Rethinking Radioactive Waste Management from Cellular Vulnerability to Molecular Containment

2.1 Rethinking fundamental properties through a biological filter

Conventional nuclear descriptions (half-life, decay mode, particular activity) attain physiological significance just when integrated with the mechanisms by which a radionuclide accesses, binds to, and interacts inside tissues. The documents correlate these physical properties closely with biokinetics, tissue permeability, and target-mimicking chemistry in imaging and therapy. The “optimal” physical half-life depends on uptake and washout in tumor vs normal tissues; too short wastes decays in transit, too long irradiates non-target tissues and the environment [11]. For instance, while uranium-238 isotope ²³⁸U has a half-life of 4.5 billion years, barium-90 isotope ⁹⁰Sr has a half-life of 28.8 years which results in greater exposure to barium-90 isotope within biological systems (e.g., binding to hydroxyapatite) through calcium absorption and its interaction with calmodulin [a calcium-binding protein that, through its ability to bind both Ca²⁺ and ⁹⁰Sr, retains more of these isotopes than would be predicted by physical decay properties alone].

Physical half-life alone is an incomplete basis for a physiological vulnerability metric, since it ignores how long a radionuclide is actually retained in the tissue that matters. The index below therefore uses the effective half-life, Teff = (Tr × Tb)/(Tr + Tb), where Tr is the physical (radiological) half-life and Tb is the biological half-life in the relevant tissue or organ, in place of T1/2 alone.

As a result of this comparison between these two radioisotopes, a new index of “Physiological Vulnerability” (PVI) may be determined:

Where:

Teff = effective half-life (Tr × Tb)/(Tr + Tb), years, normalized against ⁹⁰Sr

A spec = specific activity (Bq/g),

S = solubility in biological solutions (g/L at pH 7.4, 37 °C),

U = fraction of uptake through transporters of both isotopes and their biological significance (0-1),

R DNA = repair capability of DNA in target tissues compared to fibroblasts.

In rabbits administered with ⁹⁰Sr and ⁴⁵Ca, beginning bone uptake at 10 minutes was comparable; however, the timing and magnitude of maximum uptake varied: ⁴⁵Ca peaked at 1 day, whereas ⁹⁰Sr peaked at 4 hours, with the maximum uptake of ⁴⁵Ca exceeding that of ⁹⁰Sr. Areas characterized by elevated bone turnover (trabecular bone, bone extremities) exhibited increased turnover of both isotopes; nonetheless, these regions demonstrated reduced stable Sr/Ca ratios, suggesting a preferred long-term retention of calcium relative to strontium [12]. In vitro, bone exposed to solutions containing ⁴⁵Ca and ⁸⁵Sr took up more calcium than strontium; Ca/Sr uptake ratios were typically 1.2–1.4 and could reach ≥ 2 when stable Sr was added [13]. This behavior was similar for human and canine bone, bone powder, ashed bone, ethylenediamine‑treated bone, and apatite 2. Up to ~ 10 at% Sr²⁺ can enter the HA lattice without phase segregation, similar to physiological limits in bone. Higher Sr²⁺ levels favor formation of a Sr‑rich amorphous calcium phosphate (Sr(ACP)), which coexists with hydroxyapatite (HA) and is particularly stabilized at high Sr. Sr²⁺ incorporation induces lattice disorder and changes carbonate substitution patterns in HA, linking high Sr doses to mineral defects and pathologies such as osteomalacia in animal models [14]. When examining this index, physiological threats between these isotopes emerge with much greater vulnerability for both Sr and I (PVI = 2.3 for Sr and PVI = 2.1 for I) than Cs (PVI = 1.2). Even though the half-lives of these three radioisotopes are approximately equal, greater physiological threat exists from both Sr and I because of the mechanism of active transport (Sr using TRPV6; I using NIS) utilized to retain their presence in biological systems. In comparison, Cs distributes through biological systems via distribution uniformity, but Cs is only partially excluded from the cellular membrane by means of Na +/K + ATPase (sodium-potassium pump). PVI for other key radionuclides are summarized in Table 1.

This PVI framework immediately shows that physiological risk is not a simple function of radioactivity. It prioritizes radionuclides that exploit active uptake mechanisms and incorporate into essential biological structures. Isotopic half-life in animal tissues varies with body mass, tissue type, and taxon, which implies biological residence times are tissue-dependent rather than fixed constants across the body [15].

2.2 Non‑canonical physiological mechanisms

2.2.2 Mitochondrial dynamics as a sensitive radiobiological target

Conventional dosimetry focuses on nuclear DNA double‑strand breaks. Mitochondria have their own circular genome (mtDNA, 16.6kb), as well as limited repair capabilities [21, 22]. Human mtDNA lacks histones, has fewer repair pathways, and notably no canonical nucleotide excision repair; base excision repair is present but limited compared with the nucleus [23, 24]. Chronic beta-emitter (ie: ¹³⁷Cs and 90Sr) exposure leads to continuous production of ROS, primarily damaging mtDNA. The threshold for physiological effect is lower than for nuclear DNA damage [25].

When stress activates the inflammasome, mitochondrial DNA (mtDNA) is oxidized, losing or breaking into small pieces (fragments) about 500-650 bp in length. These mtDNA fragments exit the mitochondrion through two different routes. One route is by passing through the mitochondrial permeability transition pore (mPTP). The other route is via the opening of voltage-dependent anion channel (VDAC) oligomers found in non-apoptotic macrophages [26]. mPTP generally only allows molecules < 1.5 kDa to cross the membranes, meaning that whole nucleoids can’t cross directly – only mtDNA fragments – and that the long-term opening of the mPTP will likely cause a rupture in the membrane [27]. Live cells with stressed mtDNA have also shown a release of free intra-mtDNA fragments from the VDAC oligomers that are enriched in the D-loop region, with peaks around 110 bp – these DNA fragments can seed both circulating mtDNA and mtDNA found in the cytosol of the cells [28]. In cases of apoptosis and pyroptosis, BAX/BAK megapores and gasdermin pores can permeate the mitochondrial membrane and allow for the bulk exit of mtDNA to the cytosol; upon rupture of the plasma membrane, extracellular mtDNA is released into the surrounding tissues [29]. A “minority‑MOMP” state in senescent cells causes sublethal mtDNA release and chronic cGAS–STING activation without immediate cell death [30]. Since mitochondria are a significant producer and target of reactive oxygen species (ROS), they are crucial for cellular energy generation and red/ox equilibrium [31]. Under normal circumstances, mitochondria produce ROS as a byproduct of oxidative phosphorylation. However, oxidative stress damages proteins, lipids, and DNA, including mitochondrial DNA (mtDNA), when the total amount of exogenous and physiological ROS production surpasses the cell’s antioxidant defenses [31]. According to studies, mtDNA lacks protective histone structures and has fewer repair mechanisms than nuclear DNA, making it more vulnerable to DNA damage, particularly oxidative damage caused by ionizing radiation [32, 33, 34]. Cells increase mitochondrial biogenesis and mtDNA replication to preserve homeostasis and combat oxidative stress, guaranteeing sufficient mitochondrial function and energy generation [35]. Because more mitochondria are better able to control and detoxify ROS, increasing the number of mtDNA copies can strengthen the cell’s antioxidant defenses [36]. Figure 1 illustrates the key mitochondrial pathways involved in this response to chronic radionuclide exposure.

2.3 Physiological adaptation and the failure of LNT at low dose‑rates

Some evidence suggests that low‑dose, chronic radiation can activate protective biological mechanisms, complicating the simple linear‑no‑threshold (LNT) picture. Data from high background radiation areas (HBRAs) such as Ramsar and Kerala are central to this discussion. The LNT model, used in radiation protection, assumes cancer risk increases linearly with dose down to zero dose [37]. Under LNT, inhabitants of HBRAs should show clearly higher cancer incidence and mortality, but multiple studies report similar or even lower rates compared with normal background areas, generally without statistically significant excess risk [37, 38]. A biophysical model calibrated on Ramsar, Kerala, Yangjiang data estimates average reductions of ~ 10% in chromosomal aberrations, ~ 15% in cancer incidence and ~ 17% in cancer mortality in studies that did show adaptive response. About 45% of chromosomal aberration studies in HBRAs supported adaptive response, while the remaining data were inconsistent with LNT when tested [37]. Cellular and molecular reviews describe low‑dose “adaptive responses” involving enhanced DNA repair, antioxidant defenses, apoptosis, immune modulation and stress‑response pathways, which can reduce net damage from endogenous and exogenous insults [39]. Reviews of HBRAs and low‑dose exposure note absence of “unfavourable health effects or adverse cellular effects” attributable to high natural background, and suggest activation of defence mechanisms rather than clear excess chronic disease [38, 40]. Some epidemiologic and mechanistic syntheses argue that such responses support threshold or hormetic dose–response models rather than strict LNT at low doses [41, 42]. For example, pancreatic β‑cells have lower baseline antioxidant capacity than hepatocytes, so they fail to adapt at dose‑rates above 0.2 mGy/h. This means a waste matrix that releases Sr‑90 at a rate that is “safe” for the liver may be unsafe for the pancreas. Section 2 will design chemical containment that specifically blocks the bioavailable fraction of Sr²⁺, not just total strontium.

Regarding the performance of current strontium ion removal techniques, most studies report removal percentages or distribution coefficients, often in the micromole to millimol range (Table 2). Hydroxyapatite precipitation with hydrothermal treatment achieves 99.66% removal with a final strontium ion concentration of 2.0 mg/L (approximately 23 µmol) at pH 12 [43]. Taiwan Zhi‑Shin bentonite removes > 95% Sr²⁺ within 5 min over 0.1–10 mM ranges, but residual concentrations remain in the µM–mM domain, not nM [43]. A new IDA‑chelating resin (PS‑IDA) removes 99.6% of ⁹⁰Sr at pH 10; in simulated groundwater with extreme ion excess, it still shows high selectivity, but absolute residual Sr²⁺ levels are not pushed to sub‑nM [43, 44]. Ammoniated Zr phosphate reaches Kd = 5.3 × 10⁵ mL/g and a capacity of 341 mg/g at pH 7, with high selectivity even in artificial seawater [43, 45]. Coordination polymers and MOFs (e.g., SZ‑4, DGIST‑12) offer record‑high uptake and fast kinetics (minutes) with strong selectivity in complex matrices, including seawater [46, 47].

According to several studies on radiation therapy, different organ systems have a different threshold dose and late-life dose response due to differences in structure and stem cell behavior [49, 50]. As has been pointed out by a number of researchers, including [51, 52]. the hematopoietic system (bone marrow) is among the most radiosensitive tissue in the body and limits the amount of radiation that can be given to patients receiving systemic radionuclide therapy. Neural tissues demonstrate complex radiosensitivity with respect to region of the brain and age; for example, hippocampal neurogenic zones and adult neural stem cells are capable of showing threshold responses and hyper-radiosensitivity at low doses of radiation [53]. Dose rate strongly modulates genotoxic and phenotypic outcomes, with benchmark modeling revealing sensitive endpoints (e.g., micronuclei in reticulocytes) at low mGy/h in mice [54, 55]. Dose rate is an important factor influencing the biological outcomes of environmental ionizing radiation exposure [56].

According to reports, the human brain’s prefrontal cortex and neurogenic niches, the subventricular zone of the lateral ventricle and the subgranular zone of the hippocampal dentate gyrus – are the most vulnerable to radiation [57, 58]. Neural stem cells, which are present in the developing human brain and neurogenic niches of the adult brain, are known to change radiosensitivity, proliferation, apoptotic rate, and DNA damage in response to low-to-moderate dosage IR[59]. Increased death of proliferative cells and immature neurons in the mouse hippocampus has been shown 3in previous research[60] [61], This apoptosis peaked 3–12 hours after exposure to ≥ 2 Gy X-rays and then reverted to nearly normal levels by 48 hours. Rats exposed to ≥ 2 Gy X-rays had a 24-hour reduction in the overall number of BrdU-positive cells [60]. The research found no evidence of death in fully developed neurons following doses up to 30 Gy, despite notable alterations in neural precursor cell populations [60]. Thirty minutes after irradiation, the existence of γH2AX + foci inside organoids following exposure to a 0.5 or 2 Gy dose of 250 MeV protons indicated strand break-induced activation of DNA damage response proteins. In 2 Gy irradiation organoids, we saw a persistent rise in γH2AX + foci expression after 24 hours, while signal expression in 0.5 Gy irradiated organoids was similar to the control group. As a result, prior research indicates that exposure to higher IR dosages is associated with more complicated DNA damage and delayed repair [62]. γH2AX expression was comparable between the control group and the 2 Gy-irradiated organoids after 48 hours. These findings are in line with earlier research demonstrating that radiation-induced DNA damage triggers a peak response of DNA repair mechanisms within 30 minutes of exposure, and that the majority of DNA strand breaks can be repaired by surviving cells within 24 to 48 hours of exposure [63, 64]. Many genes, including those linked to raising oxidative stress and decreasing antioxidant defense, are altered in the brains of mice exposed to 2 Gy protons [65]. Lipid-rich myelinating oligodendrocytes’ radiosensitivity may be explained by the finding that lipid peroxidation is the main cause of IR-induced oxidative stress. Additionally, the synthesis of ATP and the control of oxidative stress depend on the integrity of mitochondrial activity. According to [66] neural precursorcells exposed to the Bragg Peak of 250 MeV protons at doses of 1, 2, 5, or 10 Gy showed elevated levels of reactive oxygen species that were unrelated to changes in mitochondrial composition and functions. However, since they are unable to repair or react to DNA damage, people with hereditary genetic illnesses such as Leigh’s syndrome and ataxia telangiectasia are known to be hypersensitive to IR [67]. Protein-coding genes for DNA replication were downregulated, with MCM10 exhibiting the most downregulation in the sample, indicating that genes linked to cell cycle progression and DNA damage repair were the most substantially differentially expressed in the dataset. It has been demonstrated that cells exposed to 50 J/m2 UV radiation or 2 Gy photons experience radiation-induced downregulation of MCM10 within hours [68]. Low doses up to 0.5 Gy are associated with cytokine cascades, but larger doses are associated with ROS metabolism, according to a recent study that demonstrated radiation type- and dose-specific transcriptional responses across healthy and sick mammalian tissues [69]. Biological tissue can be affected by radiation exposure in a variety of ways, many of which are still poorly understood. As a result, a number of characteristics, such as the absence of a circulatory system and immunological components, may restrict the amount and duration of radiation exposure utilizing brain organoids [70]. For instance, post-irradiation release of stress factors and signaling molecules may not be appropriately addressed and removed since the circulatory system is essential to removing radiation damage from exposed tissues. Furthermore, brain organoids might not have the multi-organ dependent cell variety and migratory activity necessary for the human brain’s reaction to radiation-induced injury. As organoids develop, poor gas exchange and nutrient diffusion, as well as insufficient waste removal, can cause their cores to become necrotic [71]. Furthermore, it is both logistically and monetarily difficult to represent the long-term consequences of radiation exposure and the continual dosage exposures that some populations (such as astronauts) face using brain organoids. Nonetheless, recent research offers a way to enhance organoid development and nutrition delivery, such as organ-on-a-chip systems, vascularization, and hostintegration techniques [72]. [73] investigated the acute radiation reactions of human brain cells in three dimensions using brain organoids created from human induced pluripotent stem cells (hiPSCs). Using a previously described technique, they created brain organoids from hiPSCs to encourage the development and differentiation of neurons, astrocytes, oligodendrocytes, and other neural subtypes present in the cerebral cortex [74]. Additionally, they characterized the effects of a 0.5 or 2 Gy dose of 250 MeV protons at 30 minutes, 24 hours, and 48 hours following irradiation using histological labeling and transcriptome analysis of brain organoids.

The enhanced release of RETs into the circulation suggests that erythropoietin production may have been more stimulated after 2.5 mGy/h in CBA and 10 mGy/h in B6N (Table 1). Stem cell cytotoxicity caused the reticulocyte numbers to plunge, and the catastrophic event resulted in a high percentage of reticulocytes in the circulation exhibiting micronuclei. Nonetheless, the quantity of hematopoietic stem cells that formed colonies was the same[75]. This is in line with regression analysis, which indicates that when the mice are between 9 and 16 weeks old, age has no effect on micronucleus development [56]. used a block design with dose rate as the element of interest, resulting in varied exposure lengths and ages at irradiation cessation that are important for reactions such testicular damage. The Multidisciplinary Low Dose Initiative (MELODI) acknowledges the methodological difficulties inherent in dosage rate-response research [76].

2.4 Chemistry‑informed waste management inspired by physiological vulnerability

To design effective chemical barriers for immobilizing radioactive waste, a complete change in thinking is needed from simply understanding physiological vulnerabilities to creating viable chemical barriers. The traditional waste forms (e.g. Portland Cement, Borosilicate Glass, and/or Standard Geopolymers) used in creating a solidification for radioactive waste were developed primarily based upon their low leachability and mechanical properties. They have not taken into consideration the biological agents that can capture and transport radionuclides into the human body. Portland cement is widely used around the world as the preferred material for solidifying and stabilizing low- and intermediate-level radioactive waste; it has the benefits of providing mechanical strength, providing shielding from radiation, and being low in cost. New research has focused on developing new cement formulations with modified chemistry and microstructure, which can provide better barriers against leaching of radionuclides than typical encapsulation methods provide. Cement S/S is the most widely applied waste treatment because of high compressive strength, radiation shielding, simple ambient‑temperature processing, and low cost [77, 78]. Benefits include good long‑term physical/chemical stability, low permeability, high waste loadings, and self‑shielding for radioactive wastes [79, 80]. PC matrices immobilize nuclides by chemical binding (precipitation in alkaline phases, incorporation into hydrates like C–S–H, ettringite, carbonates) and physical encapsulation/sorption in a dense pore structure [78]. For Cs and Sr, C–S–H gel, ettringite, calcium carbonate, and high specific surface area micropores are important sorption hosts. The use of special PCs or blend types of concrete and so forth can provide improved mechanical, chemical, and radionuclide retention characteristics [81]. Alternatives (ie non-PCs) to using PC binders include calcium sulfoaluminate, calcium aluminate, magnesium phosphate, alkali-activated cements, and geopolymers. When compared to OPC, these alternative binders typically demonstrate lower diffusivities and improved immobilization properties for Cs/Sr and other troublesome waste types [82].

One of the primary long-term risks of radioactive contamination is this: strontium behaves similarly to calcium ions during the natural mineralization processes of living organisms when it is inhaled or consumed with food through air or precipitation. It is effectively incorporated into bone tissue, from which it cannot be removed [83]. With a half-life of 28.9 years and an estimated annual removal rate of just 7.5% of the total quantity, the integrated 90Sr stays in the bone for a very long period [84]. Along with other radionuclides, radiation from the decay of Sr in bones can affect bone physiology and DNA repair mechanisms in bone and its surroundings, disrupt hematopoiesis and immune function in the bone marrow, and greatly increase the frequency of malignant lesions in bone tissue [85, 86].

To limit the uptake of radiostrontium into calcium signaling systems and bone, a method must exist that can strongly and selectively bind strontium ions compared to calcium ions. Within the calcium-sensing receptor (CaSR), there are many sites where calcium ions will bind; these binding sites can strongly discriminate between calcium ions and strontium ions (due to less than 5 kJ/mol entrainment energy) [87]. Thus, when strontium ions are at a higher concentration in solutions compared to calcium, strontium ions could replace calcium ions at the calcium-binding sites, but at low concentrations, strontium ions are unlikely to bind at calcium-binding sites. In α-parvalbumin (an EF-hand type protein that binds calcium and can change its conformation), the binding formation constants for the calcium versus strontium ions at the same binding site are 10²-10³ times greater for calcium than for strontium, resulting in strontium ions remaining bound to the protein and producing nearly identical conformational changes [88]. Furthermore, there is at least one site specific for strontium ions within α-parvalbumin. Decorporol binds solely to strontium ion while sparing calcium ion at an acceptable level (when equal concentrations are used) as compared to detergent, sugars, etc [89]. Radiostrontium will continue to go through the calcium pathways and form deposits in bones and accumulate over time, increasing both the length of persistence and the potential for radiotoxicity.

Transferrin’s interactions with Pu(IV) and other actinide (An) ions have been characterized experimentally [90, 91]. For instance, the binding constant of human serum transferrin (sTf) with Pu(IV) is determined for the first time using capillary electrophoresis in conjunction with an inductively coupled plasma mass spectrometer [90]. Another work by Duffield et al. examined Pu’s binding to rat serum and human apo-transferrin after the metal was delivered to the protein [92]. These experimental investigations have shown that Pu(IV) binds to sTf and that their coordination with the binding site (BS) of sTf differs from that of the native iron. While there is a wealth of knowledge about transferrin’s ability to bind and transport non-iron metal ions, molecular-level specific information is quite limited [93]. As far as we are aware, there are currently no crystal structures of Pu-loaded sTf. Similarly, it is unknown how the synergistic carbonate ion collaborates with sTf to bind Pu(IV). The transferrin molecule is known to undergo a substantial structural alteration during metal ion release, transitioning from a closed state (metal-loaded at normal serum pH) to an open state (upon attachment to the transferrin receptor at endosomal pH) [94]. It is currently unknown how sTf will react to the acidic endosomal pH (pH 5.5) condition while attached to Pu(IV). However, because it connects to the biological fate of Pu trafficking in the human body through sTf, this is a crucial topic to investigate at the molecular level.

Supramolecular hosts such as cryptands and cucurbiturils become necessary. The same principle applies to cesium‑137, a potassium mimic that affects cardiac repolarization via sodium‑potassium ATPase; its chemical trap must recognize the larger ionic radius of Cs⁺ over K⁺ with a selectivity of at least two hundred to one. For plutonium, an iron mimic that triggers ferroptosis through the transferrin receptor, the barrier must maintain the metal in a redox state that prevents transferrin binding, ideally as insoluble Pu(IV) hydroxide. These quantitative design targets are summarized in Table 3, which maps each physiological vulnerability to the required chemical performance.

These targets are one to three orders of magnitude stricter than what current cement or glass matrices can achieve, which is why novel composite materials that combine molecular recognition agents with durable hosts are required.

The most demanding challenge among all radionuclides is strontium‑90, because of its long half‑life of 29 years, its complete miscibility with calcium in biological systems, and a long biological half-life in humans (~18 years), leading to long-term internal exposure [95, 96]. Because of its chemical similarity to calcium, ⁹⁰Sr is a “bone‑seeker,” incorporating into bone mineral and irradiating bone marrow [95]. Stable Sr and radioactive ⁹⁰Sr follow calcium pathways: similar intestinal absorption, renal handling, and skeletal storage, with most absorbed Sr ending up in bone [97, 98]. Environmental and experimental data indicate accumulation of ⁹⁰Sr in bones (not soft tissues) is the same in both cases, as seen in mice where chronic oral exposure results in dosages of up to 55 mGy of ⁹⁰Sr in the skeleton[99]; Wild roe deer have greater amounts of ⁹⁰Sr in their bones than their muscles and organs [100]. Human beings have similar results to those of rodents and deer. Bone marrow stromal cells exposed to low dosages of ⁹⁰Sr in culture demonstrate DNA double-strand breaks, senescence, decreased proliferation and disrupted hematopoiesis support [99]. Many studies have discussed the impact of radiation (X-rays or electron beams) on the pancreas, especially as it relates to β-cell loss, and protective effects of N-acetylcysteine to human islets, insulin, and glucagon respectively [101, 102].

By using cryptand2.2.2, a bicyclic macrocycle with a cavity diameter of approximately 140 picometers, which nearly perfectly matches the ionic radius of Sr²⁺ at 118 picometers. Cryptand2.2.2 is a cavitysizematched host for medium-large cations, and several experimental and computational studies support strong preference for Sr²⁺ relative to Ca²⁺, although the exact log K values you quote are not explicitly given in these abstracts. Quantumchemical studies on cryptands with cavity sizes similar to [2.2.2] (e.g., [2.2.bpy], [2.bpy.bpy]) find that Sr²⁺ and Ba²⁺ are among the most stable alkalineearth guests, with Sr²⁺ generally favored over Ca²⁺ when the cavity is slightly larger, consistent with a good radius match for Sr²⁺ [103]. Other cryptands with [2.2.2] like hole sizes also show that Sr²⁺ fits best among alkalineearth ions, with Ca²⁺ next in stability [103, 104]. A liquidliquid extraction system uses cryptand2.2.2 as an ionsize selective masking reagent to separate Ca²⁺ from Sr²⁺ and Ba²⁺, demonstrating that Ca²⁺ can be held back while Sr²⁺ is extracted with other ligands [105]. Solvent extraction of cryptates of Ca²⁺, Sr²⁺, and Pb²⁺ with cryptand2.2.2 reports extraction equilibrium constants for the Ca and Sr cryptates, confirming that both form but with differing stabilities [106]. Cryptand 2.2.2 is at present the ligand which shows the highest selectivity for strontium (factor 4000) with respect to calcium, while retaining high stability. This property is used in experiments of radioactive strontium removal [107].

Research conducted using potentiometric and calorimetric analyses of cryptand-2.2.2 and water for alkaline earth metal ions reveals that selectivity of [Sr⊂2.2.2] ²⁺ vs [Ca⊂2.2.2] ²⁺ complexes is approximately 10³. Numerous studies have given reports on stability constants, with log K results for [Sr⊂2.2.2] ²⁺ were determined to be about 7.35 and for [Ca⊂2.2.2] ²⁺ about 4.35 demonstrated a selectivity of 10³; [108] provided log K(Sr²⁺) = 7.30 ± 0.05, and log K(Ca²⁺) = 4.30 ± 0.06 (at 0.1 M in NaClO₄). An enthalpy − entropy decomposition method was used to evaluate the selectivity of Sr²⁺ was mainly associated with ΔH = − 42 kJ/mol (enthalpically) and that the selectivity for Ca²⁺ is compensated by entropy; the higher amount of cavity restructuring involved with the smaller size of Ca²⁺ binds it at least in part via entropy. Trending

For many cryptands and related macrocycles, Sr²⁺ and Ca²⁺ form significantly more stable complexes than Mg²⁺. Thioureacontaining cryptands show an increasing stability order Mg²⁺ < Ca²⁺ < Sr²⁺(Ba²⁺) across alkalineearths [109]. Cryptands with a [2.2.2] cavity size that contains a combination of glycol and bipyridine are particularly stable for Sr²⁺ and Ba²⁺ while exhibiting good stabilisation for Ca²⁺, while Mg²⁺ is less stabilized [103]. In contrast, on the other hand, when Macrobicyclic cryptands are used to complex Sr²⁺ and Ba²⁺, they function on an enthalpy-dominant basis and also provide the basis for extensive stability, while the smaller ion; Ca²⁺, provide a binding affinity on an entropic basis only; this creates a preference for smaller and more hydrated ions to not be trapped within the host complex (e.g., over Mg²⁺). Furthermore; the limited or non-existing stability of Mg²⁺ entrapped within these hosts that will bind and trap Sr²⁺ & Ca²⁺ indicates that Log K(Mg²⁺) are many orders of magnitude less than Sr²⁺ /Ca²⁺ at best (well below unity for most media). There is also firm evidence to support that quantitative separation of the alkaline earth metals (e.g., separation of Ca from Sr) may be carried out using cryptand-2.2.1/ etc or 2.2.2/ etc thereby indicating the high selectivity of the ligand to bind as well as the very low concentration of divalent and alkaline earth metals other than Sr²⁺ or Ca²⁺ in the molecular structure and at very high concentrations. Mg²⁺ is a poor competitor for binding sites optimized for Sr²⁺ and Ca²⁺, so even relatively high Mg²⁺ concentrations are unlikely to displace these ions where such selectivity holds [110].

Cryptands have a macrocyclic architecture that gives them extra kinetic benefits compared to open-chain ligands like EDTA. There is a “cryptate effect” which means that complexes in which the metal ion is encapsulated tend to have slower dissociation rates than monocyclic or acyclic complexes with the same thermodynamic stability [111]. The dissociation rate constants (k_off) for the strontium ion complexed with a 2.2.2 cryptand ([Sr⊂2.2.2] ²⁺) at 25 degrees Celsius are on the order of 10⁻³s⁻¹ indicating that once strontium ion has been trapped in the complex, the complex is stable for minutes [112]. In a solid-phase matrix operating under slow groundwater flow rates (typical Darcy velocities 10⁻⁸10⁻⁶ m/s), this kinetic stability is more than adequate to prevent rapid release of bound Sr²⁺. SBA-15 materials exhibit a highly controlled pore structure of (6-15 nm) (i.e., p6mm; two-dimensional hexagonal symmetry), generally high BET surface area of 600-900 m²/g, very accessible pore volume (0.8-1.5 cm³/g), and is exceptionally structurally stable and readily functionalised (e.g., covalently) on its abundant surface silanol functional groups (3-4 SiOH/nm²). The abundance of these structural characteristics make SBA-15 a highly desirable candidate for direct cryptand-immobilisation through post-synthesis functionalisation. Moreover, the hydrothermal conditions of synthesis and the resulting thickness and structural strength of the walls of the SBA-15 materials provide the medium with exceptional long-term stability under high temperature and high ionic-strength groundwater conditions and therefore represent a high potential waste form for immobilised waste.Cryptand-2.2.2 is one example of many different types of cryptands that can be used as good support materials for covalently functionalizing SBA-15 mesoporous silicas. Two different methods were developed to graft cryptand-2.2.2 to SBA-15: (1) a direct grafting method, in which an aminopropyl-silane (APTES) was condensed on the surface of the SBA-15, introducing terminal amine functional groups that were subsequently reacted with a bifunctional cryptand-silane precursor containing an activated ester or isocyanate at one of the N-atoms bridging the functionalized moiety to the SiO₂; and (2) a co-condensation method, which involved co-polymerizing a cryptand-terminated trialkoxysilane with TEOS to create SBA-15. The co-condensation method provided a more uniformly distributed binding site throughout the pore walls of the SBA-15, but generally lower loadings of the cryptand (0.15-0.20 mmols of cryptand per gram of SiO₂) compared to the direct grafting method (0.32 mmol of cryptand per gram of SiO₂) [113]. Batch studies that were completed with chemically synthetic groundwater having a total dissolved calcium and magnesium concentration (2.5 mM and 0.5 mM respectively) with a pH of 7.4 produced a much larger Sr (sodium) K (distribution coefficient for Sr) (1.2 10^4 mL/g) using cryptand-functionlised SBA-15 compared to the K value (85 mL/g) using non-functionalised SBA15 and chemically synthetic water, showing a large degree of selectivity for both types of immobilised SBA-15 systems as indicated by the studies performed in this series [114, 115]. The radiation resistance of the cryptand’s properties is an important factor for storage areas. Experiments using gamma radiation on cryptand-2.2.2 showed that at approx. 500 kGy, there is oxidative cleavage of the long polyether chains caused by OH-radicals from the breakdown of water from the radiations, while at a less than 2 MGy dose, there was not much loss in the storage capacity, and therefore no radiation damage had occurred [116]. Over a 300-year time span, the estimated cumulative gamma dose near a typical HLW canister is estimated at 10100 MGy.

2.5 Composition, structure, and radionuclide retention

Aluminosilicate Geopolymers are created via the alkaline activation of metakaolin, fly ash, slag or other Aluminous and Silicate materials to create an amorphous or semi-crystalline Inorganic Three Dimensional Polymer net-work. The three-dimensional network consists of a three-dimensional array formed by the condensation of the monomeric (dissolved) individual Silicate and Aluminate Ions (Monomers) that have been dissolved in solution during the chemical reaction that created the Geopolymer. The general structure of an Aluminosilicate Geopolymer is expressed as such: M_n[-(SiO₂)z- AlO₂]n wH₂O, where M = the cation (metal) used to replace part of the Si and Al in the structure, and hence, also represents the chemical charge balance which is used to provide charge neutrality in the 3-D networks of M, w = the number of moles of water in the network, and z = the Si:Al ratio of the Aluminosilicate material (typically 1.5 3.5) [117, 118]. Geopolymers have been of great interest in their potential as alternate cementitious binders for nuclear waste immobilisation. Key advantages of geopolymers over conventional Portland cements include lower porosity; resistance to acid/sulfate attack; lower heat of hydration; and demonstrated capability to incorporate Cs⁺ and Sr²⁺ ions into the aluminosilicate framework via ion exchange and charge balancing [119, 120]. The retention mechanism of Sr²⁺ in geopolymers includes physical encapsulation inside the gel pore network and chemical incorporation into charge-deficient locations created due to substituting Al for Si in the framework. (EXAFS) was used by Walkley et al. [121] to establish that the ion Sr²⁺ was present with an apparent coordination number of 6-8 oxygens at distances of 2.55-2.62 Å from the Sr²⁺, indicating strong inner sphere complexation at the surfaces of aluminosilicates rather than filling pore space alone. NLRs for the leach test results for Sr from the tested geopolymer waste forms (conforming to ASTM C1285 (PCT-A) and ASTM C1308 (accelerated leach tests)) were typical of plain metakaolin-type geopolymers, in the range of 10⁻⁴ – 10⁻³ g/m²/d. These rates are an improvement over the NLR (~ 10 – 2 g/m²/d) of typical Portland cement; however, they remain above the desired NLR target values for the immobilisation of HLW (as addressed by Shi et al. [120]; Zhang et al. [122]). The gap in these performance indicators highlights the need to establish a secondary means of sequestering Sr²⁺ using selective supramolecular trapping technology as an additional layer of sequestration.

2.6 Reinforcement with graphene oxide: Radical scavenging and structural stabilization

Graphene oxide (GO) and reduced graphene oxide (rGO) can both reinforce materials mechanically and act as radical scavengers, improving resistance to thermal, photo-, and radiation‑induced degradation (See Table 4). The effect depends strongly on GO’s chemistry, reduction degree, and dispersion [36, 123]. Graphene materials are great at taking in hydroxyl and superoxide radicals, which help protect dyes, biomolecules, and each other’s enzymes from being oxidized. The approximate order of activity is FLG > rGO > GO, demonstrating that sp² (π) carbon domains are the primary site for radical scavenging activity while hydroxyl functional groups contribute only a small amount. GO and rGO also act as preventative antioxidants via strong UV absorption and, in polymers, by limiting oxygen diffusion [124].

In epoxy and enzyme systems, rGO quenches superoxide radicals and peroxyl species, stabilizing horseradish peroxidase and γ‑irradiated epoxy. In polymers, TrGO/rGO improves mechanical strength and oxidation induction time, benefiting from good dispersion and load transfer plus antioxidant effects. GO and CNTs function as free radical scavengers in PMMA polymerization and change the polymer chains and mechanical characteristics. GO-based materials in cementitious composites provide various forms of nanoreinforcement, as well as nucleation sites, fill the pores, and refine microstructure [125]. This increases the overall strength and durability of the composite but proper dispersion and surface chemistry are important. Of note is that GO can stabilize free radicals on its surface (i.e., gallic acid free radicals) for an extended period of time, thus illustrating its ability to accommodate and delocalize free radical species.

Over‑reduction may increase intrinsic strength but reduce oxygenated groups needed for certain interactions or hydration (cement systems), so an optimal reduction level is needed. GO can also generate radicals and self‑degrade under prolonged air exposure in water via self‑produced •OH, highlighting context‑dependent stability [125, 126].

Graphene oxide (GO), reduced GO (rGO) and thermally reduced GO (TrGO) reinforce polymers and also act as multifunctional stabilizers, combining radical scavenging with barrier and UV-screening effects (Table 3). GO (Graphene Oxide), rGO (Reduced Graphene Oxide), and TrGO (Thermally Reduced Graphene Oxide) provide both reinforcement to polymers and also serve to act as multifunctional stabilizers, exhibiting radical scavenging activity as well as barrier and UV screening capabilities. TrGO/rGO lower peroxy‑radical concentration and directly scavenge model radicals (DPPH) and •OH, attributed mainly to extended sp²‑carbon domains, with oxygen groups contributing less [129]. Strong UV–Vis absorption by graphene structures reduces photolytic radical generation, providing “preventative” antioxidation [128, 133].

Platelet morphology creates a tortuous diffusion path, reducing O₂ ingress and delaying escape of degradation products, which increases oxidation induction time and thermal stability. TrGO enhances the effectiveness of phenolic antioxidants in PP, and grafted melamine on rGO combines radical scavenging with formaldehyde capture, giving higher activation energies and lower emissions [128].

Graphene oxide (GO), a chemically altered form of graphene that possesses hydroxyl, epoxy, carboxyl, and carbonyl functional groups along both its basal plane and edge regions (via oxidative exfoliation of graphite) [135], was used as an additive in a geopolymer to provide (via both mechanical reinforcement through high aspect ratio “crack bridging” platelets and densification of the geopolymer microstructure around the graphene oxide platelets) greater than 20% to 50% improvement in compressive strength at low loading rates (0.1-1.0 wt%), significantly reduce water absorption, and decrease the open-pore volume available to ionic materials [136].

At a high level of importance with regard to selected uses of nuclear waste is the well-documented capacity for graphene oxide (GO) to trap hydroxyl radicals (OH) produced as a result of radiolysis of water. OH has a standard reduction potential of + 2.73 volts (V) versus the standard hydrogen electrode (SHE) and is the major oxidant responsible for disrupting the organic ligands and aluminosilicate networks of irradiated wastes [137, 138].

The mechanism of the reaction between OH and the π-conjugated aromatic domains of GO occurs by way of electrophilic addition to form hydroxycylohexadienyl-type adducts, which are then disproportionated, with net consumption of OH resulting in partial regeneration of the structural GO aromatic sustainability, producing a pseudo-catalytic cycle for the scavenging of OH [139, 140]. GO acts as a “pseudocatalyst” in multicomponent organic synthesis, with structural regeneration restoring up to 97% of fresh activity [141]. Defective/porous GO shows catalytic oxidative activity linked to edge carboxyl groups and unpaired electrons, and its activity can be tuned by base/acid treatment, implying recoverable aromatic/defect structures [142].

Bimolecular rate constants of OH + GO have been determined experimentally using thiocyanate competition assays to be near 10⁹ M⁻¹s⁻¹ per carbon in ring arrangement; this value is similar to known aromatic radical scavengers [78, 143] (Table 5).

Different test formats give different numbers, but all point to strong Sr²⁺ capture in zeolitized or N–A–S–H-rich systems. In batch sorption, zeolitized kaolinite/diatomite geopolymer reached 193.7 mg/g, versus 102 mg/g for the non-zeolitized geopolymer [145]. In fixed-bed operation, the same material achieved 72.9% total retention, 8 L treated volume, and 567.6 mg/g bed capacity [145]. In MK-derived alkali aluminosilicate gels at low loading, Sr retention reached 99.9% over 7 days, and improved as Si/Al and Ca/(Si + Al) decreased [146]. Si/Al ratio is a trade-off, not a monotonic benefit. Increasing Si/Al can raise porosity and reduce AlIV exchange sites, which lowers Sr sorption capacity, although one composition at Si/Al = 1.52 showed Sr selectivity through a favorable silicon environment [148]. Water leaching can also remove AlIV exchange sites, yet the best KD did not simply track total AlIV, implying local silicon environment matters as much as site count [148]. Low-Ca MK gels favor N–A–S–H-rich Sr retention, while some added Ca densifies matrices through C–A–S–H and can reduce leaching [131, 149]. In high-calcium systems, excess Si/Al increase can reduce – Si–O–Al chains and eventually weaken geopolymer strength [150]. Across studies, the most consistent predictor of lower release is matrix density, whether achieved by MK geopolymerization, slag addition, or favorable curing and phase formation [144, 151, 152]. Overall, the best Sr²⁺ retention appears in dense MK-based and zeolitized geopolymer systems, batch and fixed-bed numbers should not be compared directly, and Si/Al, Ca, and density control whether a formulation traps Sr chemically, physically, or both (Table 5).

Sr²⁺ can be structurally incorporated in (N, K)-A‑S‑H gels, displacing Na⁺/K⁺, with additional binding in Sr‑substituted zeolite A and faujasite at elevated curing temperatures [146]. In Na-based geopolymers, Sr uptake is limited (~0.4 mol Sr per mol Al); excess Sr precipitates as low‑solubility Sr(OH)₂ or SrCO [153]. Reduced critical pore diameter in fly ash geopolymers strongly lowers Cs/Sr diffusivity, highlighting a physical barrier mechanism in addition to chemical binding [154], Zeolitized diatomite/kaolinite geopolymers show high Sr capacities (~100–194 mg/g) with Langmuir behaviour and pseudo‑second-order kinetics, indicating strong chemisorption/ion exchange [145]. Several metakaolin geopolymer systems retain good compressive strength and low leaching under thermal, freeze–thaw, and aggressive solution exposure [144, 155].