Dual X-ray- and Neutron-Shielding Properties of Gd2O3/NR Composites with Autonomous Self-Healing Capabilities

The neutron- and X-ray-shielding, morphological, physical, mechanical, and self-healing properties were investigated for natural rubber (NR) composites containing varying gadolinium oxide (Gd2O3) contents (0, 25, 50, 75, and 100 parts per hundred parts of rubber; phr) to investigate their potential uses as self-healing and flexible neutron- and X-ray-shielding materials. Gd2O3 was selected as a radiation protective filler in this work due to its preferable properties of having relatively high neutron absorption cross-section (σabs), atomic number (Z), and density (ρ) that could potentially enhance interaction probabilities with incident radiation. The results indicated that the overall neutron-shielding and X-ray-shielding properties of the NR composites were enhanced with the addition of Gd2O3, as evidenced by considerable reductions in the half-value layer (HVL) values of the samples containing 100 phr Gd2O3 to just 1.9 mm and 1.3 mm for thermal neutrons and 60 kV X-rays, respectively. Furthermore, the results revealed that, with the increase in Gd2O3 content, the mean values (± standard deviations) of the tensile strength and elongation at break of the NR composites decreased, whereas the hardness (Shore A) increased, for which extreme values were found in the sample with 100 phr Gd2O3 (3.34 ± 0.26 MPa, 411 ± 9%, and 50 ± 1, respectively). In order to determine the self-healing properties of the NR composites, the surfaces of the cut samples were gently pressed together, and they remained in contact for 60 min; then, the self-healing properties (the recoverable strength and the %Recovery) of the self-healed samples were measured, which were in the ranges of 0.30–0.40 MPa and 3.7–9.4%, respectively, for all the samples. These findings confirmed the ability to autonomously self-heal damaged surfaces through the generation of a reversible ionic supramolecular network. In summary, the outcomes from this work suggested that the developed Gd2O3/NR composites have great potential to be utilized as effective shielding materials, with additional dual shielding and self-healing capabilities that could prolong the lifetime of the materials, reduce the associated costs of repairing or replacing damaged equipment, and enhance the safety of all users and the public.


Introduction
As the demand for greener technologies has rapidly increased in recent years following the Sustainable Development Goals (SDGs) introduced by the United Nations both thermal neutron and high-energy photon attenuations and present the advantages of Gd 2 O 3 as a radiation-protective filler in comparison to common Pb, Bi, and B compounds.
Another important factor to consider in producing radiation-shielding materials is the selection of the main matrix, for which the selection largely depends on the requirements of the intended applications. For example, applications requiring high flexibility, strength, and elongation usually rely on natural rubber (NR) or synthetic rubber (SR). For example, B 2 O 3 /NR [11], B 4 C/NR [20], and H 3 BO 3 /ethylene-propylene diene monomer (EPDM) [21] composites have been developed for use as flexible, neutron-shielding materials, while Bi 2 O 3 /EPDM [13], WO 3 /EPDM [13], BaSO 4 /NR [14], and Pb/NR [22] composites have been utilized as flexible, X-ray-shielding and gamma-shielding materials. While these composites could serve their mandatory purpose (the ability to attenuate incident thermal neutrons or X-rays (depending on filler type) with high flexibility and strength), the lack of self-healing capabilities in most common NR and SR composites has resulted in extra procedures or new materials needed to restore full function once the materials are damaged, inevitably shortening their lifetimes and increasing operational costs. To resolve these shortcomings, Xu et al. successfully developed autonomously self-healing NR composites by introducing controlled peroxide-induced vulcanization to generate ionic cross-links to NR networks via the polymerization of zinc dimethacrylate (ZDMA), which slowed the formation of non-reversible covalent cross-links while generating a reversible ionic supramolecular network to NR, enabling the ability to autonomously heal after damage [23,24]. Hence, to expand their usefulness to other applications, the concept of autonomously self-healing NR materials can be adapted for the production of radiationshielding materials, which could not only present the mentioned benefits but also improve safety for radiation users from damaged equipment.
Therefore, this current work investigates the properties of flexible Gd 2 O 3 /NR composites for their potential use as dual thermal-neutron-and X-ray-shielding materials with autonomously self-healing capabilities by introducing reversible ionic supramolecular cross-links to NR networks. In order to understand the effects of the Gd 2 O 3 fillers on the properties of the composites, the Gd 2 O 3 contents were varied with values of 0, 25, 50, 75, and 100 parts per hundred parts of rubber (phr) by weight to thoroughly investigate the properties of interest, which consisted of thermal-neutron and X-ray-shielding properties (based on the linear attenuation coefficient (µ), the mass attenuation coefficient (µ m ), the half-value layer (HVL), the tenth-value layer (TVL), and the Pb equivalence (Pb eq )), as well as mechanical (based on tensile strength, elongation at break, and hardness (Shore A) both before and after self-healing), morphological, and physical (based on density) properties. The outcomes of this work can not only present valuable information on the dual neutronand X-ray-shielding properties of the developed Gd 2 O 3 /NR composites, but may also offer a novel procedure to obtain self-healing NR composites that is beneficial for the future development of other radiation-shielding products.

Materials and Chemicals
Natural rubber (STR 5CV) with a Mooney viscosity of 60.8 (at 100 • C) was supplied by Hybrid Post Co., Ltd. (Bangkok, Thailand). The names, contents, roles, and suppliers of the chemicals used for sample preparation are shown in Table 1. An image of Gd 2 O 3 powder captured using a scanning electron microscope (SEM; Quanta 450 FEI: JSM-6610LV, Eindhoven, the Netherlands) is shown in Figure 1, which indicates that the average particle size of the Gd 2 O 3 powder was 3.4 ± 0.4 µm, as determined using ImageJ software version 1.50i (Bethesda, MD, USA).

Sample Preparation
The NR samples were prepared using two steps: mastication and then compounding. Initially, the NR was masticated on a two-roll mill (R11-3FF, Kodaira Seisakusho Co., Ltd., Tokyo, Japan) for 5 min. Then, the masticated NR was compounded with the chemicals (Table 1) for a further 15-20 min. Notably, although the content of Gd2O3 was as high as 100 phr, the much higher density of Gd2O3 (ρ = 7.4 g/cm 3 ) than that of NR (approximately 0.93-0.97 g/cm 3 [25]) resulted in the volume of Gd2O3 powder used during the compounding being much less than that of NR, making the mixing of all the chemicals on a two-roll mill possible. After the compounding, the NR samples were vulcanized using hot com-

Sample Preparation
The NR samples were prepared using two steps: mastication and then compounding. Initially, the NR was masticated on a two-roll mill (R11-3FF, Kodaira Seisakusho Co., Ltd., Tokyo, Japan) for 5 min. Then, the masticated NR was compounded with the chemicals (Table 1) for a further 15-20 min. Notably, although the content of Gd 2 O 3 was as high as 100 phr, the much higher density of Gd 2 O 3 (ρ = 7.4 g/cm 3 ) than that of NR (approximately 0.93-0.97 g/cm 3 [25]) resulted in the volume of Gd 2 O 3 powder used during the compounding being much less than that of NR, making the mixing of all the chemicals on a two-roll mill possible. After the compounding, the NR samples were vulcanized using hot compression molding (CC-HM-2060, Chaicharoen Karnchang Co., Ltd., Bangkok, Thailand) at 150 • C and a pressure of 160 kg/cm 2 for 150 secs in a mold with dimensions of either 15 cm × 15 cm × 0.2 cm or 10 cm × 10 cm × 0.2 cm. Notably, the procedure for sample preparation was mainly based on the published works of Xu et al. [23,24], while the cure time of 150 secs was selected following preliminary studies for optimized cure times, for which shorter or longer cure times resulted in the samples being too soft or too hard, respectively, which limited their useability and prevented the initiation of self-healing mechanisms from occurring [26]. The neutron shielding properties of the Gd 2 O 3 /NR composites were investigated at the Thailand Institute of Nuclear Technology (Public Organization), Bangkok, Thailand. The neutron-shielding parameters investigated in this work were the neutron transmission (I/I 0 ), the linear attenuation coefficient (µ), the half-value layer (HVL), and the tenth-value layer (TVL), and their relationships are shown in Equations (1)-(4) [14]: where I 0 is the initial intensity of the incident neutrons, I is the final intensity of the transmitted neutrons, x is the thickness of the sample, and ρ is the density of the sample. The setup for neutron-shielding measurement is schematically shown in Figure 2, with a 241 Am/Be used as a thermal neutron source. The values of I and I 0 were recorded using a 3 He neutron detector that was connected to a high-voltage supplier (Model 659, ORTEC, CA, USA), an amplifier (Model 2022, Canberra, CT, USA), and a time counter (Model TC 535P, Tennelec, TN, USA). The neutron source was positioned such that it was 0.89 m away from the NR sample and 1.00 m away from the detector. Notably, to investigate the effects of sample thickness on the neutron-shielding properties, the total thickness values of the NR samples were also varied (2, 4, 6, 8, and 10 mm). The neutron shielding properties of the Gd2O3/NR composites were investigated at the Thailand Institute of Nuclear Technology (Public Organization), Bangkok, Thailand. The neutron-shielding parameters investigated in this work were the neutron transmission (I/I0), the linear attenuation coefficient (µ), the half-value layer (HVL), and the tenthvalue layer (TVL), and their relationships are shown in Equations (1)-(4) [14]: where I0 is the initial intensity of the incident neutrons, I is the final intensity of the transmitted neutrons, x is the thickness of the sample, and ρ is the density of the sample. The setup for neutron-shielding measurement is schematically shown in Figure 2, with a 241 Am/Be used as a thermal neutron source. The values of I and I0 were recorded using a 3 He neutron detector that was connected to a high-voltage supplier (Model 659, ORTEC, CA, USA), an amplifier (Model 2022, Canberra, CT, USA), and a time counter (Model TC 535P, Tennelec, TN, USA). The neutron source was positioned such that it was 0.89 m away from the NR sample and 1.00 m away from the detector. Notably, to investigate the effects of sample thickness on the neutron-shielding properties, the total thickness values of the NR samples were also varied (2, 4, 6, 8, and 10 mm).

X-Ray-Shielding Properties
The schematic setup for X-ray-shielding measurement is shown in Figure 3. The measurement was carried out at the Secondary Standard Dosimetry Laboratory (SSDL), the Office of Atoms for Peace (OAP), Bangkok, Thailand. The X-ray-shielding parameters investigated in this work were X-ray transmission (I/I0), the linear attenuation coefficient (µ), the mass attenuation coefficient (µm), the half-value layer (HVL), the tenth-value layer

X-ray-Shielding Properties
The schematic setup for X-ray-shielding measurement is shown in Figure 3. The measurement was carried out at the Secondary Standard Dosimetry Laboratory (SSDL), the Office of Atoms for Peace (OAP), Bangkok, Thailand. The X-ray-shielding parameters investigated in this work were X-ray transmission (I/I 0 ), the linear attenuation coefficient (µ), the mass attenuation coefficient (µ m ), the half-value layer (HVL), the tenth-value layer (TVL), and the Pb equivalence (Pb eq ), for which Pb eq could be determined using Equation (5) [14]: where µ Pb is the linear attenuation coefficient of a pure Pb sheet. It should be noted that the values of µ Pb were 63.06 cm −1 and 25.99 cm −1 for the incident X-ray energies of 45 keV and 80 keV, respectively, and were numerically determined using XCOM software (National Institute of Standards and Technology, Gaithersburg, MD, USA) [14,27]. The X-ray energies of 45 keV and 80 keV were selected for the determination of Pb eq due to being the average energies of X-rays generated from an X-ray tube (YXLON MGC41, NY, USA) with the supplied voltages of 60 and 100 kV (Keithley 651B, OH, USA), respectively, used in this work. The emitted X-ray beam was collimated using a Pb collimator with a 1 mm pinhole, and the transmitted X-rays were detected and counted using a free-air ionization chamber (Korea Research Institute of Standards and Science; KRISS, Daejeon, Korea). More details for the setup of the neutron-shielding measurement are available in [14]. Similar to the neutron measurement, the total thickness values of the NR samples varied from 2 to 10 mm in 2 mm increments to investigate effects of material thickness on X-ray-shielding abilities.
Polymers 2022, 14, x FOR PEER REVIEW 6 of 18 (TVL), and the Pb equivalence (Pbeq), for which Pbeq could be determined using Equation (5) [14]: where µPb is the linear attenuation coefficient of a pure Pb sheet. It should be noted that the values of µPb were 63.06 cm −1 and 25.99 cm −1 for the incident X-ray energies of 45 keV and 80 keV, respectively, and were numerically determined using XCOM software (National Institute of Standards and Technology, Gaithersburg, MD, USA) [14,27]. The X-ray energies of 45 keV and 80 keV were selected for the determination of Pbeq due to being the average energies of X-rays generated from an X-ray tube (YXLON MGC41, NY, USA) with the supplied voltages of 60 and 100 kV (Keithley 651B, OH, USA), respectively, used in this work. The emitted X-ray beam was collimated using a Pb collimator with a 1 mm pinhole, and the transmitted X-rays were detected and counted using a free-air ionization chamber (Korea Research Institute of Standards and Science; KRISS, Daejeon, Korea). More details for the setup of the neutron-shielding measurement are available in [14]. Similar to the neutron measurement, the total thickness values of the NR samples varied from 2 to 10 mm in 2 mm increments to investigate effects of material thickness on X-rayshielding abilities.

Mechanical Properties
The mechanical properties of tensile strength and elongation at break for all the Gd2O3/NR composites were determined using a universal testing machine (Auto-graph AG-I 5kN, Shimadzu, Kyoto, Japan) following ASTM D412-06 standard testing. The tensile testing speed used for all the samples was 50 mm/min. The surface hardness (Shore A) was determined using a hardness durometer (Teclock GS-719G, Japan) following the ASTM D2240-05 standard testing method. For the determination of the self-healing capabilities of the developed Gd2O3/NR composites, samples having shapes and sizes based on ASTM D412-06 standard testing were cut into two equal pieces using a surgical knife and were immediately brough into contact. Then, after 60 min of contact, the samples were installed in a universal testing machine (Auto-graph AG-I 5kN, Shimadzu, Kyoto, Japan) to determine their tensile strength and elongation at break, following the same testing procedures as those for the uncut samples. Then, the tensile strength values of the self-healed samples were used for

Mechanical Properties
The mechanical properties of tensile strength and elongation at break for all the Gd 2 O 3 /NR composites were determined using a universal testing machine (Auto-graph AG-I 5kN, Shimadzu, Kyoto, Japan) following ASTM D412-06 standard testing. The tensile testing speed used for all the samples was 50 mm/min. The surface hardness (Shore A) was determined using a hardness durometer (Teclock GS-719G, Japan) following the ASTM D2240-05 standard testing method. For the determination of the self-healing capabilities of the developed Gd 2 O 3 /NR composites, samples having shapes and sizes based on ASTM D412-06 standard testing were cut into two equal pieces using a surgical knife and were immediately brough into contact. Then, after 60 min of contact, the samples were installed in a universal testing machine (Auto-graph AG-I 5kN, Shimadzu, Kyoto, Japan) to determine their tensile strength and elongation at break, following the same testing procedures as those for the uncut samples. Then, the tensile strength values of the self-healed samples were used for the calculation of the percentage of recoverable strength (%Recovery) using Equation (6) [18]: where TS self-healing and TS uncut are the tensile strengths of the self-healed and uncut samples, respectively [28].

Density Measurement
The densities for all the Gd 2 O 3 /NR composites were determined using a densitometer (MH-300A, Shanghai, China) with a precision of 0.01 g/cm 3 , and the determination was carried out based on the Archimedes principle [29]. Additionally, to verify the correctness of the density measurement, theoretical densities (ρ th ) for all the samples were calculated using Equation (7) [14]: where C NR is the content of NR, C Gd2O3 is the content of Gd 2 O 3 , ρ NR is the density of NR (0.95 g/cm 3 ), and ρ Gd2O3 is the density of Gd 2 O 3 (7.4 g/cm 3 ).

Morphological Studies
The morphology, dispersion of Gd 2 O 3 particles, and dispersion of Gd elements were determined using scanning electron microscopy (SEM) with energy-dispersive X-ray (EDX) spectroscopy (Quanta 450 FEI: JSM-6610LV, Eindhoven, the Netherlands) at a 10 kV accelerating voltage. Prior to the SEM-EDX studies, all specimens were coated with gold using a sputter coater (Quorum SC7620: Mini Sputter Coater/Glow Discharge System, Nottingham, UK) at a power voltage of 10 kV and a current of 10 mA for 120 secs. Table 2 indicates the experimental and theoretical densities, as well as the differences between these two values, of all the Gd 2 O 3 /NR composites investigated in this work. The results showed that the densities of the NR samples increased with increasing Gd 2 O 3 content, while the differences between the experimental and theoretical values were below 5.0% for all the samples, clearly verifying the correctness and reliability of the experimental values for further use. The positive relationship between density and filler content was due to the much higher density of Gd 2 O 3 than NR, resulting in a greater sample mass (determined at the same total volume) and, subsequently, greater overall density of the NR composites containing higher Gd 2 O 3 contents [30].

Neutron-Shielding Properties
The results for the neutron-shielding properties, consisting of I/I 0 , µ, µ m , HVL, and TVL, for all the Gd 2 O 3 /NR composites are shown in Figure 4, which indicates that the overall neutron-shielding properties of the samples increased with increasing Gd 2 O 3 content, as evidenced by the lower values of I/I 0 , HVL, and TVL and the higher values of µ and µ m in the samples containing higher contents of Gd 2 O 3 . This shielding enhancement from the addition of Gd 2 O 3 was mainly due to Gd having a much higher σ abs value than the C and H in NR (σ abs values for Gd, C, and H are 49,700 barns, 0.0035 barns, and 0.3326 barns, respectively [31]), resulting in considerably increased chances for incident thermal neutrons to be absorbed and attenuated by the composites and, consequently, leading to superior neutron-shielding properties for the Gd 2 O 3 /NR composites [32]. Figure 5 shows the disper-sion of Gd in the NR composites based on SEM-EDX and reveals that the highest elemental density of Gd was in the sample containing 100 phr Gd 2 O 3 (Figure 5d), confirming the rationale for the improved neutron-shielding properties of the Gd 2 O 3 /NR composites by the addition of Gd 2 O 3 .
Polymers 2022, 14, x FOR PEER REVIEW 9 of 18     In addition, Figure 4 indicates that a small addition of 25 phr Gd 2 O 3 to pristine NR could sharply increase the neutron-shielding properties of the samples, as seen by the sharp decreases in I/I 0 and HVL from 97% and 84 mm in pristine NR to just 55% and 2.3 mm, respectively, in the 25 phr Gd 2 O 3 /NR sample (values of I/I 0 and HVL were compared using 2 mm thick samples). This notable improvement in the neutron-shielding properties was mainly due to the sudden change in dominant neutron interactions from elastic scattering in the pristine NR to neutron absorption in the Gd 2 O 3 /NR composites, for which the latter mechanism was relatively more effective in neutron attenuation than the former [33]. However, as more Gd 2 O 3 powder was added to the composites, only a slight improvement was observed, perhaps because the samples already relied on the preferable absorption mechanism such that further addition of Gd 2 O 3 could only slightly increase the probabilities of neutrons being absorbed [34].
Another point worth mentioning is that the ability to attenuate neutrons increased with increasing sample thickness. The dependence of neutron-shielding properties on sample thickness, as illustrated in the determination of I/I 0 and shown in Figure 4a, was mostly due to more materials being available to elastically scatter (in the case of a pristine NR sample) or absorb (in the case of Gd 2 O 3 /NR samples) incident neutrons in thicker samples, subsequently reducing the transmitted neutrons (lower I/I 0 ). This relationship could also be mathematically explained using Equation (1) when re-arranged as shown in Equation (8), which depicts that the value of ln(I/I 0 ) was inversely proportional to x (thickness of the sample); hence, I/I 0 was negatively related to x: Figure 6 shows the results for the percentage of X-ray transmission for 60 kV and 100 kV X-rays, respectively, through varying thicknesses (2, 4, 6, 8, and 10 mm) of Gd 2 O 3 /NR composites. Similar to the results from the neutron-shielding measurement, the X-ray transmission decreased with increasing Gd 2 O 3 content and sample thickness. The dependence of the X-ray transmission on Gd 2 O 3 content was mainly due to the increased interaction probabilities between the incident X-rays and the materials through photoelectric absorption and Compton scattering with the addition of Gd 2 O 3 , for which their cross-sections were positively correlated to the Z and ρ values of the materials, hence improving the attenuation ability of the composites [35,36]. In addition, increasing the thickness of the samples could decrease X-ray transmission due to more Gd 2 O 3 being available to interact with the incident X-rays, resulting in fewer X-rays being transmitted through the samples.  Figure 6 shows the results for the percentage of X-ray transmission for 60 kV and 100 kV X-rays, respectively, through varying thicknesses (2, 4, 6, 8, and 10 mm) of Gd2O3/NR composites. Similar to the results from the neutron-shielding measurement, the X-ray transmission decreased with increasing Gd2O3 content and sample thickness. The dependence of the X-ray transmission on Gd2O3 content was mainly due to the increased interaction probabilities between the incident X-rays and the materials through photoelectric absorption and Compton scattering with the addition of Gd2O3, for which their cross-sections were positively correlated to the Z and ρ values of the materials, hence improving the attenuation ability of the composites [35,36]. In addition, increasing the thickness of the samples could decrease X-ray transmission due to more Gd2O3 being available to interact with the incident X-rays, resulting in fewer X-rays being transmitted through the samples. Figure 6. X-ray transmission of Gd2O3/NR composites with varying thicknesses from 2 to 10 mm and varying Gd2O3 contents from 0 to 100 phr determined at X-ray supplied voltages of (a) 60 kV and (b) 100 kV. Figure 7 shows the results of the µ, µm, HVL, TVL, and Pbeq values of Gd2O3/NR composites containing varying Gd2O3 content determined using 60 kV and 100 kV X-rays (common supplied voltages used for medical diagnostics [37]). The results imply that the overall X-ray-shielding properties of the composites generally increased with increasing Gd2O3 content, as evidenced by the lowest values of HVL and TVL and the highest values of µ, µm, and Pbeq being found in the sample with 100 phr Gd2O3. For example, the values of HVL (Pbeq) of the NR composites were reduced from 1.12 cm and 2.50 cm, respectively, (0.03 mmPb and 0.03 mmPb) for pristine NR to 0.36 and 0.65 cm (0.09 mmPb and 0.12 mmPb) for 25 phr Gd2O3/NR composites, determined at 60 kV and 100 kV X-rays, respec-  Figure 7 shows the results of the µ, µ m , HVL, TVL, and Pb eq values of Gd 2 O 3 /NR composites containing varying Gd 2 O 3 content determined using 60 kV and 100 kV X-rays (common supplied voltages used for medical diagnostics [37]). The results imply that the overall X-ray-shielding properties of the composites generally increased with increasing Gd 2 O 3 content, as evidenced by the lowest values of HVL and TVL and the highest values of µ, µ m , and Pb eq being found in the sample with 100 phr Gd 2 O 3 . For example, the values of HVL (Pb eq ) of the NR composites were reduced from 1.12 cm and 2.50 cm, respectively, (0.03 mmPb and 0.03 mmPb) for pristine NR to 0.36 and 0.65 cm (0.09 mmPb and 0.12 mmPb) for 25 phr Gd 2 O 3 /NR composites, determined at 60 kV and 100 kV X-rays, respectively, exhibiting approximately a 3-4-fold improvement in X-ray attenuation ability with the addition of just 25 phr Gd 2 O 3 .

X-Ray-Shielding Properties
Polymers 2022, 14, x FOR PEER REVIEW 11 of 18 Consequently, the incident X-rays emitted from an X-ray tube with higher supplied voltages could have fewer chances of interaction with the materials, resulting in more X-rays being transmitted and, subsequently, lower overall X-ray attenuation properties than lower-energy X-rays [39]. Notably, the values of Pbeq at a specific Gd2O3 content could be tailored according to the shielding requirements for intended applications by lowering or increasing the thickness of the sample (the sample thickness for Figure 7e was 3 mm) using Equation (5).  Table 3 shows a comparison of neutron-and X-ray-shielding properties (based on the values of HVL) from this work with other similar materials, which indicates that the current materials exhibited comparable or better neutron and X-ray attenuation capabilities than those from other reports. As a result, this comparison clearly confirmed the useability Another interesting result from Figure 7 was that the X-ray-shielding properties of the samples determined at the 60 kV supplied voltage were higher than those at the 100 kV supplied voltage due to the cross-section of the dominant photoelectric absorption (σ pe ) being inversely related to the cube of incident X-ray energy (E), as shown in Equation (9) [38]:

Comparative Neutron-and X-ray-Shielding Properties between Current and Other Similar Materials
Consequently, the incident X-rays emitted from an X-ray tube with higher supplied voltages could have fewer chances of interaction with the materials, resulting in more X-rays being transmitted and, subsequently, lower overall X-ray attenuation properties than lower-energy X-rays [39]. Notably, the values of Pb eq at a specific Gd 2 O 3 content could be tailored according to the shielding requirements for intended applications by lowering or increasing the thickness of the sample (the sample thickness for Figure 7e was 3 mm) using Equation (5). Table 3 shows a comparison of neutron-and X-ray-shielding properties (based on the values of HVL) from this work with other similar materials, which indicates that the current materials exhibited comparable or better neutron and X-ray attenuation capabilities than those from other reports. As a result, this comparison clearly confirmed the useability and potential of the NR composites for utilization as effective neutron-and X-ray-shielding materials with potential self-healing capabilities. It should be noted that the differences in the HVL values from all the materials in Table 3 could be due to several factors, such as differences in the filler types and contents used, as well as various energies of the incident radiation during measurement. Table 3. Comparison of neutron-and X-ray-shielding properties based on the half-value layer (HVL) between the results from this work and those from similar materials. Numbers in parentheses represent supplied voltages used for X-ray-shielding measurement.  Table 4 shows the mechanical properties-tensile strength, elongation at break, and hardness (Shore A) of Gd 2 O 3 /NR composites containing varying Gd 2 O 3 contents (0, 25, 50, 75, and 100 phr). The results indicate that the values of tensile strength and elongation at break generally decreased, while the hardness (Shore A) increased with increasing Gd 2 O 3 content. The decreases in tensile properties after the addition of Gd 2 O 3 could be due to poor surface compatibility between Gd 2 O 3 and the NR matrix, which possibly resulted in the formation of voids and discontinuities in the matrix, subsequently obstructing the transfer of external forces and reducing the overall strength and elongation of the materials [42]. Furthermore, the addition of high levels of Gd 2 O 3 contents, especially at 100 phr, led to high agglomeration of the Gd 2 O 3 particles due to filler-filler interactions that prevented more preferable and stronger rubber-filler interactions from occurring [43]. The SEM images depicting the dispersion of Gd 2 O 3 particles in the NR matrix are shown in Figure 8 and reveal that, while the particles were fairly evenly distributed throughout the NR matrix, some particle agglomeration was found in samples with higher filler contents, especially at 50, 75, and 100 phr (Figure 8c-e), compared to pristine NR (Figure 8a) and the sample with 25 phr filler content (Figure 8b). On the other hand, hardness (Shore A) had a strong positive relationship with Gd 2 O 3 content due to the high rigidity of the Gd 2 O 3 particles, which enhanced the overall rigidity and, hence, the hardness of the composites [44]. These findings are consistent with other reports, where the mechanical properties of materials generally decrease with the addition of high filler content, especially those developed for use in radiation protection [13,14].    Comparing the tensile properties obtained from this work with another work by Xu et al. indicated that the current pristine NR samples were approximately four times higher in tensile strength, as evidenced by the values reported in [23,24] being lower than 2 MPa for all the formulations. The differences in mechanical properties between these two works could be due to several factors, such as different formulation and cure times, as well as the types of NR used for sample preparation, which can affect the degree of cross-linking and, hence, the strength of the composites [45].

Mechanical Properties
3.6. Self-Healing Properties

Conclusions
This work developed dual neutron-shielding and X-ray-shielding NR composites containing varying contents of Gd2O3 (0, 25, 50, 75, and 100 phr) with autonomously selfhealing capabilities. The results showed that the added Gd2O3 acted as an effective protective filler against neutrons and X-rays, as evidenced by the decreases in I/I0, HVL, and TVL and the increases in µ, µm, and Pbeq of the NR composites after being added to the matrix. In addition, the results indicated that the increased filler content led to decreased tensile strength and elongation at break, whereas the hardness (Shore A) increased, mainly due to the initiation of particle agglomeration at high filler contents and poor surface compatibility between the NR matrix and the filler. The developed NR composites also offered self-healing capabilities at the fractured surfaces through a reversible ionic supramolecular network, with recoverable strength and %Recovery values in the ranges 0.30-0.4 MPa and 3.7-9.4%, respectively (after self-healing for 60 min). Based on the overall results obtained, the developed Gd2O3/NR composites showed great potential for use as novel and self-healing radiation-shielding materials that could effectively attenuate both neutrons and X-rays, thus prolonging the lifetime of the protective material and en- The reduction in the tensile strength and elongation at break of the self-healed NR composites in comparison to the original samples could be explained by the NR molecular chains in the uncut samples being originally cross-linked with a combination of covalent and ionic bonds, which resulted in relatively high tensile strength and elongation at break before the cut [46]. However, as the two damaged surfaces were gently pressed together and remained in contact for 60 min at room temperature (approximately 25 • C), a reversible ionic supramolecular network via the polymerization of ZDMA was able to recreate the sample through the mobility of NR molecular chains and, subsequently, restore some recoverable strength to the self-healed surfaces [47]. Nonetheless, the overall strengths of the self-healed samples were considerably lower than the original samples, with %Recovery values in the range of 3.7-9.1% (Figure 9c). This could be due to the reduced level of crosslink density in the samples after self-healing that could only be recreated by ionic bonds. On the other hand, the covalent bonds, which also initially presented and played major roles in providing exceptional strength to the original samples, were irreversible and, consequently, absent in the self-healed contact, resulting in much-reduced levels of recoverable strength and elongation at break for the samples [48]. Another factor that affected the self-healing mechanism was the addition of the Gd 2 O 3 particles to the NR matrix, for which the fillers were not a part of the reversible supramolecular network and, hence, hindered or blocked the initiation of self-healing [24].
Nonetheless, despite having Gd 2 O 3 contents of up to 100 phr, the recoverable strengths of the Gd 2 O 3 /NR composites were in the range of 0.30-0.40 MPa, which were in the same order of magnitude as that of pristine NR reported by Xu et al. (being in the range of 0.5-0.7 MPa, depending on self-healing times [23]), implying the useability and potential of the current self-healing materials for applications in radiation protection. Furthermore, the success of the current work could promote further attempts to develop 'smart' and more effective materials for use in radiation shielding, along with the previously reported composites of Bi 2 O 3 /PVA, Sm 2 O 3 /PVA, Gd 2 O 3 /PVA, graphene/PVA, and PbO 2 /acrylamide [18,28,49,50]. It should be noted that the %Recovery values of the NR samples with the addition of Gd 2 O 3 were higher than that of the pristine NR because the original Gd 2 O 3 /NR composite had 2-3 times lower tensile strength than pristine NR, while having similar recoverable strength after self-healing, which resulted in considerably higher %Recovery values for the Gd 2 O 3 /NR composites.

Conclusions
This work developed dual neutron-shielding and X-ray-shielding NR composites containing varying contents of Gd 2 O 3 (0, 25, 50, 75, and 100 phr) with autonomously self-healing capabilities. The results showed that the added Gd 2 O 3 acted as an effective protective filler against neutrons and X-rays, as evidenced by the decreases in I/I 0 , HVL, and TVL and the increases in µ, µ m , and Pb eq of the NR composites after being added to the matrix. In addition, the results indicated that the increased filler content led to decreased tensile strength and elongation at break, whereas the hardness (Shore A) increased, mainly due to the initiation of particle agglomeration at high filler contents and poor surface compatibility between the NR matrix and the filler. The developed NR composites also offered self-healing capabilities at the fractured surfaces through a reversible ionic supramolecular network, with recoverable strength and %Recovery values in the ranges 0.30-0.4 MPa and 3.7-9.4%, respectively (after self-healing for 60 min). Based on the overall results obtained, the developed Gd 2 O 3 /NR composites showed great potential for use as novel and self-healing radiation-shielding materials that could effectively attenuate both neutrons and X-rays, thus prolonging the lifetime of the protective material and enhancing the safety for users, as well as being a basis for the future development of 'smart' shielding materials.