Multiscale investigation of radiation-induced structural evolution and stability of Gd₂O₃ and Er₂O₃ radiosensitizer nanoparticles under radiotherapy-relevant doses

Abstract

High-Z nanoparticles (NPs) are promising radiosensitizers, yet little is known about how clinical radiation doses modify their crystal lattices and potentially impact performance and safety. Here, we investigate radiation-induced structural evolution in Gd₂O₃ and Er₂O₃ NPs radiosensitizers using an integrated multiscale approach that combines X-ray diffraction (XRD), and time-differential perturbed angular correlation (PAC) spectroscopy. The radiosensitizing properties of the NPs were evaluated by dispersing them in Fricke gel. The NPs exhibit significant dose-enhancement factors under ⁶⁰Co γ-radiation and six megavolts clinical photon beams, comparable to or exceeding their bulk counterparts. XRD reveals subtle but systematic lattice-parameter shifts with increasing dose, while PAC provides unprecedented atomic-scale sensitivity to local rearrangements of oxygen vacancies and defect clusters. Importantly, no evidence of amorphization, phase decomposition, or irradiation-induced structural collapse was observed up to 20 Gy, indicating a high degree of radiation resilience of both nanoparticle systems. The asymmetric (low-symmetry) rare-earth sites are found to be markedly more susceptible to radiation-induced distortions than the symmetric sites, establishing a structure-response relationship unique to the bixbyite lattice. These findings demonstrate that Gd₂O₃ and Er₂O₃ nanoparticles maintain their bixbyite crystal structure under clinically relevant irradiation doses while also experiencing subtle rearrangements of defects at specific sites. The methodology introduced provides a powerful experimental framework for investigating radiation-induced defect evolution and structural stability in nanomaterials exposed to high-energy environments. It also offers structural design insights for the future development of radiation-resilient rare-earth oxide nanoparticles as high-Z nanomaterials proposed for dose-enhancement applications.

High-Z nanoparticles (NPs) are promising radiosensitizers, yet little is known about how clinical radiation doses modify their crystal lattices and potentially impact performance and safety.

1. Introduction

Cancer is a major threat to global health, and remains one of the leading causes of significant morbidity and mortality worldwide.¹ Radiation therapy is one of the most effective tools for cancer treatment, providing survival and palliative benefits.^2–4 Radiation therapy involves the delivery of high-intensity ionizing radiation with great precision into the tumor, leading to its ablation. When well controlled, radiation dose and beam precision can maximize therapeutic benefits and protect healthy tissues.⁵ The interaction between ionizing radiation and cells produces electrons and free radicals, highly reactive atoms or molecules with unpaired electrons, which propagate through the surrounding space and deliver energy. This interaction process is what makes radiation therapy one of the main procedures for cancer treatment.^6,7

To increase the efficiency of radiation therapy, nanoparticle (NP)-based radiosensitizers have attracted increasing attention.^8–13 When exposed to ionizing radiation, radiosensitizers can locally enhance the power of radiation therapy without increasing the radiation effects on the surrounding healthy tissues. They can be classified as physical, chemical, or biological. Physical radiosensitizers are those in which radiation dose enhancement is achieved by introducing materials, typically nanoparticles, containing high-atomic-number (high-Z) elements. When radiation interacts with those NPs the photoelectric absorption is intensified, generating both secondary electrons, which will deliver a high amount of energy in a short distance, and reactive oxygen species, leading to extensive DNA damage.^12,14 In addition, NPs can take advantage of the leaky blood vessels found in tumors, allowing them to accumulate more easily at the tumor site.¹⁵ Their small size helps them penetrate tumor tissue better than larger molecules, increasing the probability that ejected electrons deposit their energy within the tumor, thereby reducing damage to healthy tissues.^16,17

Application of high-Z based NPs for dose enhancement in radiation therapy has been extensively investigated in the recent years. For example, porous platinum (Z = 78) NPs showed great potential as radiosensitizers by combining their high-Z with oxygen generation capability.¹⁸ Also, hafnium (Z = 72)-based NPs have been used in clinical trials to evaluate their safety and efficacy as dose enhancers and showed promising results.^19,20 Gold (Z = 79) NPs (AuNPs) are very well-known among the scientific community, and have been widely studied due to their radiosensitizer properties and biocompatibility.^8,21–24In vivo studies using AuNPs have shown radiation dose enhancement, resulting in significant tumor size reduction.²⁵ However, the drawback of using AuNPs is that, although they can be detected in the human body using X-ray,²⁶ materials that could be detected by other means, e.g., magnetic resonance imaging (MRI), are preferable to reduce unnecessary exposure to ionizing radiation.

Therefore, rare-earth (RE)-based NPs (RENPs) have emerged as promising candidates for radiation dose enhancement due to their high atomic number, excellent chemical stability²⁷ magnetic behavior, which allows their localization within the patient body using MRI. Notably, clinical trials have demonstrated the dual-function capability of gadolinium (Z = 64)-based NPs (GdNPs) as an MRI contrast agents and radiosensitizers in radiation therapy procedures.^28–30 Similarly, erbium (Z = 68), which has one of the highest magnetic moments among the rare-earth elements (9.58 µ_B), is a promising candidate for the same applications. Also, recent studies have been performed to better understand Er³⁺ cytotoxicity, bioactivity and potential biomedical applications and radiosensitizer properties,^31–33 thus paving the way for its prospective use in medicine.

However, even though some NPs are already being used in clinical trials, several questions and challenges remain. These include scaling up their NPs synthesis, improving strategies for their safe and effective delivery into the human body, and performing comprehensive nanotoxicity studies.³⁴ Furthermore, in some radiation therapy procedures, NPs are exposed to radiation doses in the range of tens of Grays.³⁵ Such high doses can damage the particles themselves, potentially leading to the release of RE ions into the biological environment, resulting in the accumulation of these elements in healthy tissues.^36–38 Nevertheless, most studies have focused primarily on radiation-induced cellular damage, rather than on the structural integrity of NPs under irradiation.^17,34,39 Therefore, in this study, we introduce a novel methodology, simple yet accurate and reliable, to evaluate the suitability of Gd₂O₃- and Er₂O₃- NPs providing valuable insight into their radiation-induced structural stability and defect evolution under radiotherapy-irradiation doses. Fricke gel dosimetry is employed as a physicochemical screening approach to evaluate dose enhancement in a defined volume, providing a quantitative basis for future cell-based and in vivo radiosensitization studies. The radiosensitizing efficacy and biological performance of gadolinium- and erbium-based nanoparticles have been extensively addressed in the literature, including clinical evaluation of gadolinium-based systems.^28–30 The present work is complementary to those studies, focusing on the fundamental structural response of these materials to ionizing radiation.

To evaluate functional response under ionizing radiation, Gd₂O₃-NPs and Er₂O₃-NPs were dispersed in Fricke gel, a well-established chemical dosimeter whose response is based on the dose-dependent oxidation of ferrous (Fe²⁺) ions to ferric (Fe³⁺) ions.⁴⁰ They were then exposed to doses of 3 Gy, 5 Gy and 7 Gy in two different systems: an MDS Nordion ⁶⁰Co-Gammacell 220 (MDS Nordion Inc, Ottawa, ON, Canada) and a 6 MV† photon beam generated by a Varian 6EX Linear Accelerator (Varian Medical Systems, Palo Alto, CA, USA). For comparison, the same methodology was applied to commercially available bulk Gd₂O₃ and Er₂O₃. Spectrophotometric measurements provided quantitative determination of Fe³⁺ concentration, enabling assessment of dose-enhancement behavior associated with the nanoparticle systems.^41–43 Furthermore, the relatively low cost and operational simplicity of spectrophotometers make them a practical alternative in developing countries, where access to MRI equipment in hospitals is often limited.

Despite the increasing number of studies demonstrating the radiosensitization capabilities of high-Z nanoparticles, there has been a lack of focus on understanding how therapeutic radiation doses affect the nanoparticles themselves. Radiation-induced defect formation, vacancy migration, local lattice distortions, and possible degradation of crystalline order may alter the physicochemical properties of the nanomaterials during treatment. Such effects are particularly relevant for rare-earth oxide nanoparticles, as changes in structural stability could impact their long-term effectiveness and safety. Therefore, it is crucial to directly investigate the structural evolution induced by radiation in order to assess the reliability of nanoparticle-based radiosensitizers.

Radiation-induced structural evolution of the NPs lattices was then investigated using a complementary multiscale characterization framework combining X-ray diffraction (XRD) and time-differential perturbed angular correlation (PAC) spectroscopy. XRD provides insight into global crystallographic stability, while PAC offers atomic-scale sensitivity to local symmetry changes and defect reorganization.⁴⁴ Correlating these techniques enables the detection of site-specific structural modifications that are not accessible through conventional structural probes alone. This integrated strategy provides a generalizable platform for probing defect-driven structural dynamics in nanomaterials exposed to high-energy environments, with applicability extending to nanoparticles, thin films, and bulk materials.

2. Experimental

2.1. Gd₂O₃- and Er₂O₃-NPs synthesis

Gd₂O₃-NPs and Er₂O₃-NPs were obtained via thermal decomposition, which is used to produce high-quality NPs.^45,46 This method consists in adding 2 mmol of rare-earth acetate into 20 mL of diphenyl ether with 4 mmol of oleylamine, 6 mmol of oleic acid and 10 mmol of 1,2-octanediol. The solution was gradually heated to 300 °C and maintained at this temperature under continuous stirring and reflux for 6 h. The heat was turned off and the solution was allowed to return to room temperature while stirring continued. The sample was then centrifuged, and the precipitate was dried under low pressure (about 1 kPa), at room temperature, for 24 h. The dried powder was then annealed at 600 °C for 10 h in air to remove excess organic material, resulting in a well-defined XRD and PAC spectra.⁴⁶ A small amount of NPs powder was taken from each batch to perform the initial characterization, which includes transmission electron microscopy (TEM), for size and size distribution, and XRD, for crystalline structure information. Bulk samples were commercial Gd₂O₃ (99.9% purity, trace metals basis, 278513 Aldrich) and Er₂O₃ (99.9% purity, trace metals basis, 289248 Aldrich). A diagram presenting the NPs synthesis method is shown in Fig. 1.

Fig. 1. Diagram showing the NPs synthesis, including the ¹¹¹InCl₃ addition for PAC measurements on the far right.

2.2. Sample preparation for radiosensitization tests using Fricke gel

Fricke gel consists of the mixture of ammonium ferrous sulfate, sodium chloride, xylenol orange, sulfuric acid with 270 Bloom porcine gelatin and tridistillated water.⁴⁷ The interaction between incident radiation and water molecules causes the breaking of chemical bonds, producing radicals that react with Fe²⁺ ions, supplied by the ammonium ferrous sulfate, thereby oxidizing them to Fe³⁺.⁴⁸ In a solution of low pH (approximately 1) the simplified equations are:⁴⁹

• Fe²⁺ + OH → Fe³⁺ + OH⁻

• H + O₂ → HO₂

• Fe²⁺ + HO₂ → Fe³⁺ + HO₂⁻

• HO²⁻ + H⁺ → H₂O₂

• Fe²⁺ + H₂O₂ → Fe³⁺ + OH + OH⁻

More details about the chemical reactions can be found in the literature.⁵⁰ The amount of Fe³⁺ produced depends on the energy absorbed by the solution. With the addition of xylenol orange it becomes possible to indirectly determine the number of ferric ions produced in the solution through the dose-dependent color change of the dosimeter.⁴⁰ In this study, 100 mL of Fricke gel solution was prepared using:

• 100 mL tridistillated water;

• 1 mM of ammonium ferrous sulfate (Sigma-Aldrich, 99% purity);

• 1 mM of sodium chloride (Sigma-Aldrich, 99.5% purity);

• 0.1 mM of xylenol orange tetrasodium salt (Sigma-Aldrich);

• 50 mM of sulfuric acid (Sigma-Aldrich, 99.999% purity);

• 5 wt% 270 bloom porcine gelatin (relative to 100 mL);

Initially, 75% of the total volume of water (75 mL) was mixed with the porcine gelatin in a beaker and then heated to 45 °C under continuous stirring until dissolution of the porcine gelatin was complete. The heat was removed and the solution was allowed to cool to 35 °C. In a second beaker, the sulfuric acid was added to the remaining 25% of water. This results in an exothermic reaction that accelerates the Fe²⁺ oxidation; therefore, the acidic solution was allowed to cool to room temperature before it could be mixed with the other 75% of water. Ammonium ferrous sulfate hexahydrate, xylenol orange, and sodium chloride were then added, and the final solution was kept under continuous stirring for five minutes before being dispensed into polymethyl methacrylate (PMMA) cuvettes and refrigerated (at about 5 °C) to complete gel formation (Fig. 2).

Fig. 2. Scheme showing detailed Fricke gel production, irradiation and spectrophotometry readings.

For the radiosensitization tests, a large batch of Fricke gel was prepared, and then divided into 5 sub-batches, one of which served as control (no particles added), and the other 4 containing one of the following powders: Gd₂O₃-NPs, Gd₂O₃ bulk, Er₂O₃-NPs and Er₂O₃ bulk. 10⁻¹ mM of the respective powder was added to the liquid solution after the addition of ammonium ferrous sulfate hexahydrate and xylenol orange, and the mixture was stirred continuously until the onset of gel formation to minimize particle mobility. No dedicated microscopic or scattering-based characterization (e.g., dynamic light scattering, TEM of gel cross-sections, or X-ray tomography) was performed to quantify nanoscale dispersion or to detect potential local aggregation. However, the spectrophotometric readout integrates the absorbance signal over the entire cuvette volume, providing a spatially averaged measurement that is inherently less sensitive to minor local inhomogeneities. The consistency of dose-enhancement factors across replicate cuvettes within each sub-batch further supports the reproducibility of the preparation procedure. Direct dispersion characterization is identified as an important direction for future work. The cuvettes were kept in the refrigerator for 24 h. Each sub-batch was divided into 4 groups:

• Group 1: non-irradiated cuvettes;

• Group 2: irradiated with a dose of 3 Gy;

• Group 3: irradiated with a dose of 5 Gy;

• Group 4: irradiated with a dose of 7 Gy;

The first test was performed using a MDS Nordion ⁶⁰Co-Gammacell 220 with a dose rate of 873 Gy h⁻¹ ±2.7%, at the Radiation Technology Center (CTR) of the Nuclear and Energy Research Institute (IPEN/CNEN-SP), Brazil. The Gammacell 220 contains 48 double-sealed source pencils, each containing seven ⁶⁰Co sources, set in an annular formation, and the sample is placed at the center of the ring (Fig. S1-A). Doses of 3 Gy, 5 Gy and 7 Gy were applied to the powder-dispersed Fricke gel dosimeters (NPs and bulk) and the control. The second irradiation was performed on a 6 MV photon beam generated by a Varian 6EX Linear Accelerator at the Albert Einstein Hospital. Following the recommendations by the International Atomic Energy Agency (IAEA) for dosimetry in radiation therapy beams, one layer with 5 cm of solid water (below the dosimeters) and 5 cm of bolus (on top of the dosimeters) were used.⁵¹Fig. 3 shows the detailed irradiation setup.

Fig. 3. (A) Fricke gel dosimeters positioned for irradiation at the 6 MV photon beam; (B) schematic showing dosimeters positioned for irradiation: (a) control, (b) Gd₂O₃-NPs, (c) Gd₂O₃ bulk (d) Er₂O₃-NPs and (e) Er₂O₃ bulk; (C) 5 cm of solid water (below the dosimeters) and 5 cm of bolus (on top of the dosimeters).

The two irradiation modalities were selected for complementary, rather than comparative, purposes. The ⁶⁰Co Gammacell provides a well-characterized, isotropic reference field widely used in radiation physics and dosimetry, while the 6 MV clinical photon beam represents the radiation quality most commonly employed in external beam radiotherapy. Because the two sources differ in energy spectra, beam geometry, scattering conditions, and secondary particle production, each dataset was analyzed independently, and no direct quantitative comparison of absolute dose-enhancement values between the two modalities is intended.

Samples' absorbance was measured in a spectrophotometer Agilent, Cary 5000 model, using the Fe³⁺ absorption band (575 nm) as a ref. 48. Since the Fe²⁺ oxidation is a natural process, and occurs equally throughout the gel, the non-irradiated group 1 absorbance curves were subtracted from the irradiated groups 2 to 4, for the respective composite, to obtain the absorbance change caused only by irradiation. The dose enhancement factor (DEF) for each sample was determined using eqn (1):^52,53

1

where A_W is the absorbance measured in a dosimeter with NPs/bulk and A_W0 is the control absorbance at the specific radiation dose.

It should be noted that Fricke gel dosimetry was employed in the present work as a physical dosimetric tool to evaluate dose-enhancement behavior under irradiation. The objective was not to assess biological radiosensitization, cellular response, or therapeutic efficacy, which require dedicated biological studies beyond the scope of this investigation.

2.3. Structural characterization using PAC spectroscopy

PAC spectroscopy is based on the emission of two gamma radiations in cascade as result of the nuclear decay of the probe nucleus excited state. The technique relies on the measurement of γ − γ angular correlation which is perturbed by the interaction between the nuclear moment (magnetic dipole moment and/or electric quadrupole moment) of the intermediate nuclear state and extra nuclear fields (magnetic field or electric field gradient (EFG)).^54,55 The perturbation function R(t) = A₂₂G₂₂(t), where A₂₂ and G₂₂(t) are the anisotropy and perturbation function, respectively, provides information about magnetic and electric HF interactions and is calculated from the measured time delayed (t) coincidence spectra with detectors at 90° and 180°. Specifically,

2

where f_i is the occupance of the site i and G₂₂ⁱ(t) is given by:

3

Here, n is the number of frequencies in a 3D space, s_2n is the frequency amplitude, δ_i is the width of frequency distribution, τ_R,i is the finite time resolution of the PAC detectors, ω_n is the interaction frequency, and can be a combination of the electric and magnetic interactions, or purely the electric interaction, as the one discussed in this paper. ω_n and s_2n are related to the HF splitting of the intermediate nuclear level. In the specific case where the material presents no HF magnetic frequency, ω_n depends solely upon the nuclear quadrupole frequency (ω_Q)

4

and the asymmetry parameter (η)

5

Which can assume values between 0 and 1. Here, Q is the nuclear electric quadrupole moment of the intermediate level, I is the spin of the intermediate level of the probe nuclei, e is the fundamental unit of charge, , h is the Planck constant, and V_xx, V_yy, V_zz are the elements of the EFG tensor in its principal axis system. In the analysis of the experimental data, the spin-independent quadrupole frequency is used. Details of analysis of PAC spectra are reported by da Silva et al.⁵⁶

For ¹¹¹In(¹¹¹Cd) Q = 0.83 × 10⁻²⁸ m² (0.83 b) for the . Further information about PAC spectroscopy and experimental details can be found elsewhere.^57–60 Also, the distribution width δ reflects the number of probes nuclei sensing the distortions of their environment. This number can be used to identify sites near to grain boundaries, vacancies, or impurities.⁶¹ In NPs, it can be used to locate probe nuclei located on the particle surface.⁴⁶

PAC spectroscopy was performed in the Research Reactor Center (CERPq) at IPEN/CNEN-SP with a conventional slow-fast coincidence setup using four BaF₂ detectors in 90° geometry generating 12 γ − γ coincidence spectra. For these measurements, about 20 µCi of ¹¹¹In(¹¹¹Cd) in the form of indium chloride solution, to be used as probe nuclei, was added during the NPs synthesis presented previously. After annealing at 600 °C for 10 h, a small amount of the powder was characterized by TEM. The remaining was divided into 3 groups:

• Group 1: pre-irradiation XRD;

• Group 2: pre-irradiation PAC;

• Group 3: divided into 4 sub-groups:

- Sub-group 1: irradiated with a dose of 3 Gy;

- Sub-group 2: irradiated with a dose of 5 Gy;

- Sub-group 3: irradiated with a dose of 7 Gy;

- Sub-group 4: irradiated with a dose of 20 Gy;

A dose of 20 Gy was deliberately selected because it lies within the dose range employed in stereotactic body radiotherapy (SBRT) and stereotactic radiosurgery (SRS) protocols, in which single-fraction or hypofractionated doses of 15–24 Gy are routinely used for the treatment of localized tumors. This dose therefore serves as a clinically motivated upper-bound condition that extends the investigation beyond conventional fractionated radiotherapy doses (typically 2 Gy per fraction), maximizing the likelihood of detecting subtle radiation-induced structural modifications within a clinically encountered dose range.

Samples were irradiated in the ⁶⁰Co-Gammacell at room temperature. Note: because a radioactive sample is necessary for PAC measurements, due to safety protocols, the samples irradiation using the 6 MV photon beam was not possible.

3. Results and discussion

Post-annealing NPs presented an average size of 15 nm (σ = 1 nm) for Gd₂O₃ and 15 nm (σ = 3 nm) for Er₂O₃ (Fig. S1). XRD data for both compounds were analyzed using Rietveld refinement employing Rietica software⁶² (spectra for both samples are presented in Fig. S2), and show a cubic structure that belongs to the Ia3 space group, also known as bixbyite, where the RE atoms present two forms of six coordinated first neighbor oxygen polyhedral environment. As illustrated in Fig. 4 using the VESTA software,⁶³ the unit cell contains 48 oxygen atoms and 32 RE atoms occupying two sites: site D, presenting axial symmetry (Fig. 1b), and site C, asymmetric (Fig. 1c), as expected.⁶¹

Fig. 4. (a) Unit cell of RE₂O₃ in the bixbyite crystal structure (Ia3), showing RE and oxygen atom positions. (b) Detailed symmetric site (site D, violet sphere) (c) detailed asymmetric (site C, yellow sphere) site.

3.1. Radiosensitization

The dose enhancement factor (DEF) for each sample was calculated using the average absorbance of 5 or 3 cuvettes (5 for the ⁶⁰Co and 3 for the 6 MV photon beam) and the uncertainty for each value was calculated using type A (statistical analysis of repeated measurements) and B (non-statistical sources) uncertainties. The results are shown in Tables 1 and 2, for samples irradiated at the ⁶⁰Co and 6 MV photon beam, respectively (graphs of DEF as a function of dose are presented in Fig. S8).

Table 1. Dose enhancement factor (DEF) for rare-earth NPs and bulks irradiated at the ⁶⁰Co-Gammacell.

Dose	DEF
	Gd2O3-NP	Gd2O3-bulk	Er2O3-NP	Er2O3-bulk
3 Gy	1.57 ± 3	1.43 ± 2	1.47 ± 1	1.38 ± 1
5 Gy	1.19 ± 2	1.22 ± 1	1.20 ± 1	1.19 ± 1
7 Gy	1.19 ± 4	1.13 ± 1	1.13 ± 1	1.12 ± 1

Table 2. Dose enhancement factor (DEF) for rare-earth NPs and bulks irradiated at the 6 MV photon beam.

Dose	DEF
	Gd2O3-NP	Gd2O3-bulk	Er2O3-NP	Er2O3-bulk
3 Gy	1.094 ± 2	1.056 ± 2	1.051 ± 1	1.065 ± 7
5 Gy	1.069 ± 5	1.046 ± 4	1.042 ± 5	1.036 ± 5
7 Gy	1.045 ± 5	1.030 ± 4	1.06 ± 1	1.029 ± 6

Fricke gel dosimeters with rare-earth oxides present higher absorbance than those without them (Fig. S4 to S7). According to Schreiner,⁴⁰ the production yield of ferric ions in a Fricke gel dosimeter is increased when additional “chemical paths” for the conversion of Fe²⁺ in Fe³⁺ are added. The interaction between high-Z atoms and the incident radiation provides electrons, which present higher linear energy transfer, increasing the probability of water molecule bonds breaking and forming free radicals, thus creating additional “chemical paths” for Fe²⁺ oxidation.

It is interesting to note that similar DEFs were found when NPs or bulk of the same oxide are used. Apparently, in this case, the presence of a high-Z material, regardless of size, is sufficient to increase the dose delivered to the gel, despite the possibility of reabsorption.⁶⁴ However, higher doses resulted in smaller DEFs. This is expected because high doses produce more free radicals, thereby resulting in a higher radical recombination probability.⁶⁵

In addition, DEFs obtained on dosimeters irradiated in the Gammacell were higher than those exposed to the 6 MV photon beam. It should be noted that this difference is not attributed to a systematic advantage of one irradiation modality over the other, but rather reflects the distinct beam geometries, energy spectra, and experimental setups employed (see Section 2). The primary purpose of including both modalities was to demonstrate that dose-enhancement effects are present across two independent and clinically relevant irradiation conditions. Initially, one might attribute this difference to the distinct beam energies involved (6 MeV in the clinical system and 1173.2 keV and 1332.5 keV from ⁶⁰Co). Because lower-energy photons have a higher probability of undergoing photoelectric interactions, leading to reduced photon scattering and increased electron ejection, more energy is deposited within a small volume, thereby increasing the DEF. However, previous studies show that the energy peak for similar 6 MV photon beam systems is around 1 MeV,‡^66–69 which is close to the two ⁶⁰Co primary energy peaks. Therefore, DEF should be similar for both systems. Instead, this difference could be due to the different setups used in the two procedures: layers of solid water and bolus were used on the 6 MV photon beam (see Fig. 3), versus direct exposure of the dosimeters to the gamma beam from ⁶⁰Co (Fig. S1). Another difference is the energy distribution around the peak, which is broader in the 6 MV system than the ⁶⁰Co. More studies are needed to better understand the origin of this difference.

Overall, the results obtained here are in good agreement with those reported in the literature for NPs with a similar atomic number and in a similar beam-energy range.^14,70,71 This finding confirms the feasibility of using more cost-effective materials and equipment, such as the Fricke gel dosimeter and spectrophotometer, as physicochemical screening tools for dose-enhancement evaluation. These results motivate future biological studies—including clonogenic survival assays, ROS detection, and DNA-damage quantification—to fully characterize the radiosensitizing efficacy of these nanoparticles in cellular systems.

3.2. RENPs structural changes due to radiation therapy doses

For PAC, NPs had ¹¹¹In(¹¹¹Cd) incorporated. The XRD shown in Fig. 2 was performed on these samples after complete decay of the radioactive ¹¹¹In(¹¹¹Cd).

XRD results show peak shifts as a function of radiation dose for both Gd₂O₃-NP and Er₂O₃-NP, but no change to the overall pattern beyond a lattice expansion/contraction, indicating preservation of the bixbyite crystal structure throughout the investigated dose range. No evidence of phase transformation, amorphization, or irradiation-induced crystalline degradation was detected by XRD (Fig. 5).

Fig. 5. X-ray diffraction spectra measured at room temperature for Gd₂O₃ and Er₂O₃. The insets show in detail the displacement of the peak (222) for both samples. Vertical bars at the bottom of each graph represent theoretical peaks position. All error bars are shown and represent 1σ. However, the error bars may be smaller than the symbol.

The preservation of the diffraction pattern demonstrates that irradiation doses up to 20 Gy do not induce large-scale structural degradation. Therefore, any radiation-induced modifications are expected to originate primarily from local defect rearrangements, vacancy redistribution, and subtle distortions of the rare-earth coordination environment rather than from global crystallographic transformations. Therefore, XRD alone does not determine the origin of lattice distortions, so we use PAC to study the local structural changes.

Visual inspection reveals no appreciable differences between the non-irradiated sample and the sample exposed to 20 Gy (Fig. 6). A comparison between −R(t) for all samples is shown in the SI (Fig. S9 and S10). The data were fitted using a three-site model for the quadrupole interactions: asymmetric (black lines), symmetric (red lines), and a third site representing the probe nuclei distributed on the NP surface (green lines).⁴⁶ The blue lines are theoretical fits to the experimental data.

Fig. 6. Perturbation functions for non-irradiated Gd₂O₃-NPs and Er₂2O₃-NPs, measured at room temperature. Blue lines are theoretical fits of eqn (2) to experimental data. All error bars are shown and represent 1σ. However, the error bars may be smaller than the symbol.

The fast damping observed in the surface site (green line) indicates a large interaction of the probe nuclei not only with the oxygens in the bixbyite structure, but also with vacancies, atoms present on the capping agent, etc., resulting in a broad distribution of the electric field gradient (EFG). While the HF parameters associated with the surface site are not easily determined since they depend on many factors, such as defects on the surface of the NP, probe nuclei present on the capping material, etc., the surface site must be considered to properly obtain fraction of other sites. This is due to the large surface/volume ratio of NPs over bulk materials.

The HF parameters (site fraction, nuclear quadrupole frequency (ν_Q), asymmetry parameter (η) and width of frequency distribution (δ)), obtained from the −R(t) analysis, and the lattice parameters of Gd₂O₃-NP and Er₂O₃-NP, as a function of dose, are shown in Fig. 7 (and Tables S1 and S2).

Fig. 7. On the left and right axis are presented, respectively, the hyperfine (black sphere, red square, green triangle) and lattice (blue star) parameters for Gd₂O₃ (left panel) and Er₂O₃ (right panel) nanoparticles measured at room temperature. The solid curves are only to guide the eyes. All error bars are shown and represent 1σ. However, the error bars may be smaller than the symbol.

For non-irradiated samples, the measured quadrupole frequencies and asymmetry parameter are ν_Q = 63.3 ± 2 MHz and η = 1.00 ± 6 (asymmetric site), and ν_Q = 146.2 ± 4 MHz and η = 0 (symmetric site) for Gd₂O₃-NP, and ν_Q = 84.7 ± 5 MHz and η = 0.79 ± 1 (asymmetric site), and ν_Q = 148 ± 1 MHz and η = 0 (symmetric site) for Er₂O₃-NP, and are in good agreement with those presented in the literature.^61,72 However, dose-dependent changes in these two parameters are observed in both samples at both sites (Fig. 7).

For Gd₂O₃-NP (Fig. 7, left panel), the lattice parameter decreases with radiation dose. PAC spectroscopy data show, for the asymmetric site, a slight increase in ν_Q as a function of the dose, indicating a growing local electric field gradient. However, η decreases for the same radiation doses, suggesting changes in the local symmetry around the nucleus. No significant changes were observed for δ. At 3 Gy ν_Q, η and δ remain nearly unchanged compared to the non-irradiated sample, indicating that a higher dose is necessary to cause significant structural changes to the asymmetric site. On the other hand, for the symmetric site, ν_Q shows no significant changes. η increases to around 0.1, indicating a slight change in the symmetry of this site. δ Does not change significantly.

For Er₂O₃-NP (Fig. 7, right panel), the lattice parameter increases from 3 Gy to 5 Gy, decreases from 5 Gy to 7 Gy, and increases again at 20 Gy. This result is corroborated by PAC spectroscopy. ν_Q of the asymmetric site follows the opposite trend of the lattice parameter, decreasing when it increases, and vice versa. On the other hand, η and δ follow the same trend as the lattice parameter. δ Varies significantly, going from 0.06 to 0.15, which indicates that the probe nuclei are not well localized at this site, probably due to a distortion caused by changes in the unit cell. No significant changes in the HF parameters were observed at the symmetric site, except for η, which increased from zero in the non-irradiated sample, to approximately 0.2, indicating a modification in the symmetry of the site.

The results for the symmetric site in both compounds indicate that this site is less sensitive to radiation effects than the asymmetric site. The correlation between the decreasing lattice parameter and the changes in the HF parameters for the asymmetric site supports the occurrence of structural distortion, as indicated by the changes observed in ν_Q, η and δ.^73–75 Notably, the asymmetric site is more affected than the symmetric site, suggesting that vacancies and the resulting distortions preferentially disrupt lower symmetry environments. Crucially, however, neither Gd₂O₃ nor Er₂O₃ nanoparticles exhibit phase transformation, secondary phase formation, amorphization, or large-scale disruption of the bixbyite crystal lattice at any investigated dose up to 20 Gy, as evidenced by the absence of new diffraction peaks, additional peak broadening, or loss of crystallinity in the XRD patterns. The observed PAC modifications—changes in electric-field gradients, site symmetries, and distribution widths—are consistent with local atomic-scale vacancy rearrangements within an otherwise intact crystalline framework. Therefore, both nanoparticle systems can be regarded as structurally stable under clinically relevant irradiation doses, while exhibiting dose-dependent atomic-scale modifications that do not compromise global crystal structure integrity. This may reflect less stability at these crystallographic sites under irradiation. However, further studies are required to elucidate whether the observed distortion originates from the generation of new vacancies, the migration of pre-existing ones, or a combination of both.

A relevant safety consideration for the prospective pharmaceutical use of rare-earth oxide nanoparticles is whether radiation-induced lattice distortions could promote the release of Gd³⁺ or Er³⁺ ions into biological media. Although PAC spectroscopy detected local lattice distortions and defect rearrangements, neither XRD nor PAC results indicate significant structural degradation that would compromise the integrity of the bixbyite framework. Therefore, the current results do not provide evidence for irradiation-induced lattice breakdown, widely considered a prerequisite for significant ion leaching, within the investigated dose range. Nevertheless, direct quantification of dissolved rare-earth ions by inductively coupled plasma mass spectrometry (ICP-MS) in relevant biological media (e.g., phosphate-buffered saline or cell culture medium) was not performed in the present study, as it addresses dissolution kinetics and surface chemistry governed by factors beyond crystallographic stability alone. However, it would be valuable to directly quantify the release of Gd³⁺ or Er³⁺ ions in physiological media using techniques such as ICP-MS. This would provide complementary information and is an important topic for future investigations. Additionally, atomistic simulations, such as density functional theory (DFT) calculations or molecular dynamics (MD) simulations, could offer valuable insight into the energetics and dynamics of irradiation-induced defects in bixbyite-structured rare-earth oxides. These types of investigations are beyond the scope of the current study and will be pursued in future work.

4. Conclusions

In this work, we established a multiscale experimental framework to investigate the response of Gd₂O₃ and Er₂O₃ nanoparticles to clinically relevant irradiation doses. Fricke gel dosimetry demonstrated that both nanoparticle systems produce measurable dose-enhancement effects under both ⁶⁰Co γ-rays and a clinical 6 MV photon beam, confirming their potential as high-Z radiosensitizing materials.

Complementary XRD and PAC measurements revealed that, although the bixbyite crystal structure remains preserved throughout the investigated dose range, irradiation induces subtle but measurable modifications in local atomic environments. The combined structural analyses showed that radiation exposure promotes site-specific defect evolution, reflected by changes in lattice parameters, electric-field gradients, asymmetry parameters, and frequency-distribution widths.

These effects are particularly pronounced at the low-symmetry rare-earth sites, indicating that local structural distortions and vacancy rearrangements preferentially occur in crystallographic environments with lower symmetry. Importantly, no evidence of phase transformation, amorphization, or irradiation-induced structural collapse was observed up to 20 Gy, demonstrating the high radiation resilience of both nanoparticle systems.

The present study has some limitations that should be acknowledged. First, the work focuses on structural evolution and dose-enhancement behavior and does not include biological validation through cellular radiosensitization assays, reactive oxygen species quantification, DNA-damage analysis, or in vivo investigations. Second, direct measurements of irradiation-induced ion release, dissolution behavior, or long-term nanoparticle stability in physiological environments were not performed. Third, the microscopic mechanisms responsible for the observed defect rearrangements were inferred from experimental observations and were not supported by atomistic simulations.

Future studies should therefore combine the multiscale structural methodology presented here with biological, toxicological, and pharmacological investigations to establish direct relationships between radiation-induced defect evolution and radiosensitization performance. In addition, complementary ICP-MS measurements, long-term stability studies in biological media, and theoretical calculations based on density functional theory or molecular dynamics simulations would provide valuable insight into ion-release mechanisms, vacancy migration processes, defect formation, and the atomistic origin of the observed hyperfine interactions.

Overall, this work demonstrates that PAC spectroscopy and XRD provide a powerful and complementary approach for probing radiation-induced structural evolution in rare-earth oxide nanoparticles. The insights gained from atomic-scale defect evolution offer valuable structural design insights for the development and optimization of rare-earth nanoparticles with enhanced radiation resilience. The results contribute to a deeper understanding of defect dynamics and structural stability under irradiation and provide an experimental foundation for the future development and optimization of radiation-resilient radiosensitizing nanomaterials.

5. Disclaimer

We identify certain commercial equipment, instruments, or materials in this article to specify adequately the experimental procedure. In no case does such identification imply recommendation or endorsement by the National Institute of Standards and Technology, nor does it imply that the materials or equipment identified are necessarily the best available for the purpose.

Author contributions

Eduardo L. Correa: conceptualization, data curation, formal analysis, investigation, methodology, writing original-draft, writing – review and editing. Brianna Bosch-Santos: formal analysis, investigation, methodology, visualization, writing – review and editing Gabriel A. Cabrera-Pasca: investigation, supervision, visualization, writing – review and editing Glauco R. Veneziani: investigation, methodology. Renata O. Bressane: investigation, methodology. Tatiane S. N. Sales: investigation, methodology. Wanderson L. Ferreira: investigation, methodology. Luciano F. D. Pereira: investigation, methodology. Roberto K. Sakuraba: formal analysis, investigation, methodology. Rajendra N. Saxena: supervision, validation. Maria da Penha A. Potiens: funding acquisition, investigation, methodology, and Artur W. Carbonari: conceptualization, data curation, funding acquisition, methodology, project administration, resources, supervision, validation, writing – review and editing.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

All data that support the findings of this study are included within the article (and any supplementary information (SI) files). Supplementary information is available. See DOI: https://doi.org/10.1039/d6ra03417a.