Alkali-Activated Brick Aggregates as Industrial Valorized Wastes: Synthesis and Properties

Abstract

In recent works, many industrial by-products were employed as solid precursors for the synthesis of alkali-activated binders and as alternatives to Portland cement for the immobilization of hazardous, toxic and nuclear wastes. Among industrial wastes, alkali-activated brick was found to be an interesting porous composite for removing very toxic heavy metals (Pb²⁺, Cd²⁺, Co²⁺) and radio-nuclides (Sr²⁺, Cs⁺, Rb⁺) from aqueous solutions. The starting material is very attractive due to the presence of metakaolinite as a geo-polymer precursor and silica for increasing material permeability and facilitating water filtration. The alkaline reaction gave rise to geo-polymerization followed by partial zeolitization. Elemental surface micro-analysis was performed by Scanning Electron Microscopy (SEM) equipped with an Energy-Dispersive X-ray Spectrometer (EDS). The formation of crystalline phases was corroborated by X-ray diffraction (XRD) analysis. Information about ²⁹Si, ²⁷Al and ¹H nuclei environments in crystallized and amorphous aluminosilicates was obtained by ²⁹Si, ²⁷Al and ¹H MAS NMR. ²⁷Al–¹H dipolar-mediated correlations were investigated by employing dipolar hetero-nuclear multiple quantum coherence (D-HMQC) NMR, highlighting Al–O–H bonds in bridging hydroxyl groups (Si–OH–Al) that are at the origin of adsorptive properties. Aqueous structural stability and cationic immobilization characteristics before and after material calcination were investigated from acid-leaching experiments.

1. Introduction

Human activities, such as fossil fuel burning, deforestation and cement production, have contributed strongly to the excessive emission of carbon dioxide into the atmosphere and the acceleration of global-warming phenomena. In order to limit climate change, more efficient policy actions towards decarbonization must be undertaken urgently [1]. Among them, a better efficiency/control of extraction procedures and management of non-renewable natural resources and incentive re-using/recycling and valorising of industrial (by-)products have to be encouraged to diminish substantially environmental degradation and ensure environmental preservation and sustainability.

Portland cement is an industrial low-cost product that has been used for many years as a stabilization/solidification matrix for immobilizing low- and intermediate-level radioactive liquid wastes [2]. However, radio-nuclides are found to be generally difficult to stabilize within cement matrices [3], because these latter are prone to have structural and/or chemical degradations in contact with aggressive media or at high temperatures. The generation of micro-cracks in solidification matrices (concretes) may cause, with time, radioactive leaching and severe contamination in disposal sites, thereby preventing safe land disposal. To improve the physical and chemical stability of the matrix and limit radioactive leaching in aqueous aggressive media, cementitious materials, such as blast furnace slag, fly ash and metakaolin, have been proposed as promising alternatives to Portland cement for the immobilization of hazardous, toxic and nuclear radio-elements, such as cesium, strontium, cobalt and europium [4,5,6,7,8,9,10,11,12,13,14], and by employing them either alone or combined with other solid aluminosilicate precursors [3]. It is worth noting that, contrary to normal concrete, alternative cementing systems based on aluminosilicate polymers or geo-polymers are characterized by a lack of calcium and, thereby, would not suffer from effects such as carbonation or acid leaching after stabilization/solidification [15].

In order to enrich or separate radio-nuclides, the adsorption and ion exchange method remains the best (efficiently and economically) remediation technique due to a facile large-scale production, much lower operating costs and, above all, easier industrial operability. Many inorganic ions exchange materials and multi-functional core-shell adsorbents exhibit high adsorption capacity and excellent selectivity in the removal of nuclides from solutions [16,17]; but, these materials presented relevant disadvantages and difficulties in practical/industrial applications, such as (i) products costs; (ii) complicated synthesis processes consisted of two or more reaction steps, rending adsorbent preparation relatively costly and time-consuming; and mostly, (iii) small particulate sizes which are not suitable for carrying out large-scale water treatment (in dynamic adsorption) or separating powdered particles from the liquid phase. In contrast, the heterogeneous surface behavior of mixed materials (or composites) contributed to improving adsorption capacity, permeability and filtration, especially owing to the possibility of obtaining larger particle sizes which are more adapted to column experiments [18,19]. Recent studies thus revealed remarkable advantages of using geo-polymer-zeolite-A composites for immobilizing radioactive wastes [20]. Analogically with previous works on the synthesis of zeolite-geo-polymer composites using industrial solid wastes [21], a zeolite-geo-polymer composite was prepared in the lab by proceeding to an alkaline activation of a metakaolinite-rich brick.

The aim of this paper was to review the chemical, mineralogical and micro-structural properties of alkali-activated brick as a promising matrix for radio-nuclides immobilization in the perspective of large-scale applications.

In the first part of this work, a few comments were made about the different reaction steps of the “brick-composite” synthesis.

In the second part, chemical and mineralogical analyses of alkali-activated brick aggregates were performed by using an environmental scanning electron microscope (ESEM) equipped with an energy-dispersive X-ray spectrometer (EDS), and their crystalline characteristics were determined by using X-ray diffraction.

In the third part, the generation of crystallized and amorphous aluminosilicates was evidenced and quantified by means of ²⁹Si and ²⁷Al MAS NMR spectroscopy. Information about ¹H nuclei environments was obtained by ¹H MAS NMR. ²⁷Al–¹H dipolar-mediated correlations were investigated by employing dipolar hetero-nuclear multiple quantum coherence (D-HMQC) NMR with the aim of highlighting the occurrence of Al–O–H bonds and their implication in bridging hydroxyl groups (Si–OH–Al).

In the fourth part, the aqueous structural stability and cationic immobilization characteristics of the brick-based composite before and after its calcination were investigated from acid-leaching experiments and analytical data were discussed.

2. Materials and Methods

2.1. Materials Preparation and Making Cost

The raw material used in the experiments was obtained from a brick made locally by craftsmen in the Bangui region (Central African Republic). This brick contains mainly quartz (60–65 w%) and metakaolinite (20–26 w%) and, to a lesser extent, iron oxide/hydroxide, illite and titanium dioxide [22]. Before alkaline treatment, the brick was broken into grains and sieved with sizes ranging from 0.7 to 1.0 mm. Second, brick pellets were treated with sodium hydroxide according to the following optimized synthesis conditions [23]: 10 g of Bangui brick reacted in 40 mL of a diluted NaOH solution (0.6 mol·L⁻¹) at room temperature for one night under a slow shaking at a speed of 120 rpm. This procedure was afterward followed by a fixed-temperature increase of the mixture at 90 °C for a constant reaction time of six days. Finally, the recovered grains were rinsed several times with MilliQ water and dried at 90 °C for 24 h.

Making cost of synthesized composite: Briefly, a raw brick which was provided by African craftsmen (in the Central African Republic) and used here as starting material weighed about 5 kg, and its price is 75 CFA (i.e., 0.11 Euro; 1 CFA = ~0.0015 Euro); while solid sodium hydroxide which was provided by local soap factories and used here as reagent cost 30,000 CFA/25 kg (i.e., 45 Euros). Knowing that 4 liters of a NaOH solution at a concentration of ~0.6 mol·L⁻¹ were necessary to treat 1 kg of ground brick, the authors estimated the cost for synthesizing 1 kg of adsorbent at ca 0.20 Euro (without taking into account labor costs, which are very low in this country). Hence, the low making cost of the brick-based composite should enable its use as a good low-cost adsorbent for radioactive wastewater treatment.

As for the NH₄⁺-saturated alkali-brick, an ion-exchange reaction between Na⁺ ions in brick frameworks and NH₄⁺ ions in the aqueous phase was carried out by mixing 1 g of alkali-brick grains (diameter 0.7–1.0 mm) with 100mL of an NH₄Cl solution (2 mol·L⁻¹). The suspension was shaken gingerly for 24 h in a thermostatic orbital shaker. The resulting brick grains were afterward filtered, and the filtration operation was repeated three times with the aim of ensuring a total exchange of extra-framework sodium with ammonium. Finally, brick grains were rinsed several times with Milli-Q water and dried at 70 °C for 24 h.

2.2. Chemicals

All chemicals employed in the experiments were analytical grades. Sodium hydroxide, ammonium chloride and hydrochloride acid were supplied by DISLAB (Paris, France).

2.3. Electron Microscopy Analysis

Micrographs of representative specimens of the brick before and after chemical treatment were recorded by using an environmental scanning electron microscope equipped with an Energy Dispersive X-Ray Spectrometer EDS X flash 3001 (ThermoFisher 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 alkali-brick grains and examined by ESEM/EDS. For that, a narrow beam scanned selected areas of brick pellets for chemical analysis. Atomic quantifications and mathematical treatments were undertaken using QUANTA-400 software in order to determine the averaged elemental composition of the surface brick and to detect chemical/elemental variability.

2.4. X-ray Diffraction Analysis

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

2.5. MAS NMR Analysis

The ²⁹Si MAS-NMR experiments were acquired at 79.5 MHz on a Bruker 9.4 T spectrometer (Palaiseau, FR) equipped with a Bruker 7.0 mm MAS probe operating at a spinning frequency (ν_rot) of 4 kHz. The ²⁹Si MAS-NMR spectra were recorded with a pulse length of 5 µs (π/2 flip angle), 256 transients, and a recycle delay (rd) of 30 s. This short rd value was possible because of the presence of paramagnetic species (as iron(III) oxides/hydroxides) homogeneously distributed within the brick matrix. The ²⁹Si MAS-NMR spectra were decomposed using the Gaussian/Lorentzian model of the dmfit software (version 2.0) [24]. The ¹H, ²³Na and ²⁷Al MAS-NMR spectra were recorded at 800, 211.7 and 208.5 MHz, respectively, on a Bruker 18.8 T spectrometer equipped with a Bruker 3.2 mm CP-MAS probe operating at a ν_rot of 20 kHz. The ²³Na (I = 3/2) MAS-NMR experiments were recorded with a pulse length of 1 µs (~π/5 flip angle), 1024 transients, and a recycle delay of 2 s. The ²⁷Al (I = 5/2) MAS-NMR spectra were obtained with a pulse length of 1 µs (~π/10 flip angle), 1024 transients, and a recycle delay of 2 s. The ¹H MAS-NMR experiments were recorded with a π/2 pulse length of 3.5 µs, 128 transients and a 5s rd using the DEPTH sequence in order to suppress the signal coming from the measurement probe. The presence of Al–O–H bonds was investigated by correlation NMR with the ²⁷Al(¹H) dipolar heteronuclear multiple quantum coherence (D-HMQC) NMR technique [25,26]. The 2D maps were recorded under rotor-synchronized conditions at 18.8 T with 2048 × 40 acquisition points. Each direct slice was recorded with ²⁷Al and ²⁹Si π pulses of 16 and 10 µs, 1024 transients, a rd of 2 s and a 2 ms SR4²₁ re-coupling scheme. All the ²⁹Si, ¹H, ²³Na and ²⁷Al chemical shifts were referenced as 0 ppm to TMS, TMS, NaCl solution (1 M), and Al(NO₃)₃ solution (1 mol.L⁻¹), respectively. It is noteworthy that the spectra acquired at high field (¹H, ²³Na and ²⁷Al) were recorded on similar sample mass (40 mg) in order to compare the peak intensities on the different NMR spectra.

2.6. Acid Leaching Tests

Leaching of different elements (Si, Al and Na) was carried out on alkali-brick samples (mass of composite: 500 mg) in different HCl solutions (volume of leachant: 50 mL) with pH values ranging from 0.39 to 4.83 at room temperature. Under similar experimental conditions, leaching of radioactive elements (Cs and Rb) and Si, Al and Na was carried out on Cs (or Rb)-doped alkali-brick in a 0.01 M HCl solution (pH~2). Each collected leachate was vacuum filtered through a 0.45 μm membrane filter. The concentrations of leached elements present in these leachates were determined by using an inductively coupled plasma optical emission spectrometer (ICP–OES, model 5110 VDV, Agilent Technologies).

3. Results and discussion

3.1. Synthesis Process

During the reaction course of brick-composite synthesis, the metakaolinite mineral initially present in the raw brick was dissolved in an alkali medium by forming sodium poly/mono silicates and sodium aluminates intermediately; these ionic species were subsequently polymerized into gel-state sodium alumino-silicate(s) which impregnated insoluble brick pellets. This polymerization/impregnation was followed by an induction period during which nucleation/crystallization took place at the surfaces of unreacted brick grains (mainly quartz) [27,28]. These authors [27,28] also stated that the presence of a high amount of quartz (60–65 w%) as a “secondary matrix” in brick composite contributed to facilitating the flow of inlet solutions through fixed-bed columns. Note that alkali brick had recently been assimilated to an adsorptive membrane prepared from blended zeolites into a layered geo-polymer-quartz matrix [29]. This mixed material might be considered as a surface composite membrane possessing a dual function of adsorption and filtration: with zeolitic/geo-polymeric aggregates favoring adsorption capacity, while quartz matrix allowing mostly low operating pressure and high permeability flux [29].

3.2. Microscopic and Crystalline Studies

3.2.1. Microscopic (ESEM/EDS) Analysis

ESM analysis of raw brick revealed the presence of porous and non-textural surfaces with major fissures and cracks, except for quartz crystals [22,30]. In addition, elemental (ESEM/EDS) analysis indicated the presence of: (i) silicon (Si), aluminum (Al), oxygen (O) and iron (Fe) as major elements; and (ii) titanium (Ti), magnesium (Mg), potassium (K), and calcium (Ca) as minor elements. Some of these elements were found to be distributed across brick surfaces with apparent localization, evidencing the existence of mainly SiO₂ (as quartz) and alumino-silicate (as metakaolinite) and, to a lesser extent, iron oxides/hydroxides and rutile [22].

As for alkali-activated brick aggregates, ESEM analysis revealed that the morphological aspect of the treated brick was changed significantly into aggregated cubic and spherical shapes (Figure 1). The size of these new specimens varied from 7 µm to 10 µm for cubic ones and from 3 µm to 6 µm for spherical ones. Cubic crystals were found to be comparable with those observed previously for the A-type zeolite [31,32], while spherical shape crystals were morphologically similar to those reported in the literature for zeolite NaP [33,34,35].

The spatial distribution of the framework elements Al, Fe, Na and Si is displayed in Figure 2. The ESEM/EDS mapping procedure gives the color overlay shown in this Figure, where the elemental distributions for Al, Fe, Na and Si are represented in red, green, yellow, and blue, respectively. Element distribution images indicate a positive correlation between Al, Si and Na (Figure 2B) due to the presence of sodic alumino-silicates as Na-zeolites. Conversely, there is a negative correlation between sodium and silicon in Si-rich zones which are composed of quartz crystals (in blue regions of the reconstituted image; see Figure 2A). Note as well the presence of micro-specimens of TiO₂ (rutile) in the elemental distribution for titanium; see Figure 2B.

Quantitative ESEM/EDS analysis permitted us to show that the elemental composition of cubic and spherical specimens corresponded well to those of low-silica zeolites with atomic ratios of Si/Al close to 1. Indeed, the atomic percentage of aluminum and silicon were 14.97 ± 2.05 atomic % and 16.12 ± 1.46 atomic %, respectively, which led to an averaged atomic ratio Si/Al of 1.08 (or Al/Si = 0.926). As for sodium, we found: 14.77 ± 2.67 atomic %. The presence of high amounts of sodium on the surfaces of cubic and spherical particles confirmed the presence of sodic zeolites.

3.2.2. X-ray Diffraction (XRD) Analysis

Figure 3a displays the XRD patterns of raw-brick powder. XRD analysis revealed that the material contains quartz as the principal crystalline mineral and, to a lesser extent, illite and rule. In what follows, minerals were identified on the basis of ‘2θ’ reflection angles and “hkl” Miller indices (“hkl” indices are given in the parenthesis).

Quartz: 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]. Illite: 8.8° (001), 17.9° (004), 19.8° (021), and 34.3° (034) [ICDD (International Centre for Diffraction data): 00-009-0343]. Rutile: 27.4° (110) and 36.1° (101) [ICSD Collection Code: 168140].

Figure 3b displays the XRD patterns of alkali-brick powder. Compared to the XRD patterns of raw brick, one can observe new XRD signals in addition to those ascribed to quartz, illite and rutile. These reflections were attributed to zeolites LTA and NaP.

LTA: 7.2° (200), 10.2° (220), 12.5° (222) and 21.7° (600 and 442) [36]. NaP: 12.5° (101 and 110), 17.7° (200 and 002), 21.7° (211, 112 and 121), 28.1° (310, 301, and 103), 33.4° (132, 123, 231, 213, 312, and 321) and 46.1° (134) [36].

3.3. Solid MAS MNR Study

In the past, Magic Angle Spinning (MAS) Nuclear Magnetic Resonance (NMR) spectroscopy was successfully used for providing relevant information about structure changes, chain lengths and elements ratios of (three-dimensional) geo-polymeric gels present in alkali slag and fly ash as aluminosilicates dominated (Si + Al) industrial by-products [37]. This led us to undertake detailed NMR analyses of alkali brick to gain a comprehensive insight into the different ²⁹Si, ²⁷Al, and ¹H atomic environments constituting the heterogeneous nature of alkali-activated brick. Ammonium chloride (NH₄Cl) was used as a probe for bringing information about the characteristics and mobility in brick-composite frameworks by means of ¹H MAS NMR [1D ¹H and 2D ²⁷Al(¹H) D-HMQC].

Moreover, before undertaking the analysis of NMR data, it was important to mention that it was observed no noticeable broadenings of NMR peaks related to paramagnetic NMR effects arising from the interactions of nuclear spins (¹H, ²³Na, ²⁹Si or ²⁷Al atoms) with unpaired d electrons (Fe(III). This could result from the presence of too-low Fe contents detected in alkali-brick samples by ESEM/EDS (ranging from 0.2% to 0.7%), suggesting that resonating nuclei detected were certainly located far from paramagnetic centers.

3.3.1. ²⁹Si MAS NMR

Figure 4 displays solid-state MAS NMR data for ²⁹Si nuclei in alkali brick before and after different chemical treatments. The ²⁹Si MAS NMR spectrum of an alkali-brick sample indicated several signals (ranging from −70 ppm to ~−135 ppm), which characterized the resonances of framework silicon atoms in Q⁴(mAl) type environments with m varies between 0 and 4. Note globally that in the ²⁹Si MAS NMR spectra, the relatively high intensity of resonances in the range (−84)–(−90) ppm reflected the importance of Al-rich structural units inside brick frameworks.

In order to facilitate the interpretation of ²⁹Si MAS NMR data, the experimental ²⁹Si MAS NMR spectra were de-convoluted (Figure 4). Components were identified through the de-convolution of the spectra into individual Gaussian/Lorentzian peaks representing different silicon centers: Qⁿ(mAl) with n and m capable of varying from 0 to 4. The de-convolution criterion followed at best published ²⁹Si NMR results relative to the different chemical environments surrounding the silicon nucleus belonging to the following compounds [38,39,40,41,42]: (i) zeolites NaA and NaP; (ii) aluminosilicate gels produced during the alkali activation of metakaolin; and (iii) quartz as an unreacted mineral present in the starting material. It is noteworthy to mention that other un-reacted starting materials such as illite and biotite (which are nevertheless present at low levels in the analyzed material) would also contribute to the resonances detected and, as a consequence, lead to some degree of error in the decomposition of the ²⁹Si MAS NMR spectra.

The de-convolution of the ²⁹Si MAS NMR spectrum of alkali brick revealed a series of peaks at −84.5, −89.2, −94.0, −100.7, −107.5, −113.3 and −119.5 ppm (±0.3 ppm), see Figure 4 and

Table 1. De-convolution data was obtained for the ²⁹Si MAS NMR spectra of alkali-activated brick before and after slight acidification (pH 5) and ammonium treatment.

29Si MAS NMR (de-Convolution)	Q4(4Al) (in gel)	Q4(4Al)ze and Q4(3Al)gel	Q4(2Al)	Q4(2Al) and/or Q4(1Al)	Q4(0Al)	Q4(0Al)	Q4(0Al)
Alkali-activated brick
δiso/ppm	−84.5	−89.2	−94.0	−100.7	−107.5	−113.3	−119.5
FWHM/ppm	2.2	5.1	6.7	8.0	4.8	8.0	9.8
Relative intensity(%)	9.2	19.5	22	16	11	13	9.2
Alkali-activated brick (pH 5)
δiso/ppm	−84.5	−88.7	−93.5	−100.7	−107.5	−113.3	−120.7
FWHM/ppm	2.2	5.0	6.7	8.2	4.8	8.0	9.8
Relative intensity(%)	11	20	23	14	11	14	7
NH4+-loaded alkali-activated brick
δiso/ppm	−84.5	−89.7	−94.6	−101.5	−107.5	−113.3	−120.2
FWHM/ppm	2.2	4.9	6.7	8.0	4.8	8.0	9.8
Relative intensity(%)	11	21	20	15	12	13	8

The chemical shift (δ_iso), full width at half maximum (FWHM) and relative peak intensity (I%) are given with errors of ±0.3 ppm, 0.4 ppm and 1.0%, respectively.

. The NMR line at −89.2 ppm corresponded well to the chemical shift of Si atoms in Q⁴(4Al) sites in NaA zeolite [39] and NaP zeolite [41]. The NMR resonances at −94.0 and −100.7 ppm were close enough to those detected for NaP zeolites by Criado and co-workers [41]: These authors ascribed them to silicon tetrahedra surrounded by a certain coordination number of ²⁷Al atoms ranging from 3 to 1, respectively (Q⁴(1–3Al)). As a whole, the NMR peaks at −94.0 and −100.7 ppm might certainly be due to Si(OSi)(OAl)₃, Si(OSi)₂(OAl)₂ and/or Si(OSi)₃(OAl)₁, respectively. As for the particular NMR peak at −84.5 ppm, it was associated with the presence of sodalite-type Q⁴(4Al) units in brick gel [41,43]. The −84.5 ppm signal was comparable enough to that observed previously for alkaline inorganic polymers generated through the alkaline activation of pure metakaolinite [40]. Such polymers exhibited additional peaks at around −87/−89 ppm and −94 ppm, which were superimposed on those of NaA and NaP zeolites [40]. These resonances were, therefore, assigned partly to tetrahedrally coordinated and fully polymerized (Q⁴) silicon nuclei surrounded by m AlO₄ tetrahedra and associated with the chemical ²⁹Si environments in alkali-activated metakaolin. Indeed, according to Fernandez-Jiménez and co-workers [40], the ²⁹Si MAS-NMR spectrum of metakaolin illustrating its chemical transformation into an alumino-silicate gel during alkaline activation exhibited three well-defined signals at around −84.6, −89, and −94 ppm. These chemical shifts were similar enough to those obtained for chemically bonded ceramic/cementitious materials of alkali-activated metakaolin, revealing a wide signal with several overlapped peaks, and their de-convolution indicated the presence of Q⁴(4Al), Q⁴(3Al) and Q⁴(2Al) units at −84, −89 and −92 ppm, respectively [44,45]. Note further that the relatively high intensity of the −89.2 ppm signal observed in Figure 4 was explained by the fact that zeolites NaA and NaP contributed strongly to this signal. The −89.2 ppm resonance, which was associated with Q⁴(4Al) units in zeolites NaA and NaP, was assigned to the resonance of ²⁷Al nuclei in double four-membered rings (with sodalite cages) of brick zeolites [38,40]. It was also interesting to note that, after the adsorption of NH₄⁺ ions onto alkali brick, the ²⁹Si sites were somewhat better distinguished and, hence, the different MAS NMR lines could be more easily quantified (Figure 4 and Table 1).

In addition, it was found that by using X-ray diffraction, brick quartz was apparently not (or barely) attacked by alkaline treatment [46]. A resonance at −107.7 ppm in the ²⁹Si MAS NMR spectrum of alkali brick (Figure 4) revealed Q⁴(0Al) units in quartz [47,48]. The low intensity of this signal in our ²⁹Si MAS-NMR experiments was explained by the relatively short recycle delay. This delay did not allow for a full relaxation of the quartz signal, leading to an under-estimated intensity compared to the other ²⁹Si resonances. In addition to the Q⁴(0Al) signal corresponding to the quartz mineral, other Q⁴(0Al) signals were observed at −113.3 and −119.5 ppm (see Figure 4 and Table 1). According to the literature, these peaks could be attributed to highly polymerized amorphous silica structures generated through condensation of silicate tetrahedrons during metakaolin geo-polymerization, thus leading to an increasing fraction of Q⁴(0Al) [45,49].

From ²⁹Si NMR data (see Table 1), it was possible to estimate the silica contribution of un-reacted aluminosilicate gel, Si(gel)%, to the overall ²⁹Si MAS NMR peak of “zeolitized” brick-metakaolinite from the expression (in atomic %):

$$\mathrm{Si}(\mathrm{gel})\%=\frac{\mathrm{I}\%(\mathrm{gel})}{\mathrm{I}\%(\mathrm{gel})+\mathrm{I}\%(\mathrm{zeolites})}\times 100\%$$

(1)

where I%(gel) and I%(zeolites) represent the intensity percentages of ²⁹Si resonances ascribed respectively to aluminosilicate gel and zeolites in the ²⁹Si MAS NMR spectrum of alkali-activated brick. Note that in this calculation: (i) we considered only the areas of the signals at −84.5, −94.0 and −100.7 ppm, which were assigned to Q⁴(4Al), Q⁴(2Al), Q⁴(2-1Al) units in aluminosilicate gel(s) [41,43]; and (ii) we did not take into account polymerized amorphous silica structures with Q⁴(0Al) (i.e., the ²⁹Si NMR peaks at −113.3 ppm and −119.5 ppm, see Table 1). We found at least 70 atomic % of ²⁹Si nuclei present in the “aluminosilicate” gel(s) of the partially zeolitized brick.

From the quantification of the different Q⁴(mAl) species of modified bricks (see Table 1), Engelhard’s formula (Equation (2)) was used here to estimate the Si/Al molar ratio in aluminosilicates at composite surfaces [50].

$$\frac{\mathrm{Si}}{\mathrm{Al}}=\frac{{\displaystyle {\sum }_{\mathrm{m}-0}^{\mathrm{m}=4}{\mathrm{Q}}^4(\mathrm{mAl})}}{{\displaystyle {\sum }_{\mathrm{m}=9}^{\mathrm{m}=4}\frac{\mathrm{m}}{4}·{\mathrm{Q}}^4(\mathrm{mAl})}}$$

(2)

We found an atomic Si/Al ratio ranging approximately from 1.35 to 1.53. These Si/Al values result from macroscopic chemical (NMR) analysis of silicon elements in the material. Conversely, the ESEM/EDS results obtained corresponded rather to microscopic chemical analysis of brick surfaces (see Section 3.2.1 and

Table 2. Elemental (ESEM/EDS) analyses of alkali-activated brick aggregates were performed at different magnifications.

ESEM/EDS		O	Na	Mg	Al	Si	K	Ca	Ti	Fe
Alkali-brick surface (global) a	At. %	60.98	9.08	0.37	10.97	17.29	0.56	0.07	0.18	0.50
	σ	0.89	0.25	0.02	0.68	0.60	0.25	0.03	0.05	0.18
Targeted zeolitic zones b	At. %	53.66	9.56	0.91	14.89	18.24	1.00	0.41	0.13	0.79
	σ	1.71	5.35	0.22	1.75	3.33	1.39	0.55	0.03	0.53
Zeolite specimens targeted c	At. %	50.75	12.37	0.55	16.61	18.38	0.25	0.44	0.07	0.44
	σ	6.29	2.07	0.16	2.65	1.48	0.03	0.39	0.02	0.24

^a see Figure S3A; ^b see Figure S3B; ^c see Figure S3C.

) and cannot be directly compared with NMR data. For instance, by using the ESEM/EDS technique, we instead found the following atomic Si/Al ratios: (i) Si/Al = 1.58 ± 0.16 for large, targeted zones of surface alkali-brick recovered with both quartz and zeolites; (ii) 1.22 ± 0.08 for targeted zeolites-rich zones; and (iii) 1.11 ± 0.07 for specific zeolitic specimens. As a consequence, it could be stated from ESEM/EDS investigations that alkali-brick grains were mostly coated with zeolitic frameworks (NaA and NaP) (as shown in Figure 1 and Figure 2) and would barely be coated with alumino-silicate gel(s) (as detected macroscopically by NMR). To our mind, composite gels ought to be more present in layer(s) beneath zeolitic coatings.

3.3.2. ¹H and ²³Na MAS NMR

Figure 5 displays solid-state MAS NMR data for ¹H nuclei in fine alkali-brick grains in which brick metakaolinite was thoroughly transformed into zeolites. The ¹H MAS NMR spectrum of zeolitized brick exhibits an intense signal spreading between ~2 ppm and ~9 ppm (Figure 5A). In this chemical shift range, it can be seen a principal resonance at 4.6 ppm accompanied by weak and sharp peaks at 4.2 ppm and 3.7 ppm. All these ¹H NMR resonances were ascribed to a wide distribution of various water molecules and proton species taking place in the zeolitic frameworks (α and β cages) of zeolitized brick [22]. As for the badly resolved resonances in the δ¹H range of ~1–3 ppm, they were either non-acidic or weakly acidic hydroxyl groups. These were assigned to isolated silanol/aluminol protons in brick minerals.

After the addition of an NH₄Cl solution ([NH₄Cl] = 0.5 mol·L⁻¹) into a treated-brick suspension, the ¹H MAS NMR spectrum of the recovered material revealed that zeolitic protons shifted significantly to a lower field (Figure 5B). A single resonance appeared at δ = 7.0 ± 0.1 and was assigned to NH₄⁺ ions which were adsorbed onto the Brønsted sites of the treated brick, leading to NH₄⁺-zeolite complexes [51,52,53]. As for Na⁺ ions adsorbed onto treated brick, ²³Na MAS NMR analysis indicated their near disappearance after NH₄Cl treatment, confirming the involvement of an ion-exchange reaction between sodium(I) and ammonium(I) at the brick-water interface according to an equimolar heterogeneous process (Figure 6):

>SO⁻Na⁺ + NH₄⁺→ >SO⁻NH₄⁺ + Na⁺

(3)

Note that the residual ²³Na signals observed in Figure 6B might be due to structural sodium bound to other brick minerals.

On the other hand, in order to generate the H-form of zeolites, alkali-activated brick was acidified progressively at a pH value close to 5 with an HCl solution ([HCl] = 10⁻³ mol·L⁻¹). This pH value was chosen lower than the pH at the point of zero charge, PZC (pH_PZC = 5.85) or the pH at the iso-electric point (pH_IEP = 5.9), see Figure S1, indicating the predominant occurrence of positively charged brick surfaces associated with hydroxonium ions. Acidified brick samples were afterward analyzed by ¹H and ²³Na MAS NMR spectroscopy. The ¹H MAS NMR spectrum of the acidified brick revealed the presence of a sharp peak at δ = 4.7 ppm ascribed to H₃O⁺ ions bound to brick zeolites, see Figure 7A. This signal was indeed due to hydrogen nuclei in bridging (structural) hydroxyl groups that underwent a rapid exchange between water molecules and hydroxonium ions in zeolitic cages. Such an observation agreed noticeably well with previous studies about the formation of hydroxyl groups in the low-silica zeolite with n_Si/n_Al = 1 (during the exchange of Na⁺ cations by ammonium ions), particularly showing close signals in the chemical shift range of 3.6–4.8 ppm due to Si(OH)Al groups in the α and β-cages [54]. The resonance at δ_1H = 4.70 ppm was also found to be similar enough to that observed in the spectra of X and Y zeolites and ascribed to Si(OH)Al groups in sodalite cages [52,55].

Acidified alkali-brick was subsequently treated with an NH₄Cl solution ([NH₄Cl] = 0.5 mol·L⁻¹). The ¹H MAS NMR spectrum of the recovered sample displayed three new proton resonances (Figure 7B). The peaks at δ₁ ≅ 7.3 ppm and δ₁’ = 7.0 ppm were assigned both to NH₄⁺ ions adsorbed onto Brønsted sites of framework zeolites: LTA and NaP (note that the δ₁ and δ₁’ peaks were not yet assigned to specific zeolites). Whereas the peak at δ₂ = 5.3 ppm was attributed to ammonium species involved in fast exchanges with NH₄⁺ (δ₁ and δ₁’) and hydroxonium ions in excess. A comparable magnetic phenomenon was already evidenced for other zeolites [56,57]. Moreover, by rotating the rotor at a higher speed (from 20 kHz to 60 kHz), the temperature of the rotor increased noticeably, which contributed to affecting the magnetic characteristics of the sample. The obtained ¹H MAS NMR spectrum then exhibited a line broadening and ¹H NMR signals merging, see Figure 8. This could be explained by hydrogen nuclei that underwent a rapid exchange between ammonium and hydroxonium ions. This suggested the possibility of a transfer of all four hydrogen nuclei from one ammonium ion to bridging hydroxyl groups. The small number of water molecules in the treated brick should also contribute to an increase in the mobility of ¹H nuclei along bridging Si−O⁻−Al chains. The resulting merging signal was centered at δ₃ = 5.07 ppm. Note that the weak resonance detected at the lower field in the ¹H MAS NMR spectrum (δ₁″ = 6.75 ppm) was still ascribed to the own protons of ammonium ions directly bound to bridging Si−O⁻−Al sites (as Si−O⁻NH₄⁺−Al).

It could then be contended that upon H₃O⁺/NH₄⁺ exchange, the resonance of zeolitic ¹H nuclei were low-field-shifted by Δδ₁ = 2.45 ppm (from 4.70 to ~7.15) and Δδ₂ = 0.60 ppm (from 4.70 to 5.30), which globally reflected an increasing acid strength of brick protons. Such low-field-shifts induced by ammonium ions were previously observed in ¹H Solid-state NMR studies of Brønsted acid sites of other zeolites [52,53,56,57] and flame-derived silica/alumina [51] in which silanols with neighboring aluminum atoms generated bridging AlOHSi groups. However, it is worth noting that ¹H, ²³Na and ²⁷Al MAS NMR analyses permitted to reveal a progressive decomposition of brick zeolites as the medium pH decreased below 5 (see Figure S2).

3.3.3. ²⁷Al MAS NMR

Figure 9 displays solid-state MAS NMR data for ²⁷Al nuclei in alkali brick before and after NH₄Cl treatment. Globally, dominating ²⁷Al resonances observed for studied brick samples fall in the range δ = 59.0–63.6 ppm. Such chemical-shift variations in the local environment of ²⁷Al(IV) within alkali-brick might suggest that Al(IV) sites would exist in distinct aluminosilicate frameworks and would be differentially affected by structural factors, i.e., bond angles/lengths in T–O–T, with T = Si or Al [14,58].

In the ²⁷Al MAS NMR spectrum of an alkali-brick sample, most aluminum atoms are tetrahedrally coordinated (Al^IV), which results in three ²⁷Al MAS NMR signals, see Figure 10A. The particular NMR resonances detected in the chemical shift range 57–61 ppm (see Figure 10A) were partially attributed to tetrahedrally coordinated aluminum (bound via oxygen to four ²⁹Si atoms to give Al(OSi)₄ units) of both NaA and NaP zeolites [34,59,60,61,62,63,64,65].

According to the literature [34,63,64,65], a slightly lower chemical shift (about 1 ppm) in the δ²⁷Al range 57.6–59.3 ppm was more expected for NaP than for NaA. As for the NMR resonance at 63.2 ppm (see Figure 10A), it was ascribed to tetrahedrally coordinated aluminum in geo-polymeric gels, which were not transformed into zeolites [66]. Note that, for a series of sodalites (Si/Al = I), Jacobsen et al. (1989) [67] observed a linear dependence of ²⁹Si and ²⁷Al chemical shifts on mean “Si–O–Al” bond angles (θ):

δ(²⁹Si) = −0.702·(θ) + 11.626

(4)

δ(²⁷Al) = −0.724·(θ) + 163.982

(5)

From Equation (4), we attempted to calculate the mean Si–O–Al angle for brick hydroxyl-sodalite(s) (BS) with a ²⁹Si chemical shift δ(²⁹Si) = −84.5 ppm (see Figure 4). We found: θ_BS = 136.93°. This angle value permitted us to subsequently estimate the corresponding chemical shift of ²⁷Al nuclei in BS by using Equation (5). We found δ(²⁷Al) = 64.83 ppm. This value was nevertheless higher than that observed experimentally: δ(²⁷Al)_exp. = 63.2 ppm. In contrast, by using Lippmaa et al.’s formulae (Equation (6); [68]) indicating the linear dependence of the ²⁷A1 chemical shift on the mean [Al−O−Si] bond angle in framework alumino-silicates, we found a δ(²⁷Al) value: δ(²⁷Al) = 63.53 ppm, which was much closer to the experimental one.

δ(²⁷Al) = −0.50·(θ) + 132 (ppm)

(6)

As for ²⁹Si nuclei of alumino-silicate gel(s) detected at −89.2 ppm (which superimposed on those of NaA and NaP zeolites), −94 ppm and −100.7 ppm, these chemical shifts would correspond to Q⁴(3Al), Q⁴(2Al) and Q⁴(1Al) units, respectively; According to Ramdas-Klinowski’s works [69], the ²⁹Si chemical shifts of brick gels could be correlated linearly with the structural parameter, ∑d_TT, according to:

δ(²⁹Si) = 143.03 − 20.34·(∑d_TT/Å)

(7)

in which ∑d_TT is defined as:

∑d_TT = [3.37 × m + 3.24 × (4 − m)] sin(θ/2)

(8)

where m is the number of aluminum atoms in the Si(mAl) unit (in this study, m = 1, 2, 3). Using Equations (7) and (8), one could evaluate the mean Si–O–Al angle (T = Si or AI) in the framework alumino-silicate. We found θ = 141.2°, 140.4° and 142.0° respectively for Q⁴(3Al), Q⁴(2Al) and Q⁴(1Al) units of brick gel(s), see

Table 3. Chemical shifts, δ(²⁹Si) of ²⁹Si nuclei of the different Q⁴(mAl) units present in modified-brick frameworks.

	Q4(4Al) (in Gel)	Q4(4Al) (in Zeolite)	Q4(3Al) (in Gel)	Q4(2Al) (in Gel)	Q4(1Al) (in Gel)
Alkali-activated brick aggregates
δ(29Si)	−84.5	−89.2	−89.2	−94.0	−100.7
SiOAl(Si), θ° ( δ(27Al))	142.2 (60.9	148.8 (57.6)	140.4 (61.8)	140.4 (61.8)	142.0 (61.0)
NH4+-loaded alkali-brick
δ(29Si)	−84.5	−89.7	−89.7	−94.6	−101.5
SiOAl(Si), θ° ( δ(27Al))	~139 (~62.5)	147.6 (58.2)	~139 (~62.5)	~139 (~62.5)	~139 (~62.5)

.

From these angle values, it was possible to assess the corresponding chemical shifts of ²⁷Al nuclei present in the different Q⁴(mAl) units of brick gel(s) by employing Equation (6). We found δ(²⁷Al) = 61.4 ppm, 61.8 ppm and 61.0 ppm for Q⁴(3Al), Q⁴(2Al) and Q⁴(1Al) units of brick gel(s), respectively. These δ(²⁷Al) values were consistent with the broad NMR resonances ranging from ~59 to ~61.5 ppm, which were observed in the ²⁷Al MAS NMR spectrum of alkali-brick (see Figure 10A). Likewise, the mean Si–O–Al angle (T = Al) in framework zeolites of alkali-brick was evaluated from Equations (7) and (8): we found: θ = 148.8° for zeolitic Q⁴(4Al) units (see Table 3); the corresponding chemical shift of ²⁷Al nuclei present in these Q⁴(4Al) units was determined from Lippmaa et al.’s expression (Equation (6) [68]): δ(²⁷Al) = 57.6 ppm. Also, from Engelhardt et al.’s expression (Equation (9) [43]), the mean Si–O–Al angle (T = AI) in framework zeolites of alkali brick could be estimated (θ = 147.3°): in this case, the corresponding chemical shift value of ²⁷Al nuclei (δ(²⁷Al) = 58.3 ppm) was found to be slightly higher than that determined above.

δ(²⁹Si) = −0.570·(θ) − 5.230

(9)

A weak signal observed at 3.7 ppm was attributed to the presence of a small amount of octahedrally coordinated aluminum (Al^VI). ²⁷Al nuclei corresponding to penta-coordinated aluminum (Al^V) were also present inside alkali-brick networks and gave rise to a signal at approximately δ ≅ 20–30 ppm.

After NH₄Cl treatment, the ²⁷Al MAS NMR spectrum of the recovered material changed noticeably in comparison with that of the alkali-brick (Figure 9 and Figure 10). The incorporation of NH₄⁺ ions at high concentration into the alkali-brick suspension strongly displaced alkali cations (Na⁺) from the extra-framework sites and fulfilled the charge balancing role (as shown in Figure 6 relative to the Na⁺/NH₄⁺ exchange at the brick surface). As observed by ²⁷Al MAS NMR (Figure 10B), the addition of ammonium to alkali-brick frameworks led to a broader signal at 62.5 ppm as a consequence of an overlapping of the ²⁷AlO₄ resonances in the range ~59–63 ppm with an increase in the full width at half height (FWHH). This might be explained by the fact that the interactions of NH₄⁺ ions with Q⁴ AlO₄ sites of aluminosilicate gel(s) would contribute to a decreasing relaxation of the distortional bonding strain around Al^IV atoms. Such strains would lead to a shortening of Al–O bonds with improved tetrahedral symmetry when compared to Al bonding strains caused by H-bondings between Al sites and a water-sodium(I) mixture [70]. To support this, we reported here some relevant findings regarding protons’ effects on the environment of silicon and aluminum framework atoms of alumino-silicates [71]. The authors in reference [71] concluded that the removal of protons from double four-membered rings (D4R: being present in sodalite structures) led to a strengthening of Si–O and Al–O bonds (and thereby a decrease of their lengths). In the resulting anionic framework, the Si–O–Al angle (138.4°) became larger (by 7.3°) than that at the bridging hydroxyl site (131.1°). In addition, in Section 3.3.2, it was shown that the adsorption of NH₄⁺ ions onto alkali-brick resulted in acidification of brick particles, as evidenced by the significant shifting towards lower field observed by ¹H MAS NMR (see Figure 7B). This finding implied that similar decreases of Si–O–Al angles should exist in framework brick aluminosilicates from the anionic material (>SO⁻Na⁺) to the (pseudo)-protonated form, >SO⁻H₃O⁺NH₃, as previously suggested [51,52,53,56,57]. In this assumption, the mean Si–O–Al angles of NH₄⁺-loaded alkali-brick were evaluated by using Lippmaa et al.’s formulae, Equation (6). “Si–O–Si(Al)” angles data showed that: (i) the mean Si–O–Al angle in alkali-brick (148.8°) would be slightly higher (by 1.1°) than that at bridging hydroxyl sites of “protonated” alkali-brick (147.6°). As for aluminosilicate gels, the decrease of the mean Si–O–Al angle should range from ~1° to ~3°.

3.3.4. 2D MAS NMR (²⁷Al-{¹H} D-HMQC)

In order to observe ²⁷Al–¹H dipolar-mediated correlations, 2D ²⁷Al(¹H) D-HMQC experiments were performed on ammonium-loaded H-form of alkali-brick. As shown in Figure 11, two correlations were observed in the two-dimensional ²⁷Al(¹H) D-HMQC spectrum: one between ¹H nuclei of NH₄⁺ (δ_1H = 7.0 ppm) and Al^IV (δ_27Al ≅ 64 ppm); and the other between ¹H nuclei of H₃O⁺ (δ_1H = 3.7 ppm) and Al^IV (δ_27Al ≅ 62 ppm).

The observation of an Al^IV−NH₄⁺ cross-peak in Figure 11 at (64, 7.0) ppm suggested that Brønsted acid sites Si−OH−Al, with Al^IV atoms were able to transfer protons of bridging hydroxyl groups to ammonia molecules. This chemical surface transfer led to a significant low-field shift of the ¹H NMR signal, as already observed previously for different types of zeolites [52,53,56,57]. Moreover, the ¹H MAS NMR signal reached a maximal intensity at about 5 ppm. At this resonance maximum, some of the NH₄⁺ ions were involved directly in fast exchanges with ammonia adsorbed onto Brønsted sites (δ = 7.0 ppm) and hydroxonium partially present in the H-form of brick zeolites (δ = 3.7 ppm). Such an NH₄⁺/H⁺ exchange on brick surfaces indicated that not all bridging hydroxyl groups were involved in the signal at low field (i.e.: δ_1H = 7.0 ppm). This could be explained by the fact that the hydrogen-exchange rate between ¹H nuclei of bridging hydroxyl sites and the four hydrogen nuclei of the ammonium ions became close to that relative to the chemical exchange of bridging hydroxyl groups between the sites. Such mixing exchanges then gave rise to a merging signal which occurred at a higher field shift (and therefore at a lower acidity strength) than that observed for ammonia interacting directly with negatively-charged bridging hydroxyl sites (Si−O⁻−Al groups), as suggested by Wang and his co-workers [53]. Note that by using ¹H MAS NMR and by rotating the rotor at different speeds, the transfer of hydrogen nuclei from ammonium ions to bridging hydroxyl groups was evidenced above (in Section 3.3.2 and Figure 8).

3.4. Aqueous Structural Stability and Cationic Immobilization Characteristics

In order to assess acidity effects on the aqueous structural stability of the brick composite at room temperature, a series of leaching tests were carried out at the following medium pH values: 0.39, 0.63, 0.80, 1.09, 1.27, 1.40, 1.67, 2.15, 2.32, 2.51, 3.70, 4.35 and 4.83. After a leaching time of 4 hours, the amounts of brick elements (Si, Al and Na) leached from the brick composite were determined and represented in the histogram of Figure 12. As seen in this figure, there are some noticeable differences between the leaching patterns of framework elements (Al and Si) and that of the extra-framework element (Na) with medium acidification. The concentrations of Al and Si present in the leachates remained low at reaction pH values above 2.5, suggesting a relative acid stability of the aluminosilicate matrix. Conversely, sodium was released at high concentrations with increasing acidification, revealing a facile exchange of Na⁺ ions by H₃O⁺ ions during leaching. This was in agreement with recent thermodynamic data revealing weak acid characteristics of Brønsted acid sites in alumino-silicate frameworks of alkali-activated brick [29]. At pH values below 2.5, the acid degradation of the brick composite occurred strongly, and hence the concentrations of Al, Si and Na in the leachates became important (Figure 12).

On the other hand, the equilibrium reactions between the Na-exchanged forms of the brick composite and aqueous solutions containing one, two or three cationic metals, such as Cd²⁺, Fe²⁺, Ni²⁺ and Pb²⁺ (Me²⁺), were previously studied thermodynamically on the basis of both the general properties of divalent metals in water and the global (zeolitic and geo-polymeric) surface characteristics of alkali-brick [46]. These investigations clearly revealed the importance of first hydrated radius, ionic potential and hydration-free energy, and second kinetics and mass-transfer/diffusion on the course of the global ion-exchange process at the brick−water interface [46]. Moreover, thermodynamic calculations concerning interfacial 2Na⁺/Me²⁺ ion-exchange reactions revealed that the replacement of the exchangeable cation (Na⁺) on the composite by Me²⁺ ions in the solution occurred favorably [46]. This proved that the immobilization efficiency of the composite for cationic metals is better than for Na⁺ ions. To support this, acid leaching of alkali-brick samples doped with radio-active cations (Cs⁺ and Rb⁺) was carried out in a 0.01 M HCl solution (pH~2) at room temperature and under limited acid conditions close to material decomposition (as shown in Figure 13).

As expected, the concentrations of Al, Si, Na and radioactive elements in the leachates were found to be relatively high as a result of the acid degradation of the brick composite. By using the same leaching procedure, we afterward leached brick-composite samples which were previously calcined at 600 °C and 1000 °C for 4 h. Leaching results obtained at 600 °C indicated two features: (i) the leach fractions of Al and Si decreased noticeably, showing a slightly increased stability of calcined material; and (ii) the amounts of Na⁺, Cs⁺ and Rb⁺ (as extra-framework monovalent cations, M⁺) in the leachates diminished as well, showing a positive effect on M⁺ immobilization. This relative (weak) immobilization could be explained by the loss of water molecules in hydrated cationic shells during calcination, enabling the transformation of outer-sphere “M(H₂O)_x⁺-brick” complexes into inner-sphere “M⁺-brick” complexes (as evidenced by MAS NMR, Boughriet et al., unpublished works). As for leaching results obtained at 1000 °C, they revealed that the concentrations of Al, Si, Na, Cs and Rb present in the leachate became very low, evidencing both a greater stabilization of the brick matrix and a strong immobilization of Na⁺, Cs⁺ and Rb⁺ ions, as already observed for alkali-activated (geo-polymerized) blast furnace slag/fly ash [3,13,72]. This latter point suggested that Na⁺, Cs⁺ and Rb⁺ ions were presumed to be encapsulated inside a material that was structurally modified by heating, as previously mentioned for fly ash-silica fume-based geo-polymers after their calcination [13]. However, more studies on the brick-based composite must be undertaken to gain relevant information about structural/crystalline modifications induced by high temperatures.

4. Conclusions

The brick material was treated with sodium hydroxide using careful temperature control, giving rise to a geo-polymerization followed by a progressive zeolitization of geo-polymeric gels. X-ray diffraction analysis revealed the formation of crystalline structures identified as low-silica zeolites (NaA and NaP). ESEM/EDS microscopic analysis permitted observation of the cubic and spherical specimens at the brick surface and the evaluation of their averaged elemental composition, thus confirming the occurrence of low-silica zeolites with atomic ratios Si/Al close to 1. Solid MAS NMR spectroscopy permitted us to obtain a comprehensive insight into the different ²⁹Si, ²⁷Al and ¹H nuclei environments existing in brick-composite networks. ²⁹Si and ²⁷Al MAS RMN studies revealed the still existence of geo-polymerized gels in combination with low-silica zeolites and allowed an estimation of the “geo-polymeric silicon” contribution inside the studied material. ¹H MAS NMR studies showed the existence of bridging Si−O⁻H₃O⁺−Al chains and their implication in rapid exchanges with hydrogen nuclei of water molecules and ammonium ions. Protons transfer from ammonium to bridging Si−O⁻−Al sites was confirmed by using two-dimensional ²⁷Al(¹H) D-HMQC NMR. Detailed spectral analyses of ¹H, ²⁹Si and ²⁷Al NMR resonances allowed the assessment of “Si–O–Si(Al)” angles. Structural angle calculation permitted us to evidence the effects of extra-framework cations (Na⁺, H₃O⁺ and NH₄⁺) on negatively-charged bridging hydroxyl sites (Si−O⁻−Al groups).

Acid-leaching tests revealed that heating treatment was an effective method to stabilize the significant composite structure, as well as to improve the immobilization of radio-active elements (Cs and Rb).

5. Future Prospects

Alkali-activated brick represents a novel alkali-activated binder on which negative surface charges represent potential immobilization sites for radio-active cationic species. Geo-polymers and low-silica zeolites would be considered interesting binding phases for radio-elements uptake. The low Si/Al ratio in these compounds contributed to high adsorption density that would facilitate ionic exchanges between Na⁺ ions at the solid surface and radioactive ions (Cs⁺, Rb⁺, Sr²⁺ and Co²⁺) in the solution. However, a comprehensive insight into interfacial phenomena must still be clarified on the basis of kinetics and thermodynamics investigations. The knowledge obtained here on the compositional, structural and interfacial (ionic) properties of brick-based composite would help: (i) to interpret the interfacial ion exchange; (ii) to understand the uptake and leaching mechanism of radioactive ions; and (iii) to ensure better nuclear waste immobilization and nuclear radiation shielding after calcination/vitrification.

Pyrolysis−Fourier transform infrared (Py-FTIR) spectrometric studies of alkali-brick at elevated temperatures are in prospect for understanding the pyrolysates composition and their evolution over time and analyzing the difference between Na and H form geo-polymer. In addition, studies are underway to evaluate the state of the immobilized radio-elements and their degree of encapsulation inside composite frameworks after calcination at different temperatures. For that, X-ray photoelectron spectroscopy (XPS) should be useful for etching matrix surfaces by ionic Ar+ bombardment at different durations and analyzing surface elements versus formed-crater depth.