Investigating the stable structures of yttrium oxide clusters: Y_n clusters as promising candidates for O₂ dissociation

Abstract

This study presents threshold photoionization (PI) spectra for a series of yttrium oxide clusters (Y_nO_m, n = 2–8, m = 2–4) in the photon energy range of 192 to 300 nm (6.46 to 4.13 eV). Density functional theory (DFT) is employed to explore the stable structures of these clusters. For Y_nO₂ clusters, experimental PI spectra are compared with calculated spectra for the lowest-energy and near-lowest-energy structural isomers. Stable structures contributing to the experimental PI spectra are identified. Experimentally corrected adiabatic ionization energies for Y_mO₂ clusters are determined. The newly identified lowest-energy structure for Y₂O₂ differs from those in previous literature studies, while larger clusters show better agreement, primarily varying in oxygen binding sites. Molecular oxygen-absorbed configurations of yttrium oxide clusters are generally unstable or energetically unfavorable, with O₂ activation occurring via charge transfer from yttrium to oxygen. Climbing image nudged elastic band (CI-NEB) calculations indicate that Y_nO₂ forms in the ground state when an O₂ molecule is absorbed onto low- or under-coordinated sites such as corners or edges of Y_n clusters. This process involves the dissociation of the O–O bond, followed by the adsorption of individual O atoms at different sites on the Y_n clusters. Analysis of the total density of states (TDOS) and partial density of states (PDOS) reveals an increased orbital density near the Fermi level, indicating a strong reaction affinity between Y and O atoms.

---

1. Introduction

Gas-phase metal oxide clusters are valuable model systems for studying reaction kinetics and energetics at the molecular level without external perturbations. Their unique properties enable highly accurate quantum chemistry methods to explore single-site catalytic mechanisms.¹ Identifying active sites is essential for understanding catalysis and designing better catalysts. Among metal oxides, transition metal oxides such as those of yttrium and lanthanum are widely studied for their catalytic and supportive roles in industrial and environmental applications. These include methane oxidative coupling,²n-butane cracking,³ ammonia decomposition⁴ and CO₂ adsorption.^5,6 Yttrium oxide clusters, in particular, act as electron donors and reduce substrate alkalinity, enhancing catalytic activity. Their ability to form in both stoichiometric and non-stoichiometric ratios makes them ideal for investigating charge transfer mechanisms that are critical to catalytic processes in bulk materials. Clusters of yttrium and yttrium monoxide are reactive toward CO₂ reduction⁷ and CO oxidation,⁸ with their activity determined by factors such as their structure, spin states, charge, and bond dissociation energies. Some clusters exhibit high stability and low reactivity, while others are highly active.

The study of yttrium and yttrium oxide clusters has been an area of active research for several decades. Knickelbein⁹ (1995) first investigated these clusters in the gas phase using a laser vaporization setup, reporting ionization energies (IEs) for Y_n and Y_nO (n = 2–31) but finding no magic behaviour. Subsequent work by Pramann et al.¹⁰ (2001) provided photoelectron spectra of anionic clusters, shedding light on the electronic structure evolution with oxygen and yttrium addition. Yang and co-workers later assigned ground-state structures for smaller clusters, while Yang and Xiong extended these calculations to Y_nO (n = 2–14).¹¹ Reed and Duncan¹² explored photodissociation pathways, revealing sequential Y₂O₃ losses in YO(Y₂O₃)_x⁺ clusters. Furthermore, Rahane et al.¹³ performed ab initio studies on (Y₂O₃)_n clusters, linking their structures to bulk cubic C–Y₂O₃. Our previous work¹⁴ experimentally assigned the ground-state structures of neutral and cationic Y_nO (n = 2–8) clusters based on agreement between experimental and simulated photoionization (PI) spectra, along with their adiabatic ionization energies (AIEs). We identified new structures of Y₂O, Y₃O, and Y₈O and found that ionization primarily arises from a delocalized 5s orbital rather than a localized 4d orbital. These insights into the bonding and electronic properties of yttrium oxide clusters form the foundation for our current investigation into clusters with higher oxygen content. Recent advancements include theoretical and experimental investigations to assign stable structures for small yttrium clusters, both neutral and cationic, using techniques like density functional theory (DFT) and coupled cluster methods (CCSDT).^11,15–19

In the present study, threshold photoionization spectra of yttrium oxide clusters containing 2–8 yttrium atoms and 2–4 oxygen atoms are investigated. The stable structures of yttrium monoxide clusters transform significantly upon oxidation. The lowest-energy structures of mainly the metal dioxide clusters are assigned by comparing the calculated PI spectra with the experimental spectra. The molecular oxygen-adsorbed configurations of yttrium oxide clusters are found to be unstable or energetically unfavourable. The O₂ adsorption on yttrium clusters is studied using climbing image nudged elastic band (CI-NEB) methods to obtain essential insights into the atomic and electronic interactions governing the adsorption process. By examining minimum energy path (MEP) profiles, DFT elucidates the reaction energetics and adsorption dynamics. The O₂ activation energy for the reaction is found to be 6 to 10 eV for the 2 to 8 atom yttrium clusters, which proceeds by charge transfer from yttrium to oxygen. Hybridization between the molecular orbitals of O₂ and the electronic states of yttrium leads to energy level shifts and a redistribution of charge density, which elucidates the bonding mechanism.

2. Experimental

The threshold photoionization spectroscopy experiment and data evaluation methods have been discussed previously.²⁰ In short, yttrium oxide clusters were generated via laser vaporization of yttrium foil in a pulsed flow of He/O₂ gas mixture (0.1% O₂). Cluster cooling and growth occurred within a 30 mm long, 3 mm diameter cluster formation source, extended by another 30 mm by attaching a 1 mm diameter nozzle. The clusters were photoionized with a tunable optical parametric oscillator laser (Ekspla, Lithuania, NT342B) within the ion extraction region of the reflectron time-of-flight mass spectrometer (Kore Technology, England). The photoionization spectra for all sizes were obtained by scanning the appropriate wavelength range of 192–300 nm (6.46 to 4.13 eV) at a low laser fluence to avoid multiphoton ionization. Measurements were repeated multiple times at different averages of ionizing events to minimize the short- and long-term fluctuations in the measurements and were checked for consistency. The PI spectra were normalized with respect to the cluster ion signal, obtained at a reference photoionization wavelength and at a fixed photon fluence to minimize cluster production fluctuation.

3. Computational

Density functional theory with Becke's three-parameter hybrid functional and the correlation functional of Lee, Yang, and Parr (B3LYP)²¹ was used to investigate the structures of yttrium oxide clusters. The calculations were performed using the GAMESS²² and Vienna Ab initio Simulation Package (VASP)^23,24 suite of programs. The LANL2DZ²⁵ basis set with an effective core potential (ECP) was used to represent 11 valence (4s²4p⁶4d¹5s²) and 28 core electrons of yttrium, respectively, and all eight electrons of oxygen. Initial geometries for the clusters were chosen from the previously reported structures of oxides of yttrium, lanthanum, and other transition metals.^17,18,26 Oxygen binding was investigated by probing all possible edge or face-capping sites of the metal clusters. The PI spectrum was calculated using the ezSpectra²⁷ program by calculating Franck–Condon (FC) overlap factors of all allowed vibrational transitions between the neutral and cation states of the lowest-energy isomer of a cluster. Structural assignments were carried out from the resemblance of the calculated PI spectra of a geometric isomer, not always the ones calculated as the lowest energy, with the experimental profile. Experimentally corrected adiabatic ionization energy (AIE) was determined from the experimental PI spectrum as detailed in our previous publication.¹⁴ The AIE corresponds to the 0⁰₀ transition, i.e. ionization from the ground state of the neutral to the ground state of the cation.

To investigate the NEB method, we performed DFT calculations using the VASP.^23,24,28,29 Initially optimized energy structures from GAMESS were further refined through re-optimization for the climbing image nudged elastic band (CI-NEB) calculations. During this process, the interactions between ion cores and valence electrons were described using the projector augmented wave (PAW) method.³⁰ The Perdew–Burke–Ernzerhof (PBE) exchange–correlation functional³¹ was employed for all calculations. Convergence criteria were set to 10⁻⁶ eV for total energy and 0.005 eV Å⁻¹ for atomic forces, with a plane-wave energy cutoff of 520 eV. To eliminate spurious interactions between the cluster and the surrounding vacuum, a 20 Å separation was maintained in all Cartesian directions. Structural optimizations used an 8 × 8 × 8 gamma-centered k-point grid for Brillouin zone sampling.

Following structural optimization, the CI-NEB method implemented in the VASP was used to identify transition states and saddle points along the reaction pathway. The initial and final states for CI-NEB were carefully chosen from the optimized geometries to ensure smooth interpolation between configurations. Intermediate images along the reaction path were iteratively relaxed until the perpendicular forces converged below 0.02 eV Å⁻¹. The MEP reveals the reaction barriers and transition dynamics, offering detailed insights into the energy landscape.

4. Results and discussion

DFT-calculated ground-state and other low-energy geometric structures of the neutral Y_nO₂ (n = 2–8) clusters, as obtained using GAMESS, are presented in Fig. 1 and the remaining in Fig. S1 of the ESI.† A few calculations were also repeated using Gaussian to check the consistency of relative energy order among low-lying structural isomers.³² Further details of the ground-state structures of the neutral form and the corresponding cation, such as the structures with bond lengths, Cartesian coordinates, and vibrational frequencies, are provided in Tables S1 and S2 of the ESI,† respectively.

Fig. 1 Stable structural isomers of Y_nO₂ (n = 2–8) clusters arranged in the increasing order of energy. The names of the structural isomers that contribute to the observed PI spectrum are underlined. The point group symmetry, spin states, and DFT calculated relative energies in eV are provided. The blue and red spheres denote yttrium and oxygen atoms, respectively.

4.1. Stable structures of Y_nO₂ (n = 2–8) clusters

The calculated lowest-energy structure of Y₂O₂ is rhombohedral in D_2h symmetry and in a singlet spin state (Fig. 1), corroborating well with the experimental photoionization spectrum as shown in Fig. 2. An O₂ molecule-absorbed configuration on a Y₂ dimer was calculated as the lowest energy configuration by Yang and Xiong.¹⁷ This initial configuration with a stretched O–O bond length of 1.41 Å compared to 1.21 Å in a free O₂ molecule is converged to the rhombohedral structure. The Y₂O₂ in its lowest energy state is similar to Ce₂O₂.³³ With further O–O bond stretching by 0.1 Å, the same initial configuration converges to an open Y–O–Y–O bend structure, which is 1.7 eV higher. Our calculations with several other initial structures either converged to these two stable isomers or were found significantly higher in energy (Fig. S2 of the ESI†). Previous literature studies suggested a planar boat-like structure for the anion, which is deemed unlikely based on structural robustness during oxidation or reduction.³⁴ We calculate a similar structure for the cation as the neutral form, likewise Sc₂O₂.³⁴ The computational study of the structural isomers of Y₃O₂ reveals that isomer 3a, featuring a capped and a bridging oxygen atom, is the lowest-energy configuration and shows excellent agreement with experimental PI spectra (Fig. 3). This structure has also been suggested as the ground state by Xu et al. in their CCSD(T) calculations.¹⁶ Isomer 3b, a V-shaped structure with oxygen atoms each bonded via single bonds, which is practically degenerate with 3a in our calculation, has a significantly different structure in its cationic form compared to the neutral form, and its FC factor for ionization is insignificant, making ionization of this isomer unfavourable. A planar 3c isomer is found to be 0.12 eV higher in energy than the lowest-energy 3a structure, with an insignificant FC factor for ionization. Additionally, an edge-capped planar C_2v structure (3d), previously calculated as the lowest-energy structure by Yang and Xiong,¹⁷ is 0.36 eV higher in energy than the lowest-energy 3a isomer. Optimization of several initial structures converged to either 3a or 3b, highlighting their stability as the most favourable configurations (Fig. S2†).

Fig. 2 (a) The simulated PI spectrum of Y₂O₂ (violet solid line) is overlaid against the experimental spectrum (blue spheres). The calculated ZEKE spectrum for the 1 → 2 ionization process is shown below the PI spectrum in blue. The AIE is labelled with a solid black arrow. (b) DFT calculated the lowest-energy geometric structures and energy levels of the neutral and cation clusters of Y₂O₂. The solid blue arrow shows the 1 → 2 transition. (c) Simulated ZEKE spectrum of the 2a 1 → 2 ionization transition of Y₂O₂. Prominent vibronic transitions are assigned.

Fig. 3 (a) The simulated PI spectrum of Y₃O₂ (violet solid line) is overlaid against the experimental spectrum (blue spheres). The calculated ZEKE spectrum for the 2 → 1 ionization process is shown below the PI spectrum in green. The AIE is labelled with a solid black arrow. (b) DFT calculated the lowest-energy geometric structures and energy levels of the neutral and cation clusters of Y₃O₂. The solid green arrow shows the 2 → 1 transition. (c) Simulated ZEKE spectrum of the 3a 2 → 1 ionization transition of Y₃O₂. Prominent vibronic transitions are assigned.

For Y₄O₂, we calculate a bicapped tetrahedron (slightly distorted) 4a, as the lowest-energy configuration in agreement with the previous paper.¹⁷ But there are two more isomers of similar motif, 4b and 4c (Fig. 1), differing in position and coordination number of the two O atoms, which are calculated within 0.1 eV of the lowest-energy structure. However, the calculated FC intensities for allowed ionizations for the spin states S = 1, 3 of the neutral form to the S = 2, 4 of the ion, for the 4a and 4b isomers, are negligible. The FC intensity for ionization is significant for the isomer 4c, and the calculated PIE spectrum also reproduces the experimental spectrum excellently (Fig. S3†).

The Y₅O₂ cluster prefers bicapped triangular bipyramids (TBPs) and the Y₆O₂ cluster prefers bicapped octahedra. For Y₅O₂, three TBP isomers 5a to 5c (Fig. 1) are calculated within 0.5 eV of the lowest-energy isomer, 5a. The 5a isomer has the lowest energy, but its FC intensity is significantly lower than that of 5b, making it less likely to contribute to the PIE spectrum (Fig. S4 of the ESI†). The 5b isomer, which is calculated to be 0.23 eV higher in GAMESS and only 0.10 eV higher in VASP calculations, was, however, calculated as the ground state in a previous study¹⁷ The agreement between the calculated and experimental PI spectra is also fair for this isomer. The potential energy surface for Y₆O₂ is flat. Three bicapped Y₆ octahedron structures, differing in oxygen binding sites, have similar energies near the lowest energy isomer 6a, with the other two, 6b and 6c, only 0.05 eV higher. They also have similar FC overlap factors. However, only the simulated PI spectra of the 6b isomer appear like the experimental spectra (Fig. S5 of the ESI†). The other two isomers are ruled out from contributing to the observed PI spectrum.

Several bicapped pentagonal pyramids (BPPs) are calculated as stable structures for Y₇O₂. These structures differ from one another by the bridging position of the oxygen atoms. Four such structures, 7a to 7d, named in ascending order of energy, are presented in Fig. 1. The FC factor for our calculated lowest-energy structure, 7a, is significant for the ionization, however, of a steeper slope, and does not reproduce the gentle sloped experimental PIE spectrum well, as shown in Fig. S6 of the ESI.† The isomer similar to 7c was calculated as the lowest-energy isomer by the authors of a previous study.¹⁷ The FC factors for the ionizing transitions for the other isomers are 10⁻³ to 10⁻⁴ orders of magnitude smaller and show even steeper PIE spectra. The other family of structures are octahedron based and the lowest one is higher by 0.25 eV. Since none of the low-energy isomers match the observed spectra, it is not possible to assign the actual ground-state structure with confidence.

For Y₈O₂, the most stable structure is the C₁ bicapped octahedron (8a), which agrees with a previous study by Yang and Xiong.¹⁷ The calculated PIE spectrum for the lowest-energy structure exhibits a steeper slope compared to the observed spectrum, though the overall agreement is reasonable. Unlike the Y₇O₂ clusters, the FC factors for the ionization transitions of the other low-energy isomers are similar to those of the lowest-energy isomer. Nevertheless, all calculated PIE spectra exhibit steeper slopes than the experimental results, as shown in Fig. S7 of the ESI.† The AIE values for Y_nO₂ (n = 2–8) and Y₃O₃ clusters are provided in Table 1, except for the Y₇O₂ cluster where AIE could not be determined from the experimental spectrum. For Y₇O₂, the threshold ionization energy (TIE) is provided.

Table 1 Experimentally corrected AIE values in eV for Y_nO₂ (n = 2–8, except n = 7) and Y₃O₃ clusters. TIE values of Y_nO_m (n = 3–5, m = 2–4, except 3 and 4) and Y₇O₂ clusters

Clusters	AIE (eV)	Clusters	TIE (eV)
Y2O2	5.23	Y3O2	4.68
Y3O2	4.74	Y3O3	4.38
Y3O3	4.37	Y4O2	4.72
Y4O2	4.74	Y4O3	4.73
Y5O2	4.19	Y4O4	4.80
Y6O2	4.08	Y5O2	4.12
Y8O2	4.49	Y5O3	4.21
		Y5O4	4.14
		Y7O2	4.26

---

4.2. Photoionization spectra of higher oxygen containing clusters (Y_nO_m, m > 2)

The simulated and experimental photoionization spectra of Y₃O₃ are shown in Fig. S8 of the ESI† and the remaining experimental PI spectra of Y_nO_m (n = 4–5, m = 3–4) are provided in Fig. S9 in the ESI.† The PI signals were weak or slopes indiscernible for clusters with m > 4, and consistent PI spectra for clusters with n > 5 beyond dioxides were unachievable due to low production. Stable structures of Y_nO_m (m > 2) were computed at the same DFT level, revealing discrepancies with previous results except for Y₃O₃. However, agreement between experimental and calculated PI spectra was insufficient to confirm calculated ground states. Further unbiased global minima searches are required to determine the ground states reliably. Nevertheless, the low-lying structures of these clusters, as calculated by us, are provided in the ESI (Fig S10–S11†).

In lack of similarity between the calculated and observed PI spectra, adiabatic IE could not be determined for these clusters; instead, TIE was determined by extrapolating the first linear rise of the PI signal to the baseline³⁵ and these are collected in Table 1. Fig. 4 shows the variation of the TIE values as a function of the number of oxygen atoms for Y_nO_m (n = 3–5, m = 0–4). The TIEs of Y_n and Y_nO (n = 3–5) were obtained from ref. 9 and 14, respectively, and the rest from our own measurement (Fig. S9 of the ESI†). There is no clear trend for the TIEs with n; likewise, it was observed for the Nb_nO_m clusters.³⁶ The TIEs decrease monotonically with cluster size for Y₃O_m and for Y₅O_m except at m = 3, and increase gradually for Y₄O_m with a discontinuity at m = 1. Knickelbein's results show that for smaller clusters, n ≤ 10, monoxides have slightly lower IEs by 0.1–0.3 eV than the bare clusters, and beyond that IEs are similar.⁹ It was explained in our previous paper¹⁴ that for the small Y_nO clusters, photoionization occurs from the metal-centred delocalized 5s or 5d–6s molecular orbitals (MOs), which is why monoxide clusters have similar IEs as their equivalent bare clusters. For the Al₃O_n⁻, Fe₃O_n⁻ anion clusters, the electron affinities or the binding energies were shown to increase with the oxygen content except for the first step from the metal-to-metal monoxide.^37,38 This was explained from a simple electrostatic model, considering clusters as a sphere, and a net charge transfer occurs from the metal cluster to the oxygen atom upon each O adsorption.

Fig. 4 Variation in the experimental threshold ionization energies as a function of oxygen atoms in the Y_nO_m (n = 3–5, m = 0–4) clusters.

For Y₃O₃, we calculate a hexagon C_s as the lowest energy configuration, followed by a D_2v hexagon and a C_s structure, respectively, 0.13 and 0.18 eV higher. The calculated PI spectrum of the C_s structure at 0.18 eV, obtained by adding an O atom on the Y₃O₂ lowest energy configuration, best reproduces the experimental PI spectrum (Fig. S8 of the ESI†). This configuration was also calculated as the lowest energy by Xu et al.¹⁶ The HOMO of Y₃O_m (m = 1–3) clusters hold an unpaired electron. The HOMO of Y₃O primarily consists of d orbitals from all three Y atoms; whereas in Y₃O₂ and Y₃O₃, the charge is located not on all Y atoms, and with no overlap, in a non-bonding state (Fig. S12 of the ESI†). This is perhaps the reason for the somewhat steeper monotonic reduction of the IE trend from m = 1 to 3.

4.3. Reaction study of O₂ adsorption on Y_n clusters

This study investigates the adsorption mechanisms of O₂ on Y_n clusters (n = 2–8), examining both molecular and dissociative adsorption modes. The Y_n clusters are modeled with specific geometries, and the O₂ molecule is strategically positioned in various orientations to determine the most stable adsorption configuration. In molecular adsorption, O₂ remains intact and interacts with the cluster via weak van der Waals or dipole–dipole forces. In contrast, dissociative adsorption leads to the breaking of the O–O bond, with individual oxygen atoms chemisorbing onto the cluster. The adsorption energy (E_b) is calculated as E_b = E_YnO2 − (E_Yn + E_O2), where E_YnO2, E_Yn, and E_O2 are the total energies of Y_nO₂, Y_n, and O₂, respectively. In this study, the complete dissociation of the O₂ molecule at the final stage renders the adsorption energy inapplicable. For fully dissociated states, binding energy (BE) is used, where a negative BE indicates stability. Initially, BE exhibits fluctuations before stabilizing as the cluster size increases, with a saturation trend observed beyond n = 7 (Fig. S13 of the ESI†). Low-coordination sites, such as edges and corners, play a crucial role in facilitating O₂ dissociation by lowering the reaction barrier. Fig. 5 shows the reaction pathway of O₂ absorption on Y_n to form the ground state of Y_nO₂ which involves dissociation of the O−O bond and then subsequent absorption of O atoms at different sites of Y_n clusters. The activation energy for dissociation sharply increases from 2.5 eV (n = 1) to 6.7 eV (n = 2), and then slowly increases to ∼10 eV at n = 8 with a slightly lower value of 7.8 eV for n = 7. In Fig. 5, there is a large energy gap between the first (left most) and the second points in each panel. The reaction at the initial stage is very fast due to the bond breaking. The large energy gap also indicates that the electronic structure is not continuous between the first and the immediate next image. The difference could be due to the change in charge state, spin state, or electronic occupation. These events can result in a sudden jump. This type of change is very common in transition metal clusters. Changes in electronic reconfiguration due to O₂ absorption and splitting can significantly alter the energy. Furthermore, the reaction with O₂, which involves charge transfer (see Bader charge transfer between the metal cluster and O₂ in Table S3 of the ESI†), moves the system from one electronic state to another. This causes the energy to jump. This often happens in the oxidation of metal clusters and catalytic processes where electron transfer is significant. Activation energy calculations, performed using the nudged elastic band method, reveal a size-dependent trend in adsorption stability. These findings offer valuable insights into the interaction of oxygen with yttrium clusters, contributing to their potential applications in catalysis and materials science.

Fig. 5 NEB path and corresponding reaction barrier height of O₂ dissociation on the Y_n cluster for n = 1 to 8.

4.4. Electronic structure, total density of states (TDOS), partial density of states (PDOS) and Bader charge analysis of the Y_nO₂ cluster (n = 1–8)

DFT calculations provide detailed insights into the electronic structure of yttrium clusters during O₂ adsorption, emphasizing the role of hybridization between Y-d and O-p orbitals in bond formation. Fig. S14 of the ESI† presents PDOS/TDOS for Y, O₂, and Y_nO₂, showing that the initially sharp and localized PDOS of isolated Y and O₂ broadens significantly upon adsorption, spanning from −5.0 eV to 3.0 eV. This broadening signifies strong orbital interactions and charge redistribution, enhancing bonding between yttrium and oxygen. Notably, the Y-d orbitals dominate hybridization, interacting with O-p states near the Fermi level, which drives adsorption and bonding. Bader charge analysis (Table S3 of the ESI†) further supports this by revealing that charge transfer from Y_n clusters to oxygen increases with cluster size, leading to greater charge accumulation around oxygen. This enhanced charge donation weakens the O–O bond, thereby lowering the NEB barrier for O₂ dissociation. Additionally, larger clusters provide more reactive sites due to increased surface area, further facilitating dissociation. These findings highlight Y_n clusters as promising candidates for O₂ dissociation, driven by their strong orbital interactions, enhanced charge transfer, and reduced energy barriers, making them highly reactive and effective catalysts.

5. Conclusions

This study combines threshold photoionization spectroscopy with DFT calculations to investigate a series of yttrium oxide clusters (Y_nO_m, where n = 2–8 and m = 2–4). For Y_nO₂ clusters (n = 2–8), the lowest-energy structures that reproduce the experimental PI spectra were confidently identified, and experimentally corrected adiabatic ionization energies were determined. Notably, the lowest-energy structures of Y₂O₂ and Y₃O₂ clusters differ significantly from those reported in previous studies, while larger clusters show better agreement despite variations in oxygen binding sites. The similarity between neutral and cation structures highlights the robustness of transition metal oxide clusters against oxidation or reduction.

Theoretical investigations indicate that molecular oxygen-adsorbed configurations are generally unstable or energetically unfavourable across all cluster sizes. Although it was not possible to experimentally confirm whether isolated yttrium clusters dissociate O₂ molecules on their surfaces—due to the use of O₂ and helium gases in the laser vaporization process—theoretical calculations using the nudged elastic band method reveal that O₂ dissociation on the surface of yttrium clusters is initiated by charge transfer from Y to O atoms, dissociation of the O–O bond, and subsequent reabsorption of O atoms at different sites. DFT calculations offered detailed insights into the electronic structure of yttrium clusters during O₂ adsorption. The interaction between the oxygen molecule and the yttrium cluster is revealed through the analysis of TDOS and PDOS. Hybridization between the molecular orbitals of O₂ and the electronic states of yttrium leads to energy level shifts and redistribution of charge density, which elucidates the bonding mechanism.

For clusters with higher oxygen content (m > 2), theoretical structural investigations were performed, and comparisons were made with the existing literature. However, it was challenging to achieve excellent agreement between the calculated and experimental PI spectra for all sizes, other than Y₃O₃.

Author contributions

V.V.D. – investigation, data curation, formal analysis, and simulation of PI spectra. D.B. – quantum chemical, NEB, and DOS calculations, and writing – original draft. V.C. – investigation, data curation, formal analysis, and quantum chemical calculations of Y_nO_mm > 2. G.K. – quantum chemical calculations of Y_nO₂. S.B. – conceptualization, investigation, methodology, project administration, writing – original draft, and review & editing.

Data availability

The data supporting this article have been included as part of the ESI.† If any additional questions may arise, the corresponding authors remain at the readers’ disposal.

Conflicts of interest

The authors have no conflicts to disclose.

Acknowledgements

D. B. is thankful to the Science and Engineering Research Board (SERB) for the research funding CRG/2022/003249.

References

1. M. Schlangen and H. Schwarz, Effects of Ligands, Cluster Size, and Charge State in Gas-Phase Catalysis: A Happy Marriage of Experimental and Computational Studies, Catal. Lett., 2012, 142(11), 1265–1278.
2. M. Haneda, Y. Katsuragawa, Y. Nakamura and A. Towata, Promoting Effect of Cerium Oxide on the Catalytic Performance of Yttrium Oxide for Oxidative Coupling of Methane, Front. Chem., 2018, 6, 414439.
3. F. S. Al-Sultan, S. N. Basahel and K. Narasimharao, Yttrium Oxide Supported La₂O₃ Nanomaterials for Catalytic Oxidative Cracking of n-Propane to Olefins, Catal. Lett., 2020, 150(1), 185–195.
4. Y. Zhu, H. Pan, Q. Li, X. Huang, W. Xi, H. Tang, W. Tu, S. Wang, H. Tang and H. Zhang, Dynamic Reconstruction of Yttrium Oxide-Stabilized Cobalt-Loaded Carbon-Based Catalysts During Thermal Ammonia Decomposition, Adv. Sci., 2024, 11(43), 2406659.
5. V. S. Derevschikov, R. G. Kukushkin, D. A. Yatsenko, D. I. Potemkin, P. S. Ruvinskiy, K. S. Shinkevich, T. S. Glazneva and V. A. Yakovlev, Study of the Impact of Y₂O₃ Support Synthesis on Carbon Dioxide Methanation Catalyst Applications, Ind. Eng. Chem. Res., 2022, 61(43), 15810–15819.
6. M. R. Sievers and P. B. Armentrout, Activation of Carbon Dioxide: Gas-Phase Reactions of Y⁺, YO⁺, and YO₂⁺ with CO and CO₂, Inorg. Chem., 1999, 38(2), 397–402.
7. Z. Zhao, X. Kong, Q. Yuan, H. Xie, D. Yang, J. Zhao, H. Fan and L. Jiang, Coordination-Induced CO₂ Fixation into Carbonate by Metal Oxides, Phys. Chem. Chem. Phys., 2018, 20, 19314–19320.
8. H.-L. Fang, L. Xu, J. Li, B. Wang, Y.-F. Zhang and X. Huang, Catalytic Oxidation of CO by N₂O on Neutral Y₂MO₅ (M = Y, Al) Clusters: A Density Functional Theory Study, RSC Adv., 2015, 5, 76651–76659.
9. M. Knickelbein, Photoionization Spectroscopy of Yttrium Clusters: Ionization Potentials for Y_n and Y_nO (n = 2–31), J. Chem. Phys., 1995, 102(1), 1–5.
10. A. Pramann, Y. Nakamura, A. Nakajima and K. Kaya, Photoelectron Spectroscopy of Yttrium Oxide Cluster Anions: Effects of Oxygen and Metal Atom Addition, J. Phys. Chem. A, 2001, 105(32), 7534–7540.
11. Z. Yang and S. J. Xiong, Structural Electronic, and Magnetic Properties of Y_nO (n = 2–14) Clusters: Density Functional Study, J. Chem. Phys., 2008, 129(12), 124308.
12. Z. D. Reed and M. A. Duncan, Photodissociation of Yttrium and Lanthanum Oxide Cluster Cations, J. Phys. Chem. A, 2008, 112(24), 5354–5362.
13. A. B. Rahane, P. A. Murkute, M. D. Deshpande and V. Kumar, Density Functional Calculations of the Structural and Electronic Properties of (Y₂O₃)_n^0,±1 Clusters with n = 1–10, J. Phys. Chem. A, 2013, 117, 5542–5550.
14. V. V. Deshpande, V. Chauhan, D. Bandyopadhyay, A. Anoop and S. Bhattacharyya, Structure of Small Yttrium Monoxide Clusters, Chemical Bonding, and Photoionization: Threshold Photoionization and Density Functional Theory Investigations, Phys. Chem. Chem. Phys., 2024, 26(29), 20123–20133.
15. N. S. Venkataramanan, Effect of Oxygen Content and Charge on the Structure, Stability, and Optoelectronic Properties of Yttrium Oxide Clusters, J. Phys. Chem. Solids, 2015, 82, 91–100.
16. L. Xu, C. J. Xia, L. F. Wang, L. Xie, B. Wang, Y. F. Zhang and X. Huang, Structural Evolution, Sequential Oxidation and Chemical Bonding in Tri-Yttrium Oxide Clusters: Y₃O_x⁻ and Y₃O_x (x = 0–6), RSC Adv., 2014, 4(104), 60270–60279.
17. Z. Yang and S. J. Xiong, Theoretical Study of Y_nO₂ and Y_nO₂⁻ (n = 1–8) Clusters, J. Phys. Chem. A, 2010, 114(1), 54–59.
18. Z. Yang and S. J. Xiong, Structural Electronic and Magnetic Properties of Small Yttrium Trioxide Clusters from Density Functional Calculations, J. Phys. B: At., Mol. Opt. Phys., 2009, 42(24), 245101.
19. Y. Misao, T. Nagata, M. Nakano, K. Ohshimo and F. Misaizu, Phys. Chem. Chem. Phys., 2022, 24, 11096–11103.
20. S. Bhattacharyya, D. Bandyopadhyay, S. Mukund, P. Sen and S. G. Nakhate, Ionization Energies and Ground-State Structures of Neutral La_n (n = 2–14) Clusters: A Combined Experimental and Theoretical Investigation, J. Phys. Chem. A, 2022, 126(20), 3135–3144.
21. C. Lee, W. Yang and R. G. Parr, Development of the Colle-Salvetti Correlation-Energy Formula into a Functional of the Electron Density, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37(2), 785–789.
22. G. M. J. Barca, C. Bertoni, L. Carrington, D. Datta, N. De Silva, J. E. Deustua, D. G. Fedorov, J. R. Gour, A. O. Gunina, E. Guidez, T. Harville, S. Irle, J. Ivanic, K. Kowalski, S. S. Leang, H. Li, W. Li, J. J. Lutz, I. Magoulas, J. Mato, V. Mironov, H. Nakata, B. Q. Pham, P. Piecuch, D. Poole, S. R. Pruitt, A. P. Rendell, L. B. Roskop, K. Ruedenberg, T. Sattasathuchana, M. W. Schmidt, J. Shen, L. Slipchenko, M. Sosonkina, V. Sundriyal, A. Tiwari, J. L. G. Vallejo, B. Westheimer, M. Włoch, P. Xu, F. Zahariev and M. S. Gordon, Recent Developments in the General Atomic and Molecular Electronic Structure System, J. Chem. Phys., 2020, 152(15), 154102.
23. G. Kresse and J. Furthmüller, Efficiency of Ab initio Total Energy Calculations for Metals and Semiconductors Using a Plane-Wave Basis Set, Comput. Mater. Sci., 1996, 6(1), 15–50.
24. G. Kresse and J. Furthmüller, Efficient Iterative Schemes for Ab Initio Total-Energy Calculations Using a Plane-Wave Basis Set, Phys. Rev. B: Condens. Matter Mater. Phys., 1996, 54(16), 11169–11186.
25. L. E. Roy, P. J. Hay and R. L. Martin, Revised Basis Sets for the LANL Effective Core Potentials, J. Chem. Theory Comput., 2008, 4(7), 1029–1031.
26. S. Bhattacharyya and V. V. Deshpande, Threshold Photoionization and Density Functional Theory Investigations of Small Lanthanum Monoxide Clusters, La_nO (n = 2–10), J. Phys. Chem. A, 2023, 127(36), 7460–7469.
27. S. Gozem and A. I. Krylov, The ezSpectra Suite: An Easy-to-Use Toolkit for Spectroscopy Modeling, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2022, 12(2), e1546.
28. G. Kresse and J. Hafner, Ab Initio Molecular Dynamics for Liquid Metals, Phys. Rev. B: Condens. Matter Mater. Phys., 1993, 47(1), 558.
29. G. Kresse and J. Hafner, Ab Initio Molecular-Dynamics Simulation of the Liquid-Metal–Amorphous-Semiconductor Transition in Germanium, Phys. Rev. B: Condens. Matter Mater. Phys., 1994, 49(20), 14251.
30. P. E. Blöchl, Projector Augmented-Wave Method, Phys. Rev. B: Condens. Matter Mater. Phys., 1994, 50(24), 17953.
31. J. P. Perdew, K. Burke and M. Ernzerhof, Generalized Gradient Approximation Made Simple, Phys. Rev. Lett., 1996, 77(18), 3865.
32. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, G. A. Petersson, H. Nakatsuji, X. Li, M. Caricato, A. V. Marenich, J. Bloino, B. G. Janesko, R. Gomperts, B. Mennucci, H. P. Hratchian, J. V. Ortiz, A. F. Izmaylov, J. L. Sonnenberg, D. Williams-Young, F. Ding, F. Lipparini, F. Egidi, J. Goings, B. Peng, A. Petrone, T. Henderson, D. Ranasinghe, V. G. Zakrzewski, J. Gao, N. Rega, G. Zheng, W. Liang, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, K. Throssell, J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M. J. Bearpark, J. J. Heyd, E. N. Brothers, K. N. Kudin, V. N. Staroverov, T. A. Keith, R. Kobayashi, J. Normand, K. Raghavachari, A. P. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, J. M. Millam, M. Klene, C. Adamo, R. Cammi, J. W. Ochterski, R. L. Martin, K. Morokuma, O. Farkas, J. B. Foresman and D. J. Fox, Gaussian 16, Revision C.01, Gaussian Inc., Wallingford, CT, 2016.
33. R. A. Hardy, A. M. Karayilan and G. F. Metha, Investigating Charge Transfer Interactions in AuCe₂O_n Clusters Using Photoionization Efficiency Spectroscopy and Density Functional Theory, J. Phys. Chem. A, 2019, 123(46), 10158–10168.
34. J. R. T. Johnson and I. Panas, Structure, Bonding, and Redox Properties of Scandium Oxide Clusters, a Model Study, Chem. Phys., 1999, 248(2–3), 161–179.
35. P. Lievens, P. Thoen, S. Bouckaert, W. Bouwen, F. Vanhoutte, H. Weidele and R. E. Silverans, Evidence for size-dependent electron delocalization in the ionization potentials of lithium monocarbide clusters, Chem. Phys. Lett., 1999, 302, 571–576.
36. K. Athanassenas, D. Kreisle, B. A. Collings, D. M. Rayner and P. A. Hackett, Ionization Potentials of Niobium Cluster Oxides, Chem. Phys. Lett., 1993, 213(1–2), 105–110.
37. H. Wu, X. Li, X. B. Wang, C. F. Ding and L. S. Wang, Al₃O_y (y = 0–5) Clusters: Sequential Oxidation, Metal-to-Oxide Transformation, and Photoisomerization, J. Chem. Phys., 1998, 109(2), 449–458.
38. L. S. Wang, H. Wu and S. R. Desai, Sequential Oxygen Atom Chemisorption on Surfaces of Small Iron Clusters, Phys. Rev. Lett., 1996, 76(25), 4853–4856.

---

Footnote

† Electronic supplementary information (ESI) available: The supplementary material encompasses Tables S1–S3 including Cartesian coordinates of the lowest-energy structures, vibrational frequencies, and Bader charge analysis. Fig. S1–S14 include additional low-energy isomers of dioxide clusters, photoionization (PI) spectra of Y_nO₂ (n = 4–8) and Y₃O₃, PI spectra of Y_nO_m (n = 4 and 5; m = 3 and 4), low-lying structures of Y_nO₃ (n = 3–5) and Y_nO₄ (n = 4 and 5) clusters, molecular orbitals of Y₃O_m (m = 1–3), average binding energy of Y_nO₂ (n = 1–8), and DOS/PDOS of Y_nO₂ (n = 1–6). See DOI: https://doi.org/10.1039/d5dt00357a

---