Evidence of the Formation of Crystalline Aluminosilicate Phases in Glass-Ceramics by Calcination of Alkali-Brick Aggregates, Enabling Cs⁺, Rb⁺, Co²⁺, and Sr²⁺ Encapsulation

Abstract

The feasibility of using brick aggregates for the preparation of aluminosilicate “glass-ceramic” forms as a novel cementitious composite capable of immobilizing radioactive elements was examined. Raw brick was initially activated with sodium hydroxide. X-ray diffraction analysis (XRD) confirmed zeolites (Na-A and Na-P), illite, and sand (quartz) as major phases. Thermal analysis showed several successive events: dehydration/dehydroxylation of illite, followed by degradation of illite and zeolites. Upon heating to 1000 °C, scanning electron microscopy and XRD provided evidence of the presence of novel crystalline aluminosilicate forms (analcime and leucite in the form of solid solutions). Then, upon heating to 1150 °C, the thermal process led to the additional formation of mullite and an amorphous silica-rich phase. The latter resulted from silica melting taking place, owing to the involvement of low-melting-point components on sand grains. Alkali-brick particles were then doped with Cs⁺, Rb⁺, Ca²⁺, and Sr²⁺ ions (individually) and subsequently heated at different temperatures. The corrosion resistance of the heated materials was examined in a hydrochloride acid solution. The aim was to highlight (i) the enhanced cationic-immobilization capacity of crystalline aluminosilicate phases embedded inside amorphous silica, and (ii) the role of sand in the creation of brick-based glass ceramics.

1. Introduction

Owing to the carbon-free electricity production from nuclear power plants, they are becoming a serious alternative for curtailing global emissions of greenhouse gases in the atmosphere. However, during the process of the nuclear fuel cycle, a variety of radioactive wastes are generated. Nuclear wastes are known to contain rare-earth–transition metals and long-life actinide elements. These radio-nuclides can be accidentally discharged into the environment (soils, potable water sources, and the atmosphere) and, hence, over time become very dangerous to the biosphere and human health. In order to address the critical challenge of nuclear waste disposal, numerous radio-element immobilization strategies have recently been proposed in the literature [1,2,3,4,5]. The aim of these strategies is essentially to encapsulate hazardous radio-elements into physically and chemically stable matrices and to stock the resulting radioactive materials in deep geological disposal sites, thus allowing them to safely and slowly decay into stable elements. From this perspective, researchers have examined the physical barrier characteristics of various types of solid materials, such as cementitious materials [6], glass [7,8,9,10], ceramics [10,11,12,13,14,15], composite ceramics [16], and glass-ceramics [17,18,19,20,21,22]. Among them, glass-ceramics nowadays constitute the best materials for the immobilization of radio-elements [20,21,22]. This can be attribute to two main reasons. First, in glass ceramics, nuclear contaminants are bound tightly to the crystalline phase(s) encapsulated in glass matrices. Second, glass matrices behave as secondary barriers. These contribute to limiting any release of radio-nuclides in case the material comes into direct contact with aggressive media for a long time.

Nowadays, research on radio-nuclide pollution is more focused on nuclear wastewater treatments. In this regard, novel physical, chemical, and biological treatment-processes have to be tested with the goal of eliminating radioactive pollutants to upgrade water quality and safeguard public health. Thus, many remediation technologies, e.g., coagulation/flocculation [23,24,25], membrane technologies [24,26,27,28], precipitation [24,29], precipitation from molten salt electrolyte [30], electro-remediation [24], biological treatment and bio-remediation [24], ion exchange/resins [31,32,33,34,35], and adsorption [24,36,37,38,39,40], have previously been employed with the aim of removing radioactive elements from nuclear liquid wastes, and thus, of avoiding their spread into ecosystems. Among these methods, adsorption has been found to be the most effective, because it enables the generation of intermediate “radioactive solid wastes” that can then be transformed into more chemically stable materials for safe disposal. This was accomplished, e.g., by employing aluminosilicate adsorbents such as clay materials (bentonite) and industrial byproducts (fly ash) [41]. Their ability to fix radio-nuclides on their surface and to immobilize them thermally within stable aluminosilicate matrices has been well-evidenced experimentally. In addition, the generated aluminosilicate glasses possessed excellent fixation performance and could be further transformed into aluminosilicate mineral crystals through controlled processes.

Brick is a building material that is widely produced worldwide. It is considered a composite material that contains miscellaneous minerals such as quartz, kaolin/metakaolin, illite, montmorillonite, chlorite, and feldspars [42]. Its accessibility and low cost have motivated researchers to assess the possible use of brick aggregates as adsorbents (after coating with ferrihydrite) to remove inorganic and organic contaminants from wastewaters, e.g., (i) iron(II) present in anoxic (ferrous) groundwaters [43,44]; (ii) metallic pollutants such as Cu²⁺, Cd²⁺, Pb²⁺, and Zn²⁺ [45,46,47,48]; and (iii) endocrine-disrupting chemicals, such as bisphenol A (BPA) and 4-nonyl phenol (NP), as well as hormones (17α-ethynylestradiol, testosterone, and estrone) [49,50]. Moreover, alkali-activated brick was also tested as an ion exchanger. To that end, the alkaline activation of raw brick was undertaken thermally with sodium hydroxide, bringing about a series of chemical reactions (geo-polymerization, followed by zeolitization), and leading, with time, to the formation of zeolite-based composites [51,52]. In these composites, the presence of negative charges carried by [AlO₄]⁻ tetrahedra in zeolites enabled the binding of various cations such as alkali (Li, K, Na, Cs, Rb), transition–alkaline-earth (Sr), or transition metals (Co, Fe, Mn, Zn, Cu, Ni, Cd) [53,54,55,56]. However, as cationic species still remained chemically unfixed to the brick matrix in the form of hydrates/complexes, these could easily be desorbed, e.g., when the brick grains came into contact with concentrated NaCl and/or NaOH solutions [53,54]. Fortunately, our recent work has revealed the possibility of the effective immobilization of adsorbed cationic species by thermal treatment of brick pellets [57]. Moreover, according to some authors [58,59], geo-polymer-zeolite composites may be converted progressively into aluminosilicate ceramic forms after exposure to elevated temperatures. As alkali-activated brick was assimilated to a geo-polymer/zeolites/sand composite [51,52], it would not be surprising if similar thermal transformations occurred with such a material, resulting in further enhancement of the ability to immobilize radio-elements.

Owing to the beneficial characteristics of brick material (thermal and physical stability, easy processing) and its negligible environmental impact, it may offer favorable conditions for the safe disposal of radio-elements derived from radioactive nuclear liquid wastes under thermal treatment [52,57]. However, in order to achieve more practical applications, it was necessary to improve and optimize existing immobilization methods. In particular, the influence of the calcination temperature on brick activated by sodium hydroxide was still unclear. Hence, related studies needed to be conducted with the aim of better understanding the chemical/structural evolution of alkali-brick with temperature and the effects of the obtained matrix on the extent of adsorbate stabilization. From these perspectives, there was an urgent need to clarify the mechanisms involved after each thermal step.

In the present work, we explored the possibility of using a type of brick composed mainly of sand, metakaolin, and illite to synthesize ceramized-brick materials. We sought to prepare these materials under simple operational conditions and to achieve suitable radio-nuclide immobilization performance and long-term stability. To accomplish this, the alkali activation of raw brick with NaOH had to be undertaken with the aim of generating zeolites Na-A and Na-P through metakaolin zeolitization [52,57]. The presence of zeolites in alkali-brick (having crystalline aluminosilicate materials with Si and Al tetrahedrons and counter cations (Na⁺) in extra-frameworks) would make it possible to increase the ion-exchange properties of the composite, and hence, to enable the fixation of radioactive cations such as Cs⁺, Rb⁺, Co²⁺, or Sr²⁺ to the negatively charged sites present in zeolitic frameworks.

Alkali-activated brick pellets were heated at the following temperatures: 90 °C, 400 °C, 600 °C, 800 °C, 1000 °C, 1150 °C. The mineralogical, crystalline, textural, and morphological properties of heated samples were examined by means of the following techniques: X-ray powder Diffraction (XRD); Environmental Scanning Electron Microscopy (ESEM) equipped with an Energy Dispersive X-Ray Spectrometer (EDS); and Thermo-Gravimetric Analysis (TGA)/Differential Scanning Calorimetry (DSC).

The acid corrosion resistance of alkali-brick doped with Cs⁺, Rb⁺, Co²⁺, and Sr²⁺ (individually) and subsequently heated at selected temperatures was examined under aggressive attack conditions by conducting leaching experiments in hydrochloride acid solution. Leaching solutions were analyzed by either ICP-OES or ICP-MS.

2. Materials and Methods

2.1. Preparation of Brick Materials

The brick used was made in Bangui (Central African Republic) from kaolinite/illite-rich soils. It was prepared manually by mixing extracted soils and water. The obtained mud was air-dried for 48 h and then heat-treated with dry wood for a period of about 3 days at temperatures ranging from 500 °C to 800 °C. The heated material was then cooled progressively to ambient temperature for 3 days. The obtained brick was broken manually into grains using a hammer. Brick particles were sieved with mechanical sieves, and the fraction of particles with sizes varying from 0.7 to 1.0 mm was kept for our experiments. This fraction was washed with Milli-Q water and then decanted; after settling, water was eliminated, and brick grains were dried at 105 °C.

In the fabricated brick, quartz (60–65 w%), metakaolinite (20–26 w%) and illite (9–11 w%) were the major minerals, while iron oxide/hydroxide (2–3 w%) and titanium dioxide (1–3 w%) were minor minerals (Gildas Doyemet, Thesis in preparation). Alkali-brick synthesis was initially optimized by our research group [53]. Briefly, brick pellets with size diameters ranging from 0.7 to 1.0 mm were treated in different alkaline solutions (NaOH concentration ranging from 0.10 to 1.50 mol·L⁻¹) at different temperatures (50 °C, 70 °C, and 90 °C) and for a constant reaction time of six days. The recovered materials were then attacked with HNO₃ (2 mol·L⁻¹), and the resulting solutions were analyzed by ICP-OES. Investigations revealed that the quantity of sodium bound to the brick reached a maximum when alkali-brick synthesis was performed with a NaOH concentration of 0.6 mol·L⁻¹, knowing that the quantity of Na⁺ ‘extra-frameworks’ ions was considered as a key indicator of the number of reactive sites in the brick for cationic adsorption [53]. This explained why we decided to work here under the following optimized synthesis conditions.

Brick grains having 0.7–1.0 mm sizes were treated with sodium hydroxide as follows [57]: 10 g of Bangui brick was reacted in 40 mL of a diluted NaOH solution (0.6 mol·L⁻¹) at room temperature for 24 h under slow shaking at a speed of 120 rpm. The suspension was then treated at 90 °C for 6 days. The recovered grains were finally rinsed several times with Milli-Q water and dried at 90 °C for 24 h.

Alkali-brick samples doped with ‘M^+/++’ cations (Cs⁺, Rb⁺, Co²⁺, or Sr²⁺) were prepared as follows. All the experiments were carried out at the natural pH of the initially measured suspension, i.e., 7.74 for Cs⁺, 7.42 for Rb⁺, 7.37 for Sr²⁺, and 7.29 for Co²⁺. Experiments were conducted at room temperature. Preliminarily, synthetic solutions which were employed for ‘M^+/++’ adsorption onto brick were prepared by diluting 1000 mg·L⁻¹ stock solutions of ‘M^+/++’ cations with distilled water. The ‘M^+/++’ solution of 100 mL (2 mmol·L⁻¹) was put into contact with 0.5 g of alkali-brick grains in a 200 mL glass reactor until equilibrium state was attained (~240 mn). Then, the suspension was vacuum filtered through a 0.45 µm filter membrane. The recovered grains were rinsed several times with Milli-Q water and dried at 90 °C for 24 h.

Brick samples were afterwards heated at the following temperatures: 90 °C, 400 °C, 600 °C, 800 °C, 1000 °C, and 1150 °C. For each heated-brick sample, chemical durability/stability was evaluated by conducting leaching experiments in hydrochloride acid solution (1 mol·L⁻¹).

It is worth noting that in recent works, we examined the diffusional adsorption kinetics of cations Cs⁺, Rb⁺, Co²⁺, and Sr²⁺ (in single systems) in an attempt to identify the various determining factors controlling the adsorption of cationic elements onto alkali brick pellets and regulating the interfacial diffusion kinetics [57]. Mathematical treatments of kinetics (batch) data allowed us to define a complex multi-step process with distinctive diffusion regimes that were linked mainly to the pore-size distribution of the material. Under dynamic (fixed-bed column) conditions, we also evaluated the adsorption capacity of alkali-brick for Cs⁺, Rb⁺, Co²⁺, and Sr²⁺ ions; we found 10.2, 10.5, 5.5, and 18 mg of adsorbed element per gram of brick, respectively [57].

2.2. Chemicals

All chemicals employed in the experiments were of analytical grade. Hydrochloride acid, sodium hydroxide, Co(NO₃)₂·6H₂O, Sr(NO₃)₂, Cs(NO₃), and Rb(NO₃) were supplied by DISLAB (Paris, France). Only salts of non-radioactive isotopes were employed as model substances for adsorption experiments.

2.3. ICP-OES and ICP-MS Analyses

During leaching experiments, recovered solutions were analyzed using an inductively coupled plasma optical emission spectrometer (ICP–OES, model 5110 VDV, Agilent Technologies, Paris, France). As for cesium, because of the low emission intensity detected by ICP-OES, the Cs contents in the recovered solutions were determined by means of an ICP-MS spectrometer (Agilent Technologies model 7850, Paris, France).

2.4. X-Ray Diffraction (XRD)

XRD analysis was conducted at room temperature with a Bruker D8 Advance diffractometer (Bruker, Evry, France) using Ni-filtered CuKα radiation (40 kV, 40 mA). Brick samples were scanned with a step size of 0.02° and a counting time of 0.5 s per step.

2.5. Electron Microscopy

Micrographs of representative specimens of brick materials were recorded using an environmental scanning electron microscope equipped with an Energy-Dispersive X-Ray Spectrometer EDS X flash 3001 (Thermo Fisher Scientific, Courtaboeuf, France). EDS measurements were carried out at 20 kV at low vacuum (1.00 Torr), and the maximum pulse throughput was 20 kcps. Different surface areas ranging from 0.12 to 0.50 mm² were targeted on brick grains and examined by ESEM/EDS. For this, a narrow beam scanned selected areas of brick pellets for chemical analysis. Atomic EDS quantifications and mathematical treatments were undertaken by using “ESPRIP FAMILY” software (Bruker) in order to determine the average elemental composition of the surface brick and to detect chemical/elemental variability.

2.6. Nitrogen Adsorption–Desorption Isotherms

Textural characterization of activated-brick samples was determined from the N₂ adsorption desorption isotherm at 77 K using the micrometrics model Tristar II 3820 (Micromeritics Instrument Corporation, Norcross, GA, USA). Before analysis, the brick material was heated at 373 K and degassed for 2 h under primary vacuum conditions (pressure: 10 Pa).

The specific superficial area was assessed using the Brunauer–Emmet–Teller (BET) method. The pore volume was estimated using the α-plot method. The total pore volume was evaluated from the desorption isotherm branch at P/Po = 0.98, assuming complete pore saturation. Pore-size distribution was calculated using non-local density functional theory (NLDFT) and the Berrett–Joyner–Halenda (BJH) method.

2.7. Thermal (TGA and DSC) Analyses

All thermal analyses of alkali-activated brick were performed on fine powders (≤73 µm). These were heated to and from the desired temperature in the range 20–1350 °C. Thermo-Gravimetric Analysis (TGA) and Differential Scanning Calorimetry (DSC) were carried out on a Mettler-Toledo instrument (model ATG/DSC 3+, Greifensee, Switzerland). The DSC experiment was done at a ramp rate of 10 °C·min⁻¹ under a static air atmosphere. The TGA experiment was done at a ramp rate of 10 °C·min⁻¹ under flowing air.

2.8. Corrosion-Resistance Experiments

Degradation of heated-brick samples under hydrochloride acid attack was assessed at room temperature by dynamic immersion of brick grains (0.4 g) into a 1 mol·L⁻¹-HCl solution (25 mL). Each mixture was shaken vigorously for 24 h in a thermostatic orbital shaker. Afterwards, the suspension was vacuum filtered through a 0.45 μm filter membrane. In order to determine the leaching concentrations of brick elements (Al, Si, Na) and adsorbed elements (Cs, Rb, Co and Sr), each collected leachate was analyzed using ICP-OES for Al, Si, Na, Rb, Co, and Sr, and ICP-MS for Cs.

3. Results

3.1. XRD Analysis

Minerals were identified on the basis of ‘2θ’ reflection angles and “hkl” Miller indices (“hkl” indices are given in the parenthesis).

Quartz was as a major crystalline phase in the raw brick. This mineral was also detected in the XRD diffractogram of alkali-activated brick (Figure 1).

The characteristic peaks of quartz were identified: 20.9° (100), 26.6° (011), 36.5° (110), 39.5° (102), 40.3° (111), 42.4° (200), 45.8° (201), and 50.1° (112) (ICSD Collection Code: 89276). Compared to raw brick, the intensity of the quartz peaks did not seem to be affected by alkaline treatment because of the relatively strong chemical stability of quartz under our alkaline operational conditions. Illite, and to a lesser extent, rutile were present in the raw brick. Their characteristic XRD peaks were observed in the CuK_α1 X-ray powder patterns of alkali-brick: illite, 8.8° (001), 17.9° (004), 19.8° (021), and 34.3° (034) (ICDD (International Centre for Diffraction data): 00-009-0343); and rutile, 27.4° (110) and 36.1° (101) (ICSD Collection Code: 168140). Like quartz, illite and rutile were also found to be chemically stable under our alkaline operational conditions, thus allowing the observation of their reflections in the XRD diffractogram of the alkali-activated brick (Figure 1). Alkaline treatment of raw brick led to additional XRD peaks which were attributed to zeolitic crystals. Indeed, the positions of these peaks corresponded to those previously observed for zeolites Na-A and Na-P [60]; see Figure 1. The characteristic peaks of Na-A and Na-P zeolites were observed: Na-A crystals, 7.2°, 10.2°, 12.5°, 16.1, 21.7°, 24.0°, and 30.0°, which correspond to lattice plans of (200), (220), (222), (420), (600, 442), (622), and (644, 820), respectively [60]; and Na-P crystals, 12.5°, 17.7°, 21.7°, and 33.4°, which correspond to the lattice plans of (101, 110), (200, 002), (211, 112, 121), and (132, 123, 231, 213, 312, 321), respectively [60].

In the CuK_α1 X-ray powder patterns of the alkali-activated brick obtained after heating at 400 °C, and subsequently at 600 °C and 800 °C (each thermal step performed for 1 day), reflections characteristic of brick zeolites disappeared progressively (Figure 1). This signal disappearance might have been due to the dehydration of zeolitic micro-crystals, resulting in surface modifications. Indeed, changes in zeolitic surfaces would prevent X-rays from diffracting normally in the lattice planes.

After heating at 1000 °C for 24 h, characteristic reflections of brick illite disappeared completely in the CuK_α1 X-ray powder patterns of cured brick (see Figure 1), highlighting the thermal instability of illite at this temperature. Moreover, additional reflections were observed in the XRD diffractogram (Figure 1), giving evidence for the formation of cristobalite (SiO₂) and mullite ((Al₂(A1_2+2xSi_2-2x)O_10-x), where x ranges between 0.17 and 0.57 [61]), and weakly crystalline forms of MAlSi₂0₆-type phases with M = Na⁺, K⁺ (analcime, NaAlSi₂O₆, and leucite, KAlSi₂O₆). All these minerals were identified by some of their characteristic reflections (Bragg angles 2θ°, and Miller indices hkl): (i) cristobalite, 22° (101) (RRUFF ID: R060648; R061064); (ii) analcime/leucite, 25.9° (004, 400), 33.3° (431, 413, 314), 35.8° (521, 215) (RRUFF ID: R060023 for analcime, and RRUFF ID: R040107, R060300 for leucite); and mullite, 33.3° (220), 35.4° (111), 37.1–37.6° (130, 310), 41° (121), 43° (230, 320) (RRUFF ID: R141101; R141103). It is worth noting that the XRD findings agreed well with the TGA and DSC data (see Section 3.2) and those obtained by Carroll and co-workers [62]; these data were also consistent with our ESEM/EDS analyses (revealing morphological, textural, and chemical modifications of calcined alkali-brick; see Section 3.3).

After heating alkali-activated brick at 1150 °C for 24 h, the CuK_α1 X-ray powder patterns of the cured brick showed higher peak intensities for mullite and cristobalite, while those observed for analcime-leucite solid solutions seemed to be less intense (Figure 1). Such observations would suggest that temperature elevation favored the generation of mullite and cristobalite. Moreover, a broad background spreading from 17° to 34° and centered at around 24–25° was observed. The occurrence of this broad background, which already appeared weakly at 1000 °C, was found to be intimately linked to chemical modifications of zeolites (Na-A and Na-P) and sand, both producing amorphous (glassy) materials (see Section 3.2 and Section 3.3).

3.2. Thermo-Gravimetric Analysis (TGA) and Differential Scanning Calorimetry (DSC)

Thermal properties of alkali-activated brick were explored using TGA/DSC techniques, as depicted in Figure 2.

The TGA curve displayed multiple mass losses throughout successive temperature ranges. As the brick material was subjected to increasing temperatures, different observed thermal events occurred due to the physical and chemical processes taking place within brick framework. The initial ~1.2% weight loss took place in the temperature interval 25–150 °C (Figure 2). This weight loss resulted from the removal of surface water or water molecules which were weakly attached to the brick minerals. The second (weaker) weight loss, amounting to around 0.6%, was observed within the temperature range of 150 °C to 400 °C. This mass drop likely resulted from the removal or desorption of crystalline water molecules which were enclosed/trapped in the zeolitic and illitic brick frameworks. On the DSC graph, endothermic stages spreading from 80 °C to around 400 °C were thermodynamically representative of the physical processes of water extraction from brick minerals. Note that these two dehydration steps were found to be similar to those previously detected by TGA and DSC for illite [62] and zeolites Na-A and Na-P [63] by analyzing them separately. As the temperature increased, another weight drop (about 1.3 w%) was detected, and the TGA profile comprised: (i) a first step, starting at 400 °C and finishing at around 550 °C; and (ii) a second (very weak) step, spreading from ~550 °C to ~700 °C (Figure 2). The global profile was found to be comparable to that previously observed for Hungarian white illite [62]. The two TGA steps observed in the temperature interval 400–700 °C were attributed to the weight drop resulting from the dehydroxylation of illite. Indeed, in the 400–700 °C temperature range (at which illite is dehydroxylated), no weight loss was observed by TGA when analyzing quartz separately [64,65], zeolite Na-A [66] or zeolite Na-P [63] (given that all these compounds represent grossly brick minerals). Beyond 700 °C, TGA analysis revealed no significant weight loss (see Figure 2). This was expected, as no weight loss was detected in the different TGA curves of the major brick minerals if these were thermally analyzed individually, i.e., illite [62], zeolite Na-A [66], zeolite Na-P [63], and quartz [64,65]. Globally, a total weight loss of 3.1 w%, measured by TGA analysis, represented successive losses of water and illitic hydroxyls.

On the DSC graph (Figure 2), one may observe: (i) one sharp exothermic peak at 573 °C, corresponding to the α → β phase transition of quartz [67]; (ii) a broad DSC thermal profile in the 400–700 °C temperature interval (comparable to that observed by Carroll and co-workers [62]), caused by illite dehydroxylation; and (iii) beyond 700 °C a succession of endothermic stages showing that the investigated sample had undergone structural/chemical transformations and phase transitions accompanied by glass formation.

3.3. ESEM/EDS Studies

The use of the ESEM/EDS permitted us: (i) to microscopically analyze the brick before and after alkaline activation with the aim of determining the characteristics of the specimens; (ii) to follow their thermal transformation: and (iii) to quantify elements on particle surfaces.

3.3.1. Microscopic Analysis of Raw Brick

The raw brick was found to be mainly composed of silicon (Si), aluminum (Al), oxygen (O), potassium (K), and iron (Fe), and to a lesser extent, of titanium (Ti), magnesium (Mg), and calcium (Ca). In order to represent the spatial distribution of each major framework element (i.e., Si, Al, O, K, and Fe) in the ESEM/EDS mapping procedure, the following colors were used: blue, red, brown, violet and green, respectively (Figure S1). Element distribution images showed mostly Si/O-rich zones and Al/Si/O-rich zones, which were indicative of the presence of SiO₂ (as quartz) and aluminosilicates (metakaolinite and illite [68]). As for the Fe-rich zones shown in the Fe-distribution image in Figure S1C, they are indicative of the presence of iron oxide/hydroxides. Their amount represented around 2 w% of the mass of raw brick. Note that by heating at temperatures ≥ 1000 °C, the brick material became slightly brownish in color. This color change might have been due to the transformation of Fe-oxides/hydroxides phases into hematite during firing. Unfortunately, it was not possible to prove the occurrence of hematite in calcined brick by XRD diffraction because of its low content, and above all, the difficulty of detecting its main reflections (at 2θ = 33.20°, I% = 100%; and 35.68°, I% = 73.82 (RRUFF ID: R040024)), because hematite peaks were mostly masked by those of analcime/leucite and mullite. Note also the presence of the Ti-rich zones shown in the Ti-distribution image of Figure S1C, which is indicative of traces of titanium oxide. This finding was consistent with our XRD results, revealing reflections characteristic of rutile (TiO₂); see Figure 1.

3.3.2. Microscopic Analysis of Alkali-Activated Brick

After alkaline treatment of raw brick with sodium hydroxide at 90 °C for 6 days, ESEM analysis revealed that aluminosilicates in the treated brick were aggregates with either spherical or cubic shapes (Figure 3A,B).

Their sizes ranged from 8 μm to 20 μm for spherical particles and from 15 μm to 29 μm for cubic particles. The spatial distribution of framework elements Si, Al, O, and Na revealed a positive correlation between Al, Si, and Na in spherical and cubic specimens, giving evidence for the formation of sodic alumino-silicates as Na-zeolites (Figure 3C). This finding agreed well with XRD studies, revealing characteristic reflections of Na-A and Na-P zeolites (Figure 1). It is worth noting that the spherically shaped specimens observed here by ESEM/EDS were morphologically similar to those reported in the literature for zeolite Na-P [69,70,71,72], while those with cubic forms were comparable with those previously observed for A-type zeolite [73,74]. A detailed ESEM/EDS analysis was carried out on several cubic and spherical specimens that were targeted on alkali-activated brick samples with the aim of assessing the percentages of Al, Si, and Na present in the zeolitic (Na-A and Na-P) structures. Knowing that metakaolinite-derived zeolites should belong to the GIS family, the general chemical formula (based on the crystalline unit cell) of brick-zeolites could be expressed as (M)_x/m(Al)_x(Si)_16−x(O)₃₂·nH₂O, where m is the valence of alkali ions and n represents the number of water molecules per unit cell.

Next, we attempted to establish the chemical compositions of cubic and spherical specimens and compared them to those of dehydrated zeolites LTA (Na₉₅Al₉₆Si₉₆O₃₈₄) and Na-P (Na₆Al₆Si₁₀O₃₂), which were previously reported by Treacy and Higgins [60]. After averaging ESEM/EDS-analysis data and calculating standard deviations, we found:

(Na)_{93.9 ± 7.1}(?)_z(Al)_{96.8 ± 10.6}〈Si〉₉₆(O)_{390.9 ± 20.5}  (for cubic particles);

(Na)_{6.70 ± 0.92}(?)_z(Al)_{8.64 ± 0.57}〈Si〉₁₀(O)_{37.0 5 ± 1.05}  (for spherical particles).

In these formulas, the stoichiometric coefficients for Na, Al, and O were referenced to that for Si (as those reported for LTA (“Si₉₆”) and Na-P (“Si₁₀”) by Treacy and Higgins [60]), and the atomic deficit for sodium was represented by “(?)_z”. Indeed, normally the atomic deficit ought to be compensated for by one or more cations, like K⁺, Ca²⁺, and/or Mg²⁺, which were detected using ESEM/EDS at trace levels in the zeolitic frameworks. In the above formulas, one may note that the Si/Al atomic ratio was close to 1 for cubic specimens (0.99 ± 0.10, as expected for zeolite LTA), while that for the spherical specimens (1.16 ± 0.08) was somewhat lower than that expected for zeolite Na-P (i.e., 1.66). Also, in these formulas, an excess of oxygen (390.9 for cubic specimens and 37.05 for spherical ones) was observed when compared to those expected in the frameworks of LTA (i.e., “384”) and Na-P (i.e., “32”) because of traces of water molecules inside the analyzed zeolitic specimens. With this in mind, a mass loss of 0.6 w% (as observed by TGA) within the temperature range 150–400 °C resulted mainly from the removal or desorption of crystalline H₂O trapped in brick-zeolite frameworks; one could therefore roughly evaluate the number of water molecules in the predominating zeolitic structure of LTA zeolite. We found (H₂O)_18–21; this value differed somewhat from that proposed by Tracey and Higgins [60] for LTA, i.e., (Na)₉₅(Al)₉₆(Si)₉₆(O)₃₈₄(H₂O)₃₉, but was nevertheless close enough to that deduced from elemental ESM/EDS analysis of cubic particles (see the above formula), i.e., 390.9 (±20.5) −384 = 6.6–27.1, considering only superior limits.

On the other hand, the spatial K distribution image represented in Figure 3C indicates relatively high levels of potassium in certain regions of the targeted surfaces. The localization of K-rich zones suggested the presence of illite in the brick material, as already confirmed by XRD (Figure 1). To support this notion, we decided to target different potassium-rich zones in several brick grains and analyze them by employing ESEM/EDS microscopy. The amounts of major elements within alkali-brick (Si, Al, O, Na, and K) were determined for every pixels, constituting artificially targeted surfaces with the aim of reconstituting spatial element distributions, as illustrated in Figure 3B,C. Owing to such virtual reconstitutions, element-distribution images (Figure 3C) allowed us to better distinguish Si/Al/O/Na-rich zones and Al/Si/O/K-rich zones (Figure 3B) and to evidence the (co)existence of zeolites and illite in alkali brick.

The general formula of illite minerals is given by [75]:

(K)_1.0–1.5Al₄[Si_6.5–7.0Al_1.0–1.5O₂₀](OH)₄

(1)

In their 2:1 layer structure, the presence of negative charges, which are created by a part of illite-Al in tetrahedral sites, allows the occupancy to occur of monovalent cations (mainly K⁺) in interlayer sites. With the aim of appraising the chemical composition of brick illite, ESEM/EDS micro-analyses were carried out on several alkali-brick grains. By analyzing about 15 K-rich aggregates which were selected randomly on various brick zones, the atomic percentages of Si, Al, O, K, Na, and Fe were determined and averaged. Elemental analysis led to the following chemical expression:

(K)_{3.29 ± 0.36} (Na)_{1.09 ± 0.32} (Fe)_{1.45 ± 0.22} Al_{13.83 ± 1.36} Si_{16.73 ± 0.79} O_{61.10 ± 0.15}

(2)

As the following chemical formula K_1.5Al₄(Si_6.5Al_1.5)O₂₀(OH)₄ was that generally accepted in the literature for overall illite [76], and in order to simplify our stoichiometric calculations, we decided not to take into account Na and Fe in formulae determination. From the general formulae for illite minerals (Equation (1)), one may conclude that the negative charge in the atomic combination ‘Si_6.5Al_1.5’ was balanced with the positive charge in ‘K_1.5’. On the basis of this, one could obtain an averaged chemical composition of brick illite by comparing to the percentages of all elements given in Equation (2) to that of silicon. By arbitrarily fixing ‘Si_6.5’ in the proposed chemical equation, we obtain the following:

(K)_{1.28 ± 0.14} Al₄(〈Si〉_6.5Al_{1.37 ± 0.53})O₂₀(OH)_{3.74 ± 1.17}

(3)

This chemical expression was relatively consistent with that generally accepted for illite minerals (like that of Equation (1)). Note that, with the aim of compensating for the atomic deficit, we did not exclude the involvement of other elements, such as Mg and Fe, in complex illitic structures [76].

3.3.3. Microscopic Analysis of Calcined Alkali-Brick

Upon calcination of alkali-activated brick at 400 °C and 600 °C for 24 h, ESEM analysis revealed that cubic and spherical micro-particles of zeolites (Na-A and Na-P) retained the same sizes and shapes as those typically observed in starting alkali-brick (as shown in Figure 3A,B). In addition, quantitative ESEM/EDS analysis indicated no stoichiometric changes in silicon, aluminum, or oxygen, which represent the major structural elements of zeolites. Nevertheless, obtained ESEM/EDS results did not necessarily prove that chemical transformations of zeolitic aluminosilicate frameworks did not occur during the thermal process. Indeed, according to some authors [77], at moderate temperatures, cubic and spherical (zeolites) micro-crystals had to be considered as “micro-reactors” in which chemical/structural transformations, nucleation, and crystal growths remained mostly localized inside each reactive crystal (as already observed for cubic micro-crystals of zeolite NH₄A [77]). Therefore, not surprisingly, one would not expect to observe any stoichiometric changes in such ‘closed’ zeolitic systems, at least at temperatures of 400 °C and 600 °C. However, it is worth noting that additional percentages of potassium were measured as extra-framework cations on zeolitic particles after curing them at 600 °C, i.e., 0.19–1.24 at.% (instead of 0.09–0.38 at.% in unheated alkali-brick). K increases in zeolitic extra-frameworks seemed to coincide with the thermal phenomena observed in the temperature interval of 400–700 °C based on TGA and DSC; these thermal events were attributed to the dehydroxylation process of illite, as described in Figure 2.

Upon calcination at 800 °C for 24 h, ESEM images of heated alkali-brick displayed: (i) some ‘apparent’ preservations of the morphology/size/shape of zeolitic micro-crystals at this temperature stage; and (ii) in contrast, melted and degraded zones in which illite predominated (Figure S2). In the brick calcined at 800 °C, K percentages on zeolitic specimens were found to increase significantly when compared to those on alkali-brick cured at 600 °C, i.e., 1.10–2.69 at.% ( instead of 0.19–1.24 at.% at 600 °C). This observation suggested that a possible K transfer from a liquid (glass) phase generated at this temperature toward zeolitic specimens ought to occur in the cured material. This hypothesis seemed to agree well with our TGA and DSC analyses, which revealed thermal events corresponding to illite dehydroxylation, followed by the gradual transformation of dehydroxylated illite into aluminosilicate glass (containing potassium ions) at 800 °C. Such thermal decomposition of illite enabled K transfer into zeolites. To further support this mechanistic scheme, the spatial ESEM/EDS distribution of potassium on the brick surface was compared with those of major elements constituting the zeolite framework (i.e., Si, Al, O) and that of extra-framework (Na). The results clearly indicated a ‘net’ positive correlation between Al, Si, Na, and K (Figure S2C), showing the predominance of potassium in the brick zeolite. Meanwhile, there was a still negative correlation between Al, Si, Na, and K in the alkali-brick samples cured at 600 °C for the following reasons: (i) an absence or too weak degradation of illite; and (ii) a too low (24 h) reaction time (Figure 3C).

Upon calcination at 1000 °C for 24 h, ESEM analysis revealed some glassy aspects of cured alkali-brick (Figure 4A). By examining spatial ESEM/EDS element distributions in detail, one may note stronger K enrichments in aluminosilicate-rich regions (see Figure 4C) than those measured on zeolitic particles of alkali-brick heated at 800 °C (see Figure S2C).

Upon calcination at 1150 °C for 24 h, the ESEM micro-graph of heated alkali-brick represented in Figure 5 shows severe changes in the morphology of the brick material.

As can be seen in this figure, aluminosilicate minerals generated through the combination of illitic and zeolitic particles were heterogeneously dispersed in a glass matrix; both the glass regions and ceramic regions could be distinguished on the material surfaces.

Next, we considered two zones (cycles 1 and 2, shown in Figure 5), representative of glassy and ceramic states, which were targeted arbitrarily to evaluate their elemental composition. From the ESM/EDS data, we then attempted to establish a corresponding formula by taking silicon as the reference element. Thus, our ESEM/EDS analysis of cycle-1 region led to an elemental composition close to that of silica (SiO₂): 〈Si〉_1.000O_2.089Al_0.039Na_0.036K_0.009Mg_0.007. Indeed, this chemical formula corresponds to a silica glass containing low levels of both alumina and Na-K-Mg oxides: SiO₂0.020Al₂O₃0.018Na₂O0.004K₂O0.007MgO. As for the cycle-2 region containing small white inclusions, its ESEM/EDS analysis revealed an elemental composition: 〈Si〉_1.000O_3.382Al_0.830Na_0.161K_0.058Mg_0.027. This chemical formula corresponds to a glass material with the following stoichiometry: SiO₂0.415Al₂O₃0.081Na₂O0.029K₂O0.027MgO. By considering our XRD findings (that showed the existence of both analcime-leucite solid solutions ((Na⁺,K⁺)₂Al₂Si₄O₁₂) and mullite (3Al₂O₃2SiO₂) in cured brick), one may propose the following stoichiometric expression: 0.356SiO₂0.110(Na⁺,K⁺)₂Al₂Si₄O₁₂0.102(3Al₂O₃2SiO₂)0.027MgO. It is worth noting that, in the latter formula, one could not exclude the partial incorporation of MgO inside the mullite structure, as suggested previously [78].

From results reported by different authors [79,80] and those obtained here by combining XRD, TGA/DSC, and ESEM/EDS, it was possible to propose a mechanism for the thermal evolution of brick zeolites (Na-A and Na-P) at temperatures up to 1150 °C. Note that the contribution of illite to the thermal transformations of alkali-brick was not taken into consideration in the following chemical schemes. At rising temperatures (up to 1000 °C), the decomposition of brick zeolites should take place globally as follows (the chemical formula of zeolite LTA was simplified as 12x[M₈Al₈Si₈O₃₂]):

3/2[M₆Al₆Si₁₀O₃₂]_Na-P → 3MAlSi₂O₆ + 3 “Al₂O₃·2SiO₂” + 3SiO₂(amor.) + 3M₂O

(4)

3/2[M₈Al₈Si₈O₃₂]_LTA → 4MAlSi₂O₆ + 2“Al₂O₃·2SiO₂” + 2Al₂O₃ + 4M₂O

(5)

Al₂O₃ + SiO₂ → SiAl₂O_5(spinel)

(6)

where M represents an extra-framework mono-valent cation like Na⁺ or K⁺, or adsorbed cations like Cs⁺ and Rb⁺. Note that the formation of silica-containing alumina (γ-Al₂O₃) according to Equation (6) was not excluded, and is still under debate. In a mechanistic approach taking into account these XRD findings, one could ascertain that at temperatures up to 1000 °C, there was progressive formation of analcime-derived compounds (MAlSi₂O₆), ‘no-crystallized’ mullite (3Al₂O₃2SiO₂), and amorphous silica phase possibly containing silicate salts (MSiO₃). Meanwhile, at temperatures ≥ 1000 °C, the low crystallization of mullite became more effective, as observed by XRD and reported previously in Aza et al.’s works [80]. Small white grains embedded inside the glassy phase (see Figure 5) resembled mullitic forms, as described in the literature [81,82,83].

As for brick-illite, at temperatures of around 800 °C, this mineral was transformed intermediately into the amorphous dehydroxylated aluminosilicate phase, while at temperatures up to 1150 °C, dehydroxylated illite was degraded into spinel/transitional alumina (not detected here) and mullite [62]. With time, degraded-illite compounds were subsequently mixed with products derived from the decomposition of zeolite.

3.4. Implication of Fluxing Agent(s) in Thermal Process

A global ESEM/EDS analysis of several large surfaces (150 µm × 120 µm) of alkali-brick grains (chosen hazardously) revealed progressive increases of Si content at temperatures ≥ 1000 °C, while Al and Na underwent a dilution effect (Figure 6). The increase in silicon was due to silicate inputs provided by melted sand. Thiscould be explained by the fact that brick-sand was partially melted at temperatures lower than that expected for pure crystalline silica (quartz) by the involvement of low-melting-point components in contact with silica grains [84,85]. Indeed, as the structures of illite and zeolites collapsed thermally and transformed progressively into a glassy matrix, the material acquired an important fluxing action owing to large quantity of alkalis (Na and K), enabling the partial melting of silica during brick firing.

Under specific thermal conditions, it was possible to generate an additional amorphous glass phase by melting sand-silica. The significant rise of sand-silica glass was revealed by quantitative ESEM/EDS analysis of heated alkali-brick surfaces (see atomic % evolution of silicon in Figure 6) and was confirmed by the gradual increase of a broad XRD diffraction hump, detected between 2θ°: 17–34° for alkali-brick calcined at temperatures ≥ 1000 °C; see Figure 1.

In order to better highlight the implication of fluxing agents in the transformation of alkali-brick into glass-ceramic, it was important to preliminarily examine the chemical composition of alkali-brick in terms of oxides (see

Table 1. Chemical composition (in terms of oxides) of brick materials used in this work, determined via acid attack (with a mixture of HF + HNO₃ + HCl) by ICP-OES analysis (in w%).

Oxides	SiO2	Al2O3	Fe2O3	CaO	MgO	Na2O	K2O	TiO2
Raw brick	72.74 ± 1.56	8.72 ± 1.32	2.82 ± 0.03	<0.010	0.130 ± 0.033	0.138 ± 0.008	1.031 ± 0.021	1.554 ± 0.039
Activated-brick	69.87 ± 1.43	8.79 ± 0.21	2.82 ± 0.16	0.043 ± 0.025	0.138 ± 0.030	1.929 ± 0.057	1.071 ± 0.012	1.007 ± 0.591

).

The relatively high percentage of K₂O (1.03 w%) in the raw brick reflected the presence of illite, while that of Na₂O (1.93 w%) indicated the presence of Na-zeolites in the alkali-activated brick. The total percentage of alkaline oxides (K₂O + Na₂O) represented about 3 w% of the total brick mass, while that of alkaline-earth oxides (MgO + CaO) was no more than 0.2 w%. Alkaline and alkaline-earth oxides are known to act as flux materials during curing. It is, however, interesting to mention that the thermal behavior of the main fluxing agents in alkali-brick (i.e., K₂O and Na₂O) differed somewhat from each other. Indeed, potassium oxide is more capable of generating low-temperature melting eutectics with other minerals, while sodium oxide more effectively reduces viscosity [86].

In order to explain thermodynamically how the sand present in the brick could be melted at temperatures lower than it would be on its own, it was important to apply at best phase diagrams to the thermal events which occurred during alkali-brick firing. Indeed, phase diagrams provided evidence that minerals together could melt at lower temperatures than they would individually. This phenomenon, called “eutectic melting”, led to the creation of complex eutectic systems. In order to rationalize eutectic systems, we had to conceive more simplified thermodynamic approaches by elaborating eutectic diagrams in which only a limited number of minerals present in the studied material were taken into consideration. Therefore, we attempted to survey the thermal evolution of alkali-brick in the temperature range of 800–1150 °C. Knowing that the phase diagrams represented thermodynamic models that ought to involve main flux component(s), we decided to limit our thermodynamic investigations of the “leucite-quartz” system to 1 atmosphere. With this simplified approach, the leucite mineral which was generated through thermal illite-zeolite degradation was considered as the principle fluxing agent, while sand (a mineral largely in excess in the system) was considered as a relevant silica-glass source during the thermal process. By examining the phase diagram for the previously established KAlSi₂O₆ (leucite)-SiO₂ system in detail [87], one may note the eutectic point at just under 1000 °C. At the eutectic temperature, ”eutectic melting” occurred, triggering the melting of sand-silica (although individually, quartz minerals melt at much higher temperatures). Note that the additional mineral(s) present in the brick material could also participate in the lowering of the eutectic-melting temperature.

To summarize, our findings indicated that the K and Na atoms present as extra-framework cations in the alkali-brick were encapsulated in a ceramic phase at elevated temperatures (≥1000 °C). The same operational procedure was tested to encapsulate ‘radioactive’ mono-valent cations of Cs⁺ and Rb⁺ and divalent cations Co²⁺ and Sr²⁺ (considering that the studied cations were previously adsorbed onto alkali-activated brick from aqueous solution before proceeding to calcination). In order assess the degree to which the chemical durability of alkali-activated brick and the retention of zeolitic extra-framework cations were linked to the thermal experimental conditions, a series of leaching experiments were conducted in an aggressive acid medium. The results are discussed in the next paragraph.

3.5. Acid Corrosion Resistance of Brick Materials

Phenomena of degrading/melting of crystalline forms (illite, zeolites and sand) were observed in the present work at elevated temperatures. Owing to the combined use of XRD, TGA/DSC, and ESEM/EDS, a detailed thermal analysis evidenced the presence of a novel material in which both analcime-leucite solid solutions and mullite were heterogeneously distributed and enclosed inside a silica-rich matrix.

Therefore, we attempted to demonstrate the benefits of using alkali-brick as an adsorbent, and after calcination, as a chemically stable matrix for the immobilization of zeolitic extra-framework cations. To this end, it was necessary to know whether the chemical stability of heated alkali-brick was sufficient to consider it as a toxic-element immobilizer. We therefore decided to examine the corrosion resistance of the brick material by performing hydrochloride acid attack. A series of leaching tests were conducted under highly aggressive conditions by submerging 1 day calcined alkali-brick samples in an HCl solution (1 mol·L⁻¹). Chemical analyses of recovered solutions were then carried out by ICP-OES or ICP-MS to gain insights into the degradation state of the attacked material.

Before that, alkali-brick pellets were initially suspended in M^+/++ solution (M^+/++: Cs⁺, Rb⁺, Co²⁺, or Sr²⁺) to allow interfacial Na^+/M^+/++ exchanges to occur. The recovered pellets (i.e., M^+/++-doped alkali-brick) were then dried at different temperatures for 1 day: 90 °C; 600 °C, 800 °C, 1000 °C, and 1150 °C. For each heated-brick sample, a leaching test was conducted in 1 mol·L⁻¹-HCl solution for 1 day. As expected, because of the degradation of the zeolitic support under highly acidic conditions [52], silicon and aluminum (constituting zeolitic frameworks), sodium, and adsorbed elements (present in the extra frameworks) were found to be easily released from the samples treated at 90 °C (Figure 7).

This attack was further improved by the hierarchical porous characteristics of the alkali-activated brick (having micro-pores, meso-pores, and macro-pores [57]), enhancing the diffusion of aggressive ions from the immersion solution into brick aggregates. At temperatures of 600 °C and 800 °C, the effects of the acid attack diminished noticeably due to changes in the porous properties of heated brick (depletion of BET surface area and the BJH adsorption/desorption cumulative volume of pores; see

Table 2. Micro-porosity of alkali-activated brick after thermal treatment at 90 °C, 600 °C, 800 °C, 900 °C, and 1000 °C for 1 day.

Heating Temperature	90 °C	600 °C	800 °C	900 °C	1000 °C
BET surface area (SBET) in m2/g	19.319	14.421	4.254	0.801	0.002(6)
Single point adsorption total pore volume of pores (∅ < 313.549 nm at P/Po ≤ 0.9942) in cm3/g	0.0397	0.0353	0.0200	0.0031	0.0017 (at P/Po ≤ 0.2716)
BJH Adsorption cumulative volume of pores (1.700 nm ≤ ∅ ≤ 300.000 nm) in cm3/g	0.0363	0.0323	0.0205	0.0028	-
Adsorption average pore width (4 V/A by BET) in nm	8.22	9.78	18.80	15.53	-
BJH Adsorption average pore diameter (4 V/A) in nm	12.71	13.45	17.30	27.52	-
BJH Desorption average pore diameter (4 V/A) in nm	12.46	13.23	16.16	28.58	-

).

At temperatures ≥ 1000 °C, a significant decline in leaching was observed. The weak cationic leaching in the brick samples which were previously treated at elevated temperatures could be explained by the fact that illitic and zeolitic specimens were significantly degraded/transformed into novel crystalline minerals (especially into analcime-leucite and mullite) and incorporated inside a silica-rich matrix with a smooth surfaces and glassy morphology (as shown in Figure 4 and Figure 5). Furthermore, the loss of micro-porosity on the brick surfaces (Table 2) contributed to improving the acid-corrosion resistance of the material.

4. Conclusions

In the present work, a new approach was proposed to encapsulate mono- and divalent cations in glass ceramics derived from alkali-activated brick. Cations were initially bound to alkali-brick at room temperature, and the resulting material was heated at different temperatures and analyzed. The combined use of XRD, ESEM/EDS, and TGA/DSC revealed noticeable chemical, textural, and morphological transformations of alkali-brick. These transformations took place at successive temperature stages: first, at temperatures up to 400 °C, the dehydration of illite and zeolites occurred; second, at temperatures between 400 °C and 700 °C, the dehydroxylation of illite occurred, followed by its decomposition; third, at 800–1000 °C, the degradation of zeolites NaA and NaP into other aluminosilicates, identified as analcime-derived solid solutions and mullite, occurred; and fourth, at T ≥ 1000 °C, sand melted, contributing strongly to the formation of glass ceramics. Owing to the input of silica from the melted sand, the glassy-silica content increased in the heated brick, improving the chemical durability of the calcined material and assuring the coating of (thermally formed-) crystalline aluminosilicate minerals and radio-element encapsulation.

Moreover, our investigations allowed us to obtain information about the corrosion resistance of thermally treated-alkali-brick aggregates under highly aggressive acid conditions. This research provides evidence of the importance of the degradation and chemical transformation of illite and zeolites Na-A and Na-P and the melting of sand regarding the chemical durability of the material. Thus, leaching tests carried out on calcined-brick samples evidenced the efficient barrier properties of the glass matrix toward (trapped) crystalline aluminosilicate structures containing either mono-valent cations (Na⁺, K⁺, Cs⁺) or divalent cations (Co²⁺, Sr²⁺).

Globally, this work provides novel ideas for the encapsulation of radioactive elements from nuclear liquid waste by using brick aggregates as low-cost starting materials. The challenges and prospects of employing zeolite-geo-polymeric composites derived from brick will open novel avenues of research, i.e., exploring the development of these materials as industrial (low-cost) byproducts to address issues of radioactive contamination.

5. Future Prospects

Owing to the low Si/Al ratios of zeolitic-geo-polymeric forms of alkali-brick, these compounds contributed to high negative surface charges that facilitated ionic exchanges between Na⁺ ions at the solid surface and radioactive elements (Cs⁺, Rb⁺, Co²⁺, and Sr²⁺) in the aqueous phase. Nevertheless, both the amount of extra-framework cations (Na⁺) or “active” negatively charged sites and the brick porosity depended strongly upon the reaction alkalinity/temperature/time. A future objective will be to optimize the synthesis method to maximize the number of active sites and reactivity by testing other basic reagents (e.g., KOH, Mg(OH)₂ and Ca(OH)₂) and to produce more (novel) hierarchically porous materials and improve adsorption performance.

Compared to other glass-ceramic materials in the literature, calcined brick materials immersed in hydrochloride-acid solution possessed promising capacity due to their relatively low leaching degrees. However, glass ceramics produced from brick aggregates have to meet environmental safety specifications. In this respect, leaching toxicity analyses will be carried out on heated alkali-brick by our research group using Toxicity Characteristic Leaching Procedures (TCLP), according to international standard guidelines. By applying these procedures, leaching concentrations of elements will be determined and compared to standard leaching thresholds. Normalized elemental leaching rates will also be assessed. The obtained leaching rate data will be compared to those previously determined for other industrial byproducts that have been employed for glass-ceramic production and radioactive element immobilization.

In future research, exhaustive exploration of the micro-structure of alkali-activated brick and its evolution with temperature should be conducted using UV-Raman and ²⁹Si, ²⁷Al, and ²³Na MAS NMR spectroscopies. The use of NMR spectroscopy with an ultra-high magnetic field (1200MHz) would further allow us to efficiently separate and clearly resolve penta/hexa-coordinated alumina peak(s) from dominant tetrahedral peaks, as well as to correlate and quantify the signal intensities relative to Al(IV), Al(V), and Al(VI) configurations with structural transformations in brick frameworks during thermal treatments. All these studies are essential to highlight differences in the distribution of silicon and aluminum between Qⁿ units in cured brick materials. The combined use of UV-Raman and NMR techniques would also allow us to clarify the role of fluxing agents (alkaline and alkaline earth elements) as network modifiers and charge compensators in compositional and micro-structural modifications of aluminosilicate minerals in brick aggregates. These obtained data would improve scientific knowledge of the (geo)-chemistry of complex structures of natural magmas and industrial glasses.