. A promising alternative route to improve the properties of Y2O2S is to employ alternative polymorphs of yttrium oxysulfide, since such new modifications may exhibit improved electronic properties for their application as a host material in luminescence applications. To identify the potential candidates for new modifications of Y2O2S, we performed crystal structure prediction via global exploration of the energy landscape of yttrium oxysulfide, complemented by data-mining-based searches of the Inorganic Crystal Structure Database (ICSD). For the most promising structural candidates, we then investigated their electronic properties, demonstrating the richness of Y2O2S with regard to its structural and electronic features. 2. Computational Methods 2.1. Details of the Generation of Structural Candidates of Y2O2S In order to predict new crystal structures, global optimizations on the energy landscape of Y2O2S have been performed using a fast-computable empirical two-body potential:(1)V(ri→,rj→)=V(|ri→−rj→|)=V(rij)To describe the potential energy of the system,(2)Epot=∑〈i,j〉 V(ri→,rj→),the two-body potential consists of a van der Waals term and a damped Coulomb term:(3)V(rij)=εvdWσijrij12−σijrij6+e−μrijqiqj4πε0rij with μ = 1/0.1 Å and εvdW = 0.5 eV. Here,(4)σij=rion(i)+rion(j) is the sum of the effective ionic radii of the ions i and j, and qi is the charge assigned to ion i. In this study, we employ q(Y) = +3, q(O) = −2 and q(S) = −2 for the charges. The starting source of the radius parameters for the ions are crystallographic effective ionic radii, in particular those listed in the book by Emsley [], with r(Y3+) = 1.06 Å, r(O2−) = 1.32 Å and r(S2−) = 1.94 Å. For the purpose of their use in the empirical potential, the radii are multiplied by a scale factor, r_s, that is close to one, where the values of r_s are based on experience, and for special situations a range of r_s values are employed []. In this study, we use r_s = 1.1 for the Y and O ions, yielding rion(Y) = 1.166 Å and rion(O) = 1.452 Å for the effective radii of Y and O in the expression for the empirical potential. For the sulfur ions, we use two different radius parameters in the potential, r_s = 1.1 and r_s = 1.0, i.e., rion(S) = 2.134 Å and rion(S) = 1.94 Å, during the global searches on the energy landscape, i.e., we perform half of the simulated annealing runs with rion(S) = 2.134 Å and the other half with rion(S) = 1.94 Å. The reason is that the sulfur ion exhibits a rather large range of effective ionic radii in sulfide compounds, which should be reflected in searches for structural candidates; such a parameter variation has proven quite effective in determining structural candidates in ionic-type systems []. The infinite crystal structure is approximated with a periodic approximant consisting of Z = 2 formula units of Y2O2S inside a periodically repeated variable unit cell. Stochastic simulated annealing was employed as the global optimization technique, where both the positions of the atoms inside the unit cell and the unit cell parameters were allowed to be varied at random during every run. During each simulated annealing run, multiple local optimizations (stochastic quenches) were performed, yielding a total of 5000 structural candidates. The global searches were complemented by data-mining-based searches of the Inorganic Crystal Structure Database (ICSD, version 5.6.0, Data Release 2026.1) [] to identify additional structural candidates with the general formula AB2X2, following a procedure that has been successfully applied in the Ce2ON2 chemical system []. The data-mining process followed the structured Knowledge Discovery in Databases (KDD) framework. This process was executed through four sequential phases: (1) selection of the primary dataset, (2) preprocessing, (3) transformation, and (4) evaluation to interpret the final patterns []. The structural candidates obtained via the global optimization and data mining were analyzed for their symmetries and structural similarities using the SFND [], RGS [], and CMPZ [] algorithms implemented in the KPLOT software (v.9.6.15) [], while the VESTA software (Version 3) [] was employed to visualize the crystal structures. In particular, duplicate candidates were eliminated using the CMPZ algorithm. Promising candidates were identified based on their energy, the frequency of their appearance during global landscape explorations, and the degree of symmetry in their structures. Afterwards, a local optimization was performed for all promising candidates (around one hundred structural candidates from global optimization and data mining searches) using ab initio calculations. This was followed by another round of structural analysis, which generated the final set of the most relevant feasible new modifications in the Y2O2S system. 2.2. Details of the Ab Initio Calculations All total-energy and electronic-structure calculations were performed at the ab initio level using density functional theory (DFT) with the CRYSTAL code (versions 17 and 23) [,]. The local optimizations employed analytical gradients with respect to the atom positions [] and cell parameters [] and used a local optimization routine []. As a result, atoms can freely move without symmetry constraints, and cell parameters, especially cell sizes, are fully relaxed, and the cells can change shape and size []. All-electron basis sets (AEBSs) based on Gaussian-type orbitals (GTOs) were applied. In the case of yttrium, a [7s5p3d] basis set (Y_buljan_1999) was used as in refs. [,]. For S2−, a [5s4p1d] basis set was employed as in refs. [,], while for O2−, a [4s3p] basis set was applied as in refs. [,]. We tested additional Effective Core Potentials (ECPs) to include relativistic effects; however, they had no significant influence. Spin-Orbit Coupling (SOC) was not included in the calculations. Tolerances for the convergence of total energy are set to 10−7 Eh in structural and electronic property calculations. Furthermore, a k-point sampling net of size 8 × 8 × 8 with the Monkhorst–Pack scheme was employed for each modification []. We used a 30% Fock/KS mixing parameter (damping factor, α = 0.30) to stabilize Self-Consistent Field (SCF) convergence. In particular, several different functionals have been employed for the DFT calculations: the Local-Density Approximation (LDA-PZ), with a parametrization as in [], and the Generalized Gradient Approximation (GGA-PBE), with a parametrization as in [], and the hybrid B3LYP (Becke, three-parameter, Lee–Yang–Parr) functional []. We note that repeating the calculations with several ab initio computational methods for total-energy calculations, local minimizations, and electronic-structure calculations facilitates a robust assessment of data accuracy and the viability of structural candidates [,,]. By performing a series of isotropic structural deformations, we have computed the total ground-state energy (E) as a function of the cell volume (V). The E(V) curves serve as a critical tool for mapping the energy landscapes of inorganic crystals and predicting phase stability under pressure. We used the CRYSTAL program to perform an automated volume scan to compute E vs. V curves, which we then fit to various universal Equations of State (EOS). For each volume, we performed a full V-constrained geometry optimization. Finally, we used a third-order isothermal Birch–Murnaghan fit and a 4th-order polynomial fitting procedure to compute equilibrium properties and E(V) and H(p) curves. 3. Results 3.1. Energy Landscape and Crystal Structures Overall, 5000+ structural candidates were generated by the global landscape search and the data-mining procedure. After relaxing the most promising among these candidates on an ab initio (DFT) level, the four structures with the lowest energy belonged to the results of the global optimization, and one further low-energy candidate (ε-Y2O2S) to the outcome of the data-mining-based searches. The stability of the predicted crystal structures of Y2O2S is assessed using their computed total energies at the ab initio level. The modification with the lowest energy is the thermodynamically most stable one at standard conditions, i.e., for essentially zero pressure and low temperatures, while the higher-lying local minima of the energy serve as candidates for metastable phases at standard conditions and/or as possible candidates for high-temperature, high-pressure, or effective negative-pressure phases. A further investigation of stability and metastability is possible using phonon calculations or other dynamical-stability calculations. In , we show the E(V) curves for the five most relevant predicted structures of Y2O2S, i.e., the ones exhibiting the lowest energies, computed using the LDA-PZ method. We note that the α-phase (α-Y2O2S) corresponds to the lowest energy minimum (global minimum) and thus represents the thermodynamically most stable structure, in agreement with the data in the literature [,,,,]. Next in energy ranking is the predicted β-Y2O2S polymorph, followed by the γ-Y2O2S and δ-Y2O2S modifications, while the ε-Y2O2S modification appears as the highest energy minimum regardless of computational approach (). A summary of the computed total energy per formula unit of these five structures, where the ab initio calculations were performed using the LDA-PZ, B3LYP, and GGA-PBE approximations, is shown in (relative energies are given in the ). We note that the ranking of the minima by energy does not change when we employ a different functional, indicating the robustness of the results. The α-phase of yttrium oxysulfide obtained in this study as the global minimum on the energy landscape corresponds to the experimentally known phase of Y2O2S [,,,,] and is a global minimum on the energy landscape ( and ). The crystal symmetry of Y2O2S is trigonal, with the space group P-3m1 (no. 164). The α-Y2O2S structure consists of edge-connected seven-fold coordination polyhedra around the yttrium atom, which is surrounded by four oxygen atoms with an average distance of 2.23 Å and three sulfur atoms at a distance of 2.83 Å (a). The computed lattice constants are a = 3.83 Å and c = 6.69 Å using B3LYP; a = 3.80 Å and c = 6.67 Å using GGA-PBE; and a = 3.73 Å and c = 6.54 Å using LDA-PZ (). The predicted β-Y2O2S modification is a metastable phase of yttrium oxysulfide that appears in orthorhombic symmetry with the Cmc21 (no. 36) space group. This structure was frequently found in the global searches using empirical potentials; the ab initio optimization resulted in the lattice constants a = 3.65 Å, b = 15.22 Å, c = 5.93 Å using B3LYP; a = 3.64 Å, b = 15.14 Å, c = 5.88 Å using GGA-PBE; and a = 3.56 Å, b = 14.96 Å, c = 5.70 Å using LDA-PZ (). The structure of β-Y2O2S includes two types of edge-connected coordination polyhedra around the yttrium atoms. The first type is a seven-fold coordination polyhedron, with three oxygen atoms and four sulfur atoms (b). The average bond distance is 2.57 Å. The second type is an 8-fold coordination polyhedron, where yttrium is surrounded by six oxygen and two sulfur atoms, and the average cation–anion distance is 2.49 Å (b). The structure of the γ-Y2O2S phase exhibits the lowest symmetry among the five structures with the lowest energies, appearing in the monoclinic Pm (no. 6) space group. In γ-Y2O2S, yttrium is surrounded by edge-connected five- and seven-fold coordination polyhedra (c). In the five-fold coordination polyhedra, the average bond distance is 2.36 Å, while in the seven-fold coordination polyhedra, it varies between 2.49 and 2.54 Å. The unit cell parameters of the γ-phase of yttrium oxysulfide are a = 6.19 Å, b = 3.99 Å, c = 7.26 Å, β = 100.64° using B3LYP; a = 6.17 Å, b = 3.97 Å, c = 7.21 Å, β = 100.46° using GGA; and a = 6.05 Å, b = 3.89 Å, c = 7.02 Å, β = 99.94° using LDA (). The predicted δ-phase of yttrium oxysulfide also shows an orthorhombic symmetry similar to the β-phase, but with a different space group, i.e., the δ-Y2O2S modification appears in the Pmn21 (no. 31) space group. The calculated unit cell parameters are a = 3.82 Å, b = 7.50 Å, c = 6.09 Å using B3LYP; a = 3.80 Å, b = 7.43 Å, c = 6.08 Å using GGA-PBE; and a = 3.71 Å, b = 7.25 Å, c = 5.98 Å using LDA-PZ (). The δ-Y2O2S structure is based on edge-connected five- and seven-fold coordination polyhedra around the yttrium atoms, similar to the γ-Y2O2S modification (d). In the five-fold coordination polyhedra around the yttrium atoms, the average anion–cation distance is 2.36 Å, while for the seven-fold coordination polyhedra it is 2.48 Å. Finally, the ε-Y2O2S modification exhibits an orthorhombic symmetry with the same space group as the β-phase, i.e., Cmc21 (no. 36). However, the ε-phase is structurally distinct from the β-phase and all previous ones: the ε-Y2O2S structure consists of two types of edge-connected seven-fold coordination polyhedra around the yttrium atoms. The first type of these polyhedra shows an average cation–anion bond distance of 2.47 Å (four oxygen and three sulfur atoms), and in the second type (five oxygen and two sulfur atoms), the average bond distance is 2.50 Å (e). A summary of all the predicted crystal structures of Y2O2S computed using the B3LYP, GGA-PBE, and LDA-PZ functionals is shown in . 3.2. Electronic Band-Structure Calculations To explore the potential of the Y2O2S system for applications, the electronic properties and the size of the bandgap of the predicted theoretical modifications of Y2O2S were investigated and compared with the modifications reported in the literature. The DFT calculations were performed using the GGA-PBE, the LDA-PZ, and the B3LYP hybrid functionals, as in the structural relaxations. a shows the electronic band structures of the trigonal α-Y2O2S crystallizing in the P-3m1 (no. 164) space group, calculated along the high-symmetry path A-L-M-Γ–A-H–K–Γ of the Brillouin zone, as defined in ref. []. In the case of the α-modification, the valence band maximum (VBM) is located at the A point of the Brillouin zone, while the conduction band minimum (CBM) is located at the K point, indicating an indirect bandgap of 4.92 eV computed using the hybrid B3LYP functional (a), and 3.23 eV on the LDA-PZ level of calculation (). This is in good agreement with the literature data [,,,,,,]. Next, b shows the band structure of the predicted β-phase of yttrium oxysulfide. The β-Y2O2S structure appears in the orthorhombic Cmc21 (no. 36) space group, and we have computed the band structure along the high-symmetry path Γ–Y-T-Z-Γ-S-R of the Brillouin zone. In contrast to the α-phase, the VBM and CBM are both located at the Y point of the Brillouin zone in the β-modification, showing a direct bandgap of 4.07 eV using B3LYP and 2.28 eV using the LDA-PZ functional (b and ). We note that a possible indirect-to-direct bandgap transition associated with the switch between these two (meta)stable phases of Y2O2S could be of great importance in optoelectronic applications. The electronic band structure of the low-symmetry monoclinic phase of yttrium oxysulfide (γ-Y2O2S) crystallizing in space group Pm (no. 6) is discussed next (c). The calculations were performed along the path Γ–B-D-Z–Γ-Y-A-E-C-Z of the Brillouin zone. This monoclinic structure exhibits an indirect bandgap, where the VBM is located at the Γ point, and the CBM is located at the A point of the Brillouin zone with a bandgap size of 4.86 eV, calculated using the hybrid B3LYP method. Moreover, the indirect bandgap character was confirmed in the calculations using the LDA-PZ functional, yielding a bandgap of 3.17 eV (). d shows the results of the band-structure calculations of the orthorhombic δ-Y2O2S phase using the B3LYP functional. Since the δ-Y2O2S structure appears in space group Pmn21 (no. 31), the band structure is calculated along the complicated high-symmetry path Γ–X–S–Y–Γ-Z-U-R-T-Z of the Brillouin zone; this path is different from the one used for the β-phase, which also crystallizes in an orthorhombic structure but in a different space group. The δ-Y2O2S phase shows an indirect bandgap, where the VBM is located at the Γ point, and the CBM is located along the Z-U directions of the Brillouin zone. The indirect bandgap was found to be 4.58 eV using the hybrid B3LYP functional. An indirect bandgap of 2.80 eV was also found using the LDA-PZ approximation (). Finally, the electronic band structure of the ε-Y2O2S modification is shown in e. The ε-Y2O2S structure appears in the orthorhombic Cmc21 (no. 36) space group; since this is the same as that of the β-Y2O2S modification, we have computed the band structure along the same high-symmetry path Γ–Y-T-Z-Γ-S-R of the Brillouin zone (cf. b,e). In the case of the ε-phase, the VBM is located at the Z point, and the CBM is located at the Y point of the Brillouin zone, with a bandgap size of 4.91 eV using the B3LYP and 2.35 eV using the LDA-PZ functional (e and and ). We note that within the same chemical system (Y2O2S), we deal with two different modifications that exhibit the same symmetry and space group, and thus they might be expected to exhibit a similar qualitative band structure; however, the β-Y2O2S phase has a direct bandgap, while the ε-Y2O2S modification has an indirect bandgap. Thus, creating intergrowth structures that partake in both of these modifications might open new avenues for optoelectronic applications, e.g., possibly allowing controlled switching between indirect and direct bandgap behavior in a material. The calculated bandgaps obtained using different exchange–correlation functionals are summarized in and compared with previously reported theoretical and experimental values, where available. 4. Discussion The α-modification, the only experimentally known phase of yttrium oxysulfide (Y2O2S), has received considerable attention as a phosphorescent host material and is widely used in the luminescent materials industry [,,,,]. However, both with regard to the lattice constants and the bandgap values, not many experimental or theoretical data are available, and some disagreement exists among them. This is because isolating Y2O2S requires careful control of synthesis conditions, sulfurization potential, reaction templates, and thermal kinetics. The lattice constants computed in this study are a = 3.83 Å and c = 6.69 Å using the B3LYP functional, a = 3.80 Å and c = 6.67 Å using the GGA-PBE functional, and a = 3.73 Å and c = 6.54 Å using the LDA-PZ functional (). The previous theoretical data show a = 3.750 Å, c = 6.525 Å [], where the exchange–correlation energy is evaluated within the local density approximation (LDA) by using Ceperley–Alder homogeneous electron gas data [], and a = 3.730 Å, c = 6.48 Å [], by using the hybrid density functional sX-LDA for electronic structure calculations under the framework of the generalized Kohn–Sham density functional theory (GKS-DFT) []. Those calculated literature data are in close agreement with our LDA-PZ calculations. In contrast, the experimental reports of a = 3.7826 Å, c = 6.5689 Å [,] and a = 3.784 Å, c = 6.586 Å [] agree better with our present hybrid B3LYP and GGA-PBE results, consistent with the usually larger lattice parameters at the B3LYP and GGA level, compared to LDA-PZ calculations (cf. ). In this context, we note that one usually obtains slightly larger lattice parameters at the GGA level than in LDA-PZ calculations. On the other hand, LDA- and GGA-level bandgap calculations frequently underestimate the true bandgap; for these systems, the value obtained with B3LYP is often preferable. In particular, we further note that the location of the valence bands obtained from the DFT and especially from the GGA-PBE calculations is usually accurate, while the energy levels of the conduction bands are sometimes systematically underestimated, an aspect that needs to be kept in mind when comparing experimental bandgaps with computed ones using certain functionals. As a consequence, some studies explicitly adjust the computed bandgap by adding or subtracting some constant on the order of 1–2 eV; however, applying such a shift as a standard modus operandi strikes us as not really justified. As far as the electronic properties are concerned, the α-phase of Y2O2S is a WBG (4.6–4.8 eV) semiconductor that has been considered an ideal host matrix for trivalent rare-earth cations [,], as mentioned in the Introduction. However, Y2O2S is a specific wide-bandgap semiconductor that differs from more industrially accepted and advanced WBG semiconductors such as AlN [,,]. Moreover, an X-ray photoelectron spectroscopy study was conducted to obtain more detailed information on the valence band and the outermost core band, and the bandgap energy was estimated at 6.77 eV [], which greatly exceeds the commonly accepted value. On the other hand, Li et al. believe that the bandgap energy of 6.77 eV reported by Itoh and Inabe is overestimated and that the true bandgap should be around 4.6 eV () []. Furthermore, in previous theoretical calculations, the bandgap of Y2O2S was estimated at about 3.0 eV, which is 1.6 eV lower than the commonly accepted experimental value []. Assuming that the valence bands obtained from their DFT calculations were accurate and that the conduction bands were underestimated, a “scissor operation” of 1.6 eV was adopted in the conduction band so that the calculated bandgap agreed with the experimental value []. In the present study, the computed bandgap values were 3.23 eV (LDA), 3.46 eV (GGA), and 4.92 eV (B3LYP) for the α-Y2O2S modification. A survey of the existing literature underscores the pronounced impact of the exchange–correlation functional on bandgap calculations. Furthermore, considering the data for all five modifications in for the three functionals employed demonstrates that the bandgap estimates for a given structural modification vary substantially across the LDA, GGA, and hybrid-functional approaches. In particular, the present GGA and LDA calculations agree very well with the previous theoretical data for the α-modification, while the hybrid B3LYP calculations agree with the experimental data (). We also note that for each of the five modifications considered in this study, the smallest bandgap is obtained with LDA-PZ, while the largest bandgap is obtained with the hybrid B3LYP functional. However, the ranking of the size of the four indirect bandgaps was not the same for the three functionals: in the case of B3LYP, the order was Egap(α) > Egap(ε) > Egap(γ) > Egap(δ). On the other hand, for GGA-PBE, we find Egap(ε) > Egap(α) > Egap(γ) > Egap(δ), and for LDA-PZ we have Egap(α) > Egap(γ) > Egap(δ) > Egap(ε). This clearly demonstrates the remaining uncertainty implicit in the functionals typically used for bandgap calculations; nevertheless, the general trend is clear, and the differences in the energy gaps that lead to a different ranking are small. To further analyze bandgaps and perform comparative assessments of direct/indirect bandgap transitions, one can turn to Spin-Orbit Coupling (SOC) methodology [,], the Tran–Blaha modified Becke–Johnson (TB-mBJ) functional [], or Time-Dependent Density Functional Theory (TD-DFT) calculations []. However, regardless of the quantitative details of the computed bandgaps of the various feasible modifications we have considered in this study, the results clearly show that a tuning of about 1 eV in the bandgap should be possible in the yttrium oxysulfide system by switching from one (meta)stable phase to another. Beyond the fine-tuning of the indirect bandgap by using the α, γ, δ and ε modifications, the discovery of a direct-gap polymorph in the Y2O2S system, the β-phase, is of great importance and opens new possibilities for applications of Y2O2S beyond its traditional use as a phosphor host. The present results suggest that phase control could serve as a viable route for engineering the electronic properties of Y2O2S. In particular, the β-phase, with its direct bandgap, may be of interest for applications in thin-film transistors, light-emitting diodes, or photovoltaic devices, provided that synthesis routes can be developed to stabilize this metastable modification. While this clearly poses a major challenge to the experimental solid-state chemist and the materials scientist, the E(V) curves in already indicate that the β-Y2O2S modifications should be the high-pressure phase in the Y2O2S system. In addition, we have computed the enthalpy–pressure (H(p)) curves and determined the α → β transition pressure of approximately 30.8 GPa using the LDA-PZ functional (), and around 35 GPa using the GGA-PBE and 40 GPa using the B3LYP functional. Other possible synthesis routes might be designed following the recent combined chemistry approach, in which structure–prediction calculations and advanced electron microscopy techniques demonstrate the topochemical deintercalation/reintercalation of sulfur in a layered oxychalcogenide, suggesting additional options for the future design of novel metastable phases []. Synthesizing the β-phase would allow the realization of the indirect-to-direct bandgap transition (i.e., the α → β phase transition) we predict as being feasible in the Y2O2S system, which is of great importance in optoelectronic applications. This type of tuning of electronic properties constitutes another example of a polymorph-dependent bandgap transition behavior, similar to the one that has recently been observed in La2O2S, where a newly synthesized direct-gap phase, an indirect-gap high-pressure phase, and a pressure-driven phase transition were identified []. Adding another example in the form of the α → β phase transition in Y2O2S strongly suggests that polymorphism-driven electronic tuning may be a general feature of rare-earth oxysulfides. 5. Conclusions This study investigates the feasible crystalline phases of yttrium oxysulfide (Y2O2S) and their electronic properties using advanced computational methods. By employing crystal structure prediction, we explored the compound’s energy landscape through global optimization on the empirical potential energy landscape supplemented with data mining, followed by an ab initio refinement using the LDA-PZ, the GGA-PBE, and the B3LYP hybrid functionals. Our results confirm the thermodynamic stability of the experimentally known α-phase, yielding a structure in excellent agreement with the experimentally observed trigonal modification. Beyond the known phase, we identified four novel metastable polymorphs (β-, γ-, δ-, and ε-Y2O2S). The electronic structure analysis characterizes the α-phase as a wide-bandgap semiconductor with an indirect bandgap, in agreement with the experimental results. Crucially, the analysis of the electronic properties of the newly predicted polymorphs shows that a high-pressure phase, the β-modification, exhibits a direct bandgap and has the potential to realize an indirect-to-direct bandgap transition in the Y2O2S system, highlighting the material’s versatility for optoelectronic and photovoltaic applications.