Skip to main content
NCBI home page
Scientific Reports logoLink to Scientific Reports
. 2026 Jan 24;16:6128. doi: 10.1038/s41598-026-37195-6

Isotopic and defect analysis of enriched molybdenum oxide using EPR spectroscopy and DFT simulation

Razieh Hosseini 1, Javad Karimi-sabet 2,✉, Mehdi Janbazi 3,✉, Masoomeh Sharbatdaran 3, Mohsen Ashjari 1
PMCID: PMC12902113  PMID: 41580493

Abstract

Molybdenum isotopes, particularly 96Mo, 97Mo, and 98Mo—the precursors for the medical isotope 99mTc—are vital in healthcare, science of materials, and nuclear research. While electromagnetic separation efficiently enriches these isotopes, analyzing their abundance requires techniques sensitive to subtle changes in electronic structure caused by isotopic mass variations. In this study, X-band EPR spectroscopy was experimentally employed to provide a suitable and sensitive method for probing these effects. We enriched 96Mo, 97Mo, and 98Mo isotopes electromagnetically and purified the collected powders through a chemical refining process. The refined samples were characterized using XRD, SEM, ICP, ICP-MS, PL, and EPR. Ab initio calculations were applied to investigate the electronic, optical, and EPR properties with the computed band gaps, optical characteristics, and EPR spectra validated against experimental results. Our calculations identified specific native defects in the α-MoO3 crystal lattice. Crucially, only X-band EPR measurements were experimentally performed and the high-frequency (W- and J-band) EPR spectra were generated through simulations by using DFT/MD-derived EPR parameters, which allowed an assessment of the potential spectroscopic signatures of different Mo isotopes in α-MoO3 and suggest that high-frequency EPR could, in principle, serve as a tool for isotopic analysis.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-37195-6.

Keywords: Molybdenum, Isotopes, Enrichment, EPR, DFT

Subject terms: Chemistry, Materials science, Physics

Introduction

Molybdenum (Mo) is a transition metal valued for its high melting point and strength with several naturally abundant isotopes (92Mo, 94Mo, 95Mo, 96Mo, 97Mo, 98Mo and 100Mo) that differ in nuclear properties and applications. Isotopic variations affect neutron numbers, stability, and decay; 98Mo is most abundant while 93Mo is radioactive, which enables diverse uses in nuclear technology, geochemistry, and medicine1. Mo isotopes possess biogeochemical cycles and enable imaging applications, notably 98Mo in medical imaging2 and 97Mo in neutron-reaction research3. The widely used medical isotope of 99mTc derived from 98Mo benefits from favorable decay characteristics and half-life values, supporting diagnostics for the heart, lungs, thyroid, and gastric mucosa4. Additionally, 98Mo enhances steel and alloy characteristics for high-temperature and corrosive environments5. Thus, Mo isotopes, especially 96Mo, 97Mo, and 98Mo, from which 99mTc is produced, play crucial roles in healthcare, science of materials, and nuclear research.

Although ICP-MS is widely used for isotope ratio analysis, its suitability is limited for enriched and highly valuable samples, such as molybdenum. This is due to its destructive nature, the requirement for complete sample dissolution and extensive chemical preparation, its sensitivity to potential matrix effects and plasma-related interferences, the need for mass-bias correction6–9, and the high cost of both the instrumentation and measurement procedures. Therefore, a technique that is entirely non-destructive, requires no sample dissolution or chemical preparation, and allows for sample reuse after measurement is highly desirable. Based on this need, we employed the fundamental principles of Electron Paramagnetic Resonance (EPR) to demonstrate that isotope-dependent hyperfine parameters can provide the basis for a novel approach to isotopic analysis.

EPR spectroscopy has been used as a powerful tool for various applications, including dosimetry10,11, tracking catalysts12–14, identifying defects in materials15–17, and studying proteins18 in medicine19. However, there has been no report to date on the use of spectroscopy for the analysis of atomic isotopes. Changes in the electronic energy levels of an atom resulting from isotopic variations can be applied to investigate isotopic abundance20,21. Given the dependence of the Zeeman effect on the g-factor of isotopes and the Hyperfine Coupling Constant (HFCC) for them, alterations in the energy levels of paramagnetic compounds caused by variations in the atomic isotopes can be calculated.

In the presence of a magnetic field, a unique system across the whole nuclear chart can be provided by using these isotopes so that the shift in the relativistic nuclear recoil can be assessed. The g-factor and the HFCC of isotopes are fundamental parameters in EPR spectroscopy for achieving a compound with an unpaired spin (S = 1/2). Therefore, EPR spectroscopy is a suitable and sensitive method for investigating the electronic structure and splitting of degenerate magnetic energy levels. When materials with unpaired electrons are placed in a magnetic field, the degeneracy of magnetic sub-levels is eliminated. To achieve better resolution of the degenerate sub-levels, it is necessary to record the EPR spectrum at frequencies higher than the X-band frequency (9.778 GHz), such as the W-band (94 GHz) and J-band (275 GHz) frequencies. Although High-Frequency (HF) instrumentation was not available during this study, the discussion of W- and J-band EPR is presented as a proposed and simulated methodology based on validated X-band parameters. The last two decades have seen significant progress in the development of EPR spectrometers operating above 200 GHz22.

In this study, the electromagnetic separation system enriched 96Mo, 97Mo, and 98Mo isotopes. The powders collected from the packets of 96Mo, 97Mo, and 98Mo isotopes were purified by using a chemical refining process. The refined powders were then characterized using X-ray Diffraction (XRD), Scanning Electron Microscopy (SEM), Inductively Coupled Plasma (ICP), Inductively Coupled Plasma Mass spectrometry (ICP-MS), Photoluminescence (PL), and EPR. The electronic, optical, and EPR properties were investigated using ab initio calculations. The band gap, optical properties, and EPR spectra were confirmed and validated via comparison with the existing experimental results. Based on the performed calculations, defects in the α-MoO3 crystal were identified. Furthermore, the simulated W- and J-band EPR spectra-derived from DFT/MD results and validated X-band parameters-were used for exploring the potential isotopic resolution achievable at HFs rather than claiming experimental HF-EPR measurements.

This paper was divided into 3 sections: The section of results included the material structure, findings from XRD, SEM, ICP, ICP-MS and PL spectrometry, and X-band EPR for the powders collected from the packets of 96Mo, 97Mo, and 98Mo isotopes. In the section of methods, Density Functional Theory (DFT) calculations were used to investigate the electronic properties, stability of defects, and EPR spectra for the mentioned powders. Additionally, HF-EPR data shown in this work originated from simulations designed to evaluate the feasibility of HF isotopic discrimination. The paper concluded by linking the main insights and their significance.

Results and discussion

Discussion on the structural and microstructural characterization can be found in Section S2 of the Supporting Information.

Experimental PL and EPR spectra

PL spectroscopy is a sensitive, non-destructive measurement technique for studying defects in materials. It provides important information about the energy levels of these defects, even at low densities. The room-temperature PL spectra of the purified powders from the packets of 96Mo, 97Mo, and 98Mo isotopes by using a 325-nm light source are displayed on the left side of Fig. 1. The experimental PL spectra consist of an emission peak at in the energy ranges of 3.14–3.34 eV corresponding to emission near the band edge of pristine α-MoO3 and emission peaks in the energy ranges of 2.80–2.98 eV corresponding to native defects in α-MoO3.

Fig. 1.

Fig. 1

Open in a new tab

Deconvolution of PL (left) and EPR (right) spectroscopy of the purified powder from the packets of 96Mo, 97Mo, and 98Mo isotopes.

EPR spectroscopy is also a non-destructive and suitable technique for identifying defects and analyzing atomic isotopes in materials with unpaired electrons. For defects where the emitted photon energies are identical in the experimental PL spectra and for isotopic variations that cause fine splitting in energy levels, these changes are not observable in the emitted photon energy. Therefore, accurately identifying defects based solely on the emitted photon energy from PL spectra is not possible. Furthermore, information about the isotopes in the purified powders cannot be extracted from this spectrum. In such cases, it is necessary to investigate finer interactions, such as the splitting of energy levels in the presence of perturbation like a magnetic field. The energy splitting arises from the interaction of the magnetic dipole moment of the electron spin with an external magnetic field (the Zeeman effect) and, for isotopes with non-zero nuclear spin, includes the interaction of the magnetic dipole moment of the electron spin with the nuclear spin (hyperfine interaction). The EPR spectrum for a system with one unpaired spin (S = 1/2) includes these two interactions. The EPR spectra for the purified powders from the packets of 96Mo, 97Mo, and 98Mo isotopes are portrayed on the right side of Fig. 1. As can be seen, the resonance fields for the purified powders from the mentioned packets within the swept field range of 330 to 360 mT contain 1, 2, and 1 peak(s), respectively. Changes in the number, position, and intensity of the resonance field peaks for the purified powders from these packets are due to changes in the EPR parameters, namely the g-factor and the HFCC, resulting from the defects and isotopes. Additional explanations regarding the EPR energy-level splitting can be found in Section S3 of the Supporting Information.

Electronic band structure and defect formation energy

α-MoO3 has an orthorhombic crystal structure in the Pbnm space group. Each bilayer consists of two sub-layers of distorted, corner-linked MoO6 octahedra along the [100] direction with edge-sharing along the [001] direction. The interactions between the layers are Coulombic and van der Waals interactions. There are 3 types of oxygen atoms in the α-MoO3 crystal. O1 atoms are positioned symmetrically between two Mo atoms with a distance of approximately 1.93 Å and another Mo atom in a different layer at a distance of about 2.33 Å. O2 atoms are positioned asymmetrically between two Mo atoms with distances of 1.75 and 2.13 Å. O3 atoms are bonded to one molybdenum atom with a symmetric bond length of 1.67 Å. The calculated bond lengths show good agreement with experimental data23, indicating the accuracy and validity of the model.

The electronic band structure is one of the most fundamental and important characteristics of a crystalline solid. A wide range of material properties can be investigated based on this structure. Therefore, the accurate prediction of electronic band structures is a fundamental issue in computational condensed matter physics. The band structure of pristine α-MoO3, along with the total and partial density of states for the 2p orbitals of oxygen atoms and the 4d orbitals of molybdenum atoms for the 2 × 1 × 2 and 4 × 1 × 2 supercells, are presented in Fig. 2. The direct optical band gaps at the Г point of the Brillouin zone are calculated to be 3.26 and 3.12 eV for the 2 × 1 × 2 and 4 × 1 × 2 supercells, respectively. These values show a difference of approximately 0.14 eV (corresponding to ~ 4% of the band gap) and are in good agreement with the experimental value (3.21 eV)24. As is evident from the partial density of states, the VBM and CBM are primarily composed of the 2p orbitals of oxygen atoms and the 4d orbitals of molybdenum atoms, respectively. Furthermore, the electron distribution is located mostly around the oxygen atoms (Fig. 2).

Fig. 2.

Fig. 2

Open in a new tab

The band structure, total and partial density of states (left), and electron density distribution (0.40 a.u.) (right) for the 2 × 1 × 2 and 4 × 1 × 2 supercells of α-MoO3.

To assess the thermodynamic stability of defects after chemical purification, the formation energy for all the mentioned defects were calculated. The refining conditions could influence the formation energy and the presence of various defects during the preparation process of α-MoO3 powders. Given that the purification of α-MoO3 was carried out under oxygen-rich conditions, the formation energy for all defects were calculated under oxygen-rich condition. Additional computational details regarding the supercell-size test are given in Section S4 of the Supporting Information. The formation energy for all defects were calculated under oxygen-rich condition as presented in Fig. 3. It is evident from this figure that under oxygen-rich condition, the formation of the Oi1, Oi2, Oi3, VMo, Oi1VMo and Oi2VMo defects are more probable. The formation energies of the VO1, VO2, and VO3 defects under oxygen-rich conditions for neutral supercells were obtained as 3.41, 3.58, and 2.87 eV, respectively. This revealed excellent agreement with the values reported by Tahini et al., who provided formation energies in the range of 2.06–3.93 eV by using DFT calculations25, while being smaller than the values obtained by Inzani et al., who reported formation energies in the range of 6.3 to 7.4 eV using the same calculations26. To date, there are no reports on the calculation of formation energies for other defects.

Fig. 3.

Fig. 3

Open in a new tab

Calculated formation energies for all native defects (top) and formation energies of the most dominant defects in + 1 and neutral charged states as a function of the Fermi energy (bottom) under oxygen-rich condition.

By considering the + 1-charge state for the dominant defects and their dependence on the Fermi level within the band gap, we identified the stability regions of the defects with + 1 and neutral charges. Figure 3 shows that for the Oi1, Oi3, VMo and Oi1VMo defects, the + 1-charge state is dominant when the Fermi energy is within a wide range from 0 to 1.64, 1.73 eV, 1.09 and 0.89, respectively. Therefore, for Oi1 and Oi3 defects, the + 1-charge state is the stable state compared those of VMo and Oi1VMo defects. The Oi1 and Oi3 defects exist exclusively in the + 1-charge state with a negative formation energy, indicating that these defects form spontaneously and donate an electron to the lattice.

Defects are responsible for photon emission in the visible region and the magnetic behavior in the α-MoO3 structure. The emitted photons in the PL spectrum in Fig. 1 originate from native defects in this structure. Native defects create energy levels, leading to the emission of photons at different energies. Supercell-size effects are discussed in Section S5 of the Supporting Information.

The band structure for the dominant defects in the α-MoO3 structure with a + 1-charge, along with their spin density distributions, is shown in Fig. 4, and with a neutral charge, along with the electron density distribution, is depicted in Fig. S7. As can be observed, the emitted photon energies originate from the dominant defects in the + 1-charge state. The band structure for other defects in the mentioned structure with a + 1-charge is displayed in Fig. S8. In Fig. 5, the emitted photon energies from defects in α-MoO3 are compared with those of the experimental PL spectra of the purified powders from the packets of 96Mo, 97Mo, and 98Mo isotopes (Fig. 1). The experimental PL spectra for all cases show emitted photon energies in the energy ranges of 3.14–3.34 eV (blue circle), 2.98–3.05 eV (brown triangle), and 2.85–2.92 eV (green triangle). The purple cross corresponds to a very deep central level, which is not present in that spectrum. Comparing the emitted photon energies from the experimental PL spectra with the calculated photon energies from the band structures for α-MoO3 indicates that the emitted photon energy in the energy range of 3.14–3.34 eV (blue circle) originates from the pristine α-MoO3 structure and the emitted photon energies of 2.85 and 3.05 eV originate from the dominant defects (Oi1, Oi2, Oi3, VMo, and Oi1VMo) with a + 1-charge. Since the photon emission energies for the four defects are nearly identical, it is not possible to determine precisely which specific defect is present in the purified powders from the packets of 96Mo, 97Mo, and 98Mo isotopes of α-MoO3. Therefore, to accurately specify the defects and isotopes in the purified powders, a perturbation potential is required to split and lift the degeneracy of the energy levels in the band structure.

Fig. 4.

Fig. 4

Open in a new tab

The band structure for the dominant defects in the α-MoO3 structure with a + 1-charge, along with their spin density distributions (0.25 a.u.).

Fig. 5.

Fig. 5

Open in a new tab

Comparison of the energy of photons emitted from defects in α-MoO3 with the emitted photon energies from the experimental PL spectra of the purified powders from the packets of 96Mo, 97Mo, and 98Mo isotopes.

Simulation of EPR spectra and analysis of molybdenum isotopes

The primary physical contributions to the g-factor for α-MoO3 were obtained using ab initio calculations. The theoretical values of the isotopic shift in the g-factor for isotopes were derived mainly from the sum of the nuclear recoil and nuclear size contributions as reported below21:

graphic file with name d33e646.gif 1

where n represents the principal quantum number of the valence electron. Other contributions, such as nuclear deformation27 and nuclear polarization28, are very small and have a negligible effect on the difference in the g-factor of the isotopes. For molybdenum with the atomic number of 42, the shift in the g-factor originates essentially from the mass variation of the isotopes, which has a relativistic origin. The nuclear recoil and size contributions for Mo isotopes are presented in Table S3. Mo isotopes have a mass difference of approximately 2% and almost the same nuclear charge radii29. Therefore, more than 99.98% of the difference in the g-factor comes from the nuclear recoil shift. To obtain the total value of the g-factor variation for Mo isotopes, the nuclear size effect was also taken into account.

Defects with neutral charge state in the α-MoO3 structure were diamagnetic due to the absence of unpaired electrons. Consequently, their EPR spectra were was inactive. Given the existence of an experimental EPR spectrum for the purified powders (Fig. 1) similar to the defect formation energy (Fig. 3) and emitted photon energies (Fig. 4), the + 1-charge state was the dominant state for the defects.

In the absence of raw experimental data, we relied on comparison of reported resonance field positions in the EPR spectrum to accurately identify the defects in the purified powders. The resonance field position in the simulated EPR spectrum was obtained by calculating EPR parameters, such as the g-factor (Zeeman effect) and the HFCC (A in MHz) by using RelationInline graphic, where ν is the microwave frequency in GHz, ml is the nuclear spin quantum number, βe is the Bohr magneton, and B0 is the magnetic field in mT. However, although oxygen-17 isotopes had a nuclear spin of 5/2, their natural abundance was negligible (0.038%), thus contributing insignificantly to the EPR spectrum and resonance field position. 95Mo and 97Mo isotopes had a nuclear spin of 5/2 with natural abundances of 15.90 and 9.58% respectively. Due to isotopic separation, the presence or absence of these isotopes significantly affected the energy level splitting and the EPR resonance field due to their HFCCs. The HFCC for each atom is proportional to the spin density on that atom30. As evident from Fig. 4, the molybdenum atoms possess a HFCC given the presence of unpaired electron spin density on Mo atom for all dominant defects. For the 95Mo and 97Mo isotopes, this results in 6 resonance fields (Fig. S4); however, due to the limited sweep range of our spectrometer (338–358 mT), only two of these peaks were shown in Fig. 1. 96Mo, 98Mo and 100Mo isotopes had zero nuclear spin; therefore, their HFCC did not affect the EPR spectrum, resulting in a single EPR resonance field determined solely by the g-factor value of the isotope (Fig. S4). The EPR parameters calculated for all defects in the static state, including the combined nuclear recoil and size contributions for Mo isotopes of 96 to 100, are reported in Table S4. Using the calculated parameters for Mo isotopes of 96 to 100, the X-band EPR spectra for all defects were simulated as shown in Fig. S9. In Fig. 6, the resonance field positions of the simulated EPR spectra are compared with the experimental spectra for all defects. From this figure, it is evident that the Oi1VMo, VMo, and Oi1/Oi3 defects are present in the powder collected from the 96Mo, 97Mo, and 98Mo isotope packets, respectively. The peaks at the resonance fields of 353.464 and 353.816 mT in the powder collected from the 98Mo isotope packet correspond to the Oi3 and Oi1 defects, respectively. The peak intensity for the Oi3 defect is greater than that for the Oi1 defect, indicating a higher concentration of the Oi3 defect. Furthermore, their formation energies show that the Oi3 defect has a lower formation energy than the Oi1 defect. The integrated EPR intensity can be used as a measure of their relative concentrations31. Variations in the integrated intensity primarily reflect changes in defect concentration. We clearly state that our work did not attempt to extract quantitative concentrations, because this required experimental measurements of relaxation processes and broadening mechanisms at the relevant HFs. Converting these intensities into absolute concentrations requires the use of a calibration sample with a known spin concentration. Also, DFT results showed no isotope effect on defect formation energies. Differences in dominant defect species among 96Mo-, 97Mo-, and 98Mo-enriched samples arose from variations in chemical potential of local oxygen, redox microconditions, and thermal history, causing independently processed batches to relax into distinct metastable configurations.

Fig. 6.

Fig. 6

Open in a new tab

Comparison of resonance field positions of the simulated EPR spectra with the experimental spectra for all defects.

The discrepancy between the calculated HFCC for 97Mo and the experimental results led to a difference between the simulated resonance field for the VMo defect and the experimental resonance field for the powder collected from the 97Mo isotope packet. This discrepancy confirmed that static calculations were insufficient for accurately determining the EPR parameters. To account for temperature effects and precisely specify the EPR parameters for the dominant defects, Molecular Dynamics (MD) calculations were performed. Figure 7 plots the average values of the HFCC for 97Mo and the giso factor for the VMo defect as a function of the number of snapshots. Evidently, the HFCC values converge well using 150 snapshots. The effect of temperature on the giso values is negligible. In Fig. 7, the average values of the HFCC and giso for the dominant defects respectively obtained from 200 to 150 snapshots of the MD calculations and the optimized structures are compared.

Fig. 7.

Fig. 7

Open in a new tab

The average values of the HFCC for 97Mo and the giso factor: (top) for the VMo defect as a function of the number of snapshots and (bottom) for the dominant defects of the MD calculations and the optimized structures.

The calculated EPR parameters from MD calculation, including the combined nuclear recoil and size contributions for 95Mo, 96Mo, 97Mo, 98Mo and 100Mo isotopes of the dominant defects are reported in Table S5. Using the calculated EPR parameters for 95Mo, 96Mo, 97Mo,98Mo and 100Mo isotopes, the X-band EPR spectra for 95Mo, 96Mo, 97Mo, 98Mo and 100Mo isotopes and natural molybdenum of the dominant defects were simulated, which are presented in Figs. S10. In Fig. 8, the resonance fields of 95Mo, 96Mo, 97Mo,98Mo, and 100Mo isotopes and natural molybdenum of the dominant defects are compared between the simulated and experimental spectra. As is clear from this figure, for natural molybdenum, there is a significant difference between the simulated and experimental resonance fields, confirming that isotopic separation of 95Mo, 96Mo, 97Mo, 98Mo, and 100Mo isotopes has been performed using the electromagnetic system. Furthermore, the percentages of 95Mo and 97Mo isotopes for the Oi1, Oi3, and Oi1VMo defects (in powders purified from the 96Mo and 98Mo isotope packets) are very negligible (less than 5%) similar to those of 96Mo, 98Mo, and 100Mo isotopes for the VMo defect (in powders purified from the 97Mo isotope packet).

Fig. 8.

Fig. 8

Open in a new tab

Comparison of X-band field position of the simulated EPR spectra with those of the experimental spectra for 95Mo, 96Mo, 97Mo, 98Mo, and 100Mo isotopes and natural molybdenum of the dominant defects.

Due to the overlap of EPR resonance fields and the very small differences in resonance fields between 96Mo, 98Mo and 100Mo isotopes and 95Mo and 97Mo isotopes in the X-band EPR spectrum, it was very difficult to analyze the ratios of 95Mo and 97Mo isotopes in the 97Mo isotope packet and those of 96Mo, 98Mo and 100Mo isotopes in the 96Mo and 98Mo packets in this band. Although our experiments were limited to X-band due to restricted access to HF-EPR spectrometers, fundamental parameters were accurately determined through combined X-band measurements and DFT/MD calculations. Since these parameters were intrinsic to the defect–isotope system and independent of microwave frequency, HFs could only enhance Zeeman splitting and isotopic resolution.

To achieve better isotopic separation and improved resolution, simulations were performed at HFs by using W- and J-band EPR spectroscopy. Using the calculated EPR parameters for 95Mo, 96Mo, 97Mo, 98Mo, and 100Mo isotopes (Table S5), the W-band spectra of these isotopes and natural molybdenum defects were simulated (Fig. S11). Also, the absorption signals were simulated by using the validated EPR parameters obtained from X-band experiments and DFT calculations for Oi1 and VMo defects (Fig. 9). A linewidth of ~ 0.2 mT was applied to clearly illustrate the isotopic splitting. The simulations showed that full resolution of Mo isotopes 95–100 was achieved only at higher microwave frequencies (W- and J-bands). These simulations reproduced line positions but not the relaxation-driven line shapes or absolute intensities expected in real HF experiments. Thus, they demonstrated isotopic spectral separation rather than quantitative HF-EPR. Achieving quantitative analysis requires direct measurement of relaxation processes (T₁, T₂) and broadening mechanisms, which are frequency-dependent and cannot be reliably predicted by simulation alone. Our simulations showed that HF-EPR can resolve all Mo isotopes and allow estimation of their relative abundances from the separated line intensities; however, quantitative isotopic analysis still required an appropriate calibration sample.

Fig. 9.

Fig. 9

Open in a new tab

Simulated absorption signals for the X, W, and J-bands for the Oi1 (top) and VMo (bottom) defects.

In X-band measurements, 96Mo, 98Mo, and 100Mo isotopes could be analyzed relative to 95Mo and 97Mo isotopes based on their relative peak intensities. However, only graphical images of the experimental X-band EPR spectra were available and the absence of raw digital data suitable for quantitative analysis prevented precise evaluation of 95Mo and 97Mo isotopes within the packet labeled for 96Mo and 98Mo isotopes, as well as accurate assessment of 96Mo, 98Mo, and 100Mo isotopes within the packet labeled for 97Mo isotope. Although incomplete spectral separation in the X-band limited quantitative isotopic analysis—particularly due to overlapping resonance fields among 96Mo, 98Mo, and 100Mo isotopes and the lack of complete spectral data—the observed cross-contributions remained low (< 5%) for 95Mo and 97Mo isotopes within the packet labeled for 96Mo and 98Mo isotopes and for 96Mo, 98Mo, and 100Mo isotopes within the packet labeled for 97Mo isotope which was in agreement with ICP-MS benchmarks. At HFs (W/J bands), full isotopic resolution can be achieved, enabling reliable quantitative analysis through peak-intensity ratios. If modern HF-EPR output-such as that generated by instruments at the Ioffe Institute22 in Russia-were available, it would be possible to perform quantitative isotopic analysis of Mo isotopes and heavier atoms such as lanthanides (e.g., ytterbium and gadolinium) by comparing resonance-field intensity ratios. Thus, although X-band data alone cannot fully separate 96Mo, 98Mo, and 100Mo isotopes from 95Mo and 97Mo isotopes, the combination of X-band measurements and validated simulations highlights the potential of HF-EPR for Mo isotopic analysis. These results, based on simulations, indicate that complete isotopic resolution would be experimentally attainable if W- or J-band EPR instrumentation were accessible.

Online content

Comprehensive theoretical background, detailed experimental and computational methods, extended characterizations, additional analyses, and additional resources are provided in the Supporting Information.

Conclusion

This study introduced a novel EPR-based methodology with potential applicability for isotopic analysis of molybdenum by integrating X-band measurements with DFT/MD calculations to determine the fundamental EPR parameters of Mo isotopes. In the α-MoO₃ structure, the combination of PL experiments, X-band EPR spectroscopy, and DFT + U computations of defect energies and resonance fields enabled the identification of native defects-specifically Oi1, Oi2, Oi3, VMo, Oi1VMo, and Oi2VMo-as the most probable defect configurations under oxygen-rich conditions. The + 1-charge states of these defects were found to be responsible for visible-range photon emission and defect-specific EPR signatures were observed in the isotopically enriched samples, including Oi1VMo in 96Mo, VMo in 97Mo, and Oi1/Oi3 in 98Mo. However, quantitative analysis using X-band EPR was hindered by the significant overlap of resonance fields, which limited the ability to distinguish between isotopes such as 96Mo, 98Mo, and 100Mo. The combination of X-band measurements and validated simulations demonstrated the potential of HF-EPR as a tool for Mo isotopic analysis. Simulations indicated that at HFs (W/J-band), isotope spin states became fully separated, could enable precise isotopic analysis in principle. These findings were based on simulations and indicated that full isotopic resolution could be achievable experimentally if W- or J-band EPR instrumentation were available. Despite remaining practical limitations—such as the need for HF-EPR instrumentation, calibration standards for quantification, and potential line-broadening effects in real samples—the current framework provides a foundation for future experimental investigation of isotopic separation in Mo and similar systems.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1 (3.9MB, docx)

Author contributions

Javad Karimi-sabet and Mehdi Janbazi are corresponding authorsRazieh Hosseini and Mehdi Janbazi Performing DFT calculations.Razieh Hosseini and Masoomeh Sharbatdaran done the chemical refining process. Razieh Hosseini, Mehdi Janbazi and Masoomeh Sharbatdaran prepared figures.Javad Karimi-sabet, Mehdi Janbazi and Masoomeh Sharbatdaran performed Data analysis.Javad Karimi-sabet and Mohsen Ashjari supervised the project. All authors wrote the main manuscript text.All authors reviewed the manuscript .

Funding

No Funding.

Data availability

The data are supporting the finding of this study are available from the corresponding author upon reasonable request. Data for this article, including calculation data and experimental are available at https://drive.google.com/file/d/1OPs1dqsFA62yFhL1JJPfF2DbgKKbOUR_/view. Sincerely, Javad Karimi-sabet and Mehdi JanbaziCorresponding authors.

Declarations

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Contributor Information

Javad Karimi-sabet, Email: j_karimi@alum.sharif.edu, Email: jvkarimi@aeoi.org.ir.

Mehdi Janbazi, Email: mehdijanbazi@yahoo.com.

References

  • 1.Luo, Y. et al. Determination of 93Mo in radioactive samples of sulfuric acid media from nuclear facilities. Anal. Chem.94(33), 11582–11590 (2022). [DOI] [PubMed] [Google Scholar]
  • 2.Badwar, S. et al. Measurement of formation cross-section of 99Mo from the 98Mo (n, γ) and 100Mo (n, 2n) reactions. Appl. Radiat. Isot.129, 117–123 (2017). [DOI] [PubMed] [Google Scholar]
  • 3.Krane, K. S. Neutron capture by 94, 96Zr and the decays of 97Zr and 97Nb. Appl. Radiat. Isot.94, 60–66 (2014). [DOI] [PubMed] [Google Scholar]
  • 4.Hasan, S. & Prelas, M. A. Molybdenum-99 production pathways and the sorbents for 99Mo/99mTc generator systems using (n, γ) 99Mo: a review. SN Appl. Sci.2(11), 1782 (2020). [Google Scholar]
  • 5.Liu, W. et al. Impact of Mo/Ni alloying on microstructural modulation and low-temperature toughness of high-strength low-alloy steel. J. Iron. Steel Res. Int.31(7), 1746–1762 (2024). [Google Scholar]
  • 6.Ma, X. et al. A rapid method for detecting trace Rb and Cs in high-salinity Brines by ICP-MS equipped with an all-matrix sampling device. Sci. Rep.15(1), 30457 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Van Acker, T. et al. Inductively coupled plasma mass spectrometry. Nat. Reviews Methods Primers. 3(1), 52 (2023). [Google Scholar]
  • 8.Temerdashev, Z. et al. Possibilities and limitations of ICP-Spectrometric determination of the total content of Tin and its inorganic and organic speciations in waters with different salinity Levels—Part 2: separate determination of inorganic and organic speciations of Tin. Molecules28(18), 6615 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Haslova, K. Sensitive and spectral Interference-Free determination of rhodium by photochemical vapor generation inductively coupled plasma mass spectrometry. Anal. Chem.97(6), 3545–3553 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Guo, J. et al. The design of X-band EPR cavity with narrow detection aperture for in vivo fingernail dosimetry after accidental exposure to ionizing radiation. Sci. Rep.11(1), 2883 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Lima, I. S., Eder, J., Guidelli & Baffa, O. Dose enhancement factor caused by gold nanoparticles: influence of the dosimetric sensitivity and radiation dose assessed by electron spin resonance dosimetry. Phys. Med. Biol.66(21), 215013 (2021). [DOI] [PubMed] [Google Scholar]
  • 12.Seif-Eddine, M. et al. Operando film-electrochemical EPR spectroscopy tracks radical intermediates in surface-immobilized catalysts. Nat. Chem.16(6), 1015–1023 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Bonke, S. A. et al. In situ electron paramagnetic resonance spectroscopy for catalysis. Nat. Reviews Methods Primers. 1(1), 33 (2021). [Google Scholar]
  • 14.Rostas, A. et al. Long-lived hydrated FMN radicals: EPR characterization and implications for catalytic variability in flavoproteins. J. Am. Chem. Soc.140(48), 16521–16527 (2018). [DOI] [PubMed] [Google Scholar]
  • 15.Chejanovsky, N. et al. Single-spin resonance in a Van der Waals embedded paramagnetic defect. Nat. Mater.20(8), 1079–1084 (2021). [DOI] [PubMed] [Google Scholar]
  • 16.Gottscholl, A. et al. Initialization and read-out of intrinsic spin defects in a Van der Waals crystal at room temperature. Nat. Mater.19(5), 540–545 (2020). [DOI] [PubMed] [Google Scholar]
  • 17.Foo, C. et al. Characterization of oxygen defects and nitrogen impurities in TiO2 photocatalysts using variable-temperature X-ray powder diffraction. Nat. Commun.12(1), 661 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Emmanouilidis, L. et al. NMR and EPR reveal a compaction of the RNA-binding protein FUS upon droplet formation. Nat. Chem. Biol.17(5), 608–614 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.de Lázaro Irene, and David J. Mooney. A nanoparticle’s pathway into tumours. Nat. Mater.19(5), 486–487 (2020). [DOI] [PubMed] [Google Scholar]
  • 20.Pálffy, A. Nuclear effects in atomic transitions. Contemp. Phys.51(6), 471–496 (2010). [Google Scholar]
  • 21.Köhler, F. et al. Isotope dependence of the Zeeman effect in lithium-like calcium. Nat. Commun.7(1), 10246 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Edinach, E. V. et al. Electronic structure of non-Kramers Tb 3 + centers in Garnet crystals and evidence of their energy and spin transfer to Ce 3 + emitters. Phys. Rev. B. 100(10), 104435 (2019). [Google Scholar]
  • 23.Atuchin, V. V. et al. Sublimation growth and vibrational micro spectrometry of α-MoO3 single crystals. J. Cryst. Growth.318(1), 987–990 (2011). [Google Scholar]
  • 24.Chen, J. et al. Fast synthesis of α-MoO3 nanorods with controlled aspect ratios and their enhanced lithium storage capabilities. J. Phys. Chem. C. 114(18), 8675–8678 (2010). [Google Scholar]
  • 25.Tahini, H. A. et al. Mobile polaronic States in α-MoO3: an Ab initio investigation of the role of oxygen vacancies and alkali ions. ACS Appl. Mater. Interfaces. 8(17), 10911–10917 (2016). [DOI] [PubMed] [Google Scholar]
  • 26.Inzani, K. et al. A Van der Waals density functional study of MoO3 and its oxygen vacancies. J. Phys. Chem. C. 120(16), 8959–8968 (2016). [Google Scholar]
  • 27.Zatorski, J. et al. Nuclear shape effect on the g factor of hydrogen like ions. Phys. Rev. Lett.108(6), 063005 (2012). [DOI] [PubMed] [Google Scholar]
  • 28.Volotka, A. V. & Günter, P. Nuclear polarization study: new frontiers for tests of QED in heavy highly charged ions. Phys. Rev. Lett.113(2), 023002 (2014). [DOI] [PubMed] [Google Scholar]
  • 29.Charlwood, F. C. et al. Nuclear charge radii of molybdenum fission fragments. Phys. Lett. B. 674(1), 23–27 (2009). [Google Scholar]
  • 30.Janbazi, M. Radiation-Induced N-Methylacetamide and β-Alanine radicals in aqueous solutions at different pH levels: EPR spectroscopy and First-Principles calculations. J. Mol. Liq.423, 127035 (2025). [Google Scholar]
  • 31.Zvanut, M. E. et al. Photo-induced electron paramagnetic resonance: A means to identify defects and the defect level throughout the bandgap of ultrawide bandgap semiconductors. Appl. Phys. Lett.124, 4 (2024).

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1 (3.9MB, docx)

Data Availability Statement

The data are supporting the finding of this study are available from the corresponding author upon reasonable request. Data for this article, including calculation data and experimental are available at https://drive.google.com/file/d/1OPs1dqsFA62yFhL1JJPfF2DbgKKbOUR_/view. Sincerely, Javad Karimi-sabet and Mehdi JanbaziCorresponding authors.

Articles from Scientific Reports are provided here courtesy of Nature Publishing Group

RESOURCES