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

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.


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 nonrenewable 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. 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 NH 4 + -saturated alkali-brick, an ion-exchange reaction between Na + ions in brick frameworks and NH 4 + 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 NH 4 Cl 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.

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

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 mm 2 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.

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.

MAS NMR Analysis
The 29 Si 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 29 Si 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 29 Si MAS-NMR spectra were decomposed using the Gaussian/Lorentzian model of the dmfit software (version 2.0) [24]. The 1 H, 23 Na and 27 Al 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 23 Na (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 27 Al (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 1 H 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 27 Al( 1 H) 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 27 Al and 29 Si π pulses of 16 and 10 µs, 1024 transients, a rd of 2 s and a 2 ms SR4 2 1 re-coupling scheme. All the 29 Si, 1 H, 23 Na and 27 Al chemical shifts were referenced as 0 ppm to TMS, TMS, NaCl solution (1 M), and Al(NO 3 ) 3 solution (1 mol.L −1 ), respectively. It is noteworthy that the spectra acquired at high field ( 1 H, 23 Na and 27 Al) were recorded on similar sample mass (40 mg) in order to compare the peak intensities on the different NMR spectra.

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).

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].

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 SiO 2 (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].

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 Nazeolites. 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. 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 TiO 2 (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. 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.

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 29 Si, 27 Al, and 1 H 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 1 H MAS NMR [1D 1 H and 2D 27 Al( 1 H) 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 ( 1 H, 23 Na, 29 Si or 27 Al 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.

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 29 Si, 27 Al, and 1 H atomic environments constituting the heterogeneous nature of alkali-activated brick. Ammonium chloride (NH 4 Cl) was used as a probe for bringing information about the characteristics and mobility in brick-composite frameworks by means of 1 H MAS NMR [1D 1 H and 2D 27 Al( 1 H) 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 ( 1 H, 23 Na, 29 Si or 27 Al 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. 29 Si MAS NMR Figure 4 displays solid-state MAS NMR data for 29 Si nuclei in alkali brick before and after different chemical treatments. The 29 Si 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 Q 4 (mAl) type environments with m varies between 0 and 4. Note globally that in the 29 Si 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.
Ceramics2023, 6, FOR PEER REVIEW 8 3.3.1. 29 Si MAS NMR Figure 4 displays solid-state MAS NMR data for 29 Si nuclei in alkali brick before and after different chemical treatments. The 29 Si 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 Q 4 (mAl) type environments with m varies between 0 and 4. Note globally that in the 29 Si 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.  29 Si MAS NMR spectra of alkali brick before and after different chemical treatments. Deconvoluted spectra were obtained by employing the dmfit software [24].
In order to facilitate the interpretation of 29 Si MAS NMR data, the experimental 29 Si 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: Q n (mAl) with n and m capable of varying from 0 to 4. The deconvolution criterion followed at best published 29 Si 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 29 Si MAS NMR spectra.
The de-convolution of the 29 Si 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. The NMR line at −89.2 ppm corresponded well to the chemical shift of Si atoms in Q 4 (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 27 Al atoms ranging from 3 to 1, respectively (Q 4 (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 Q 4 (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].  29 Si MAS NMR spectra of alkali brick before and after different chemical treatments. De-convoluted spectra were obtained by employing the dmfit software [24].
In order to facilitate the interpretation of 29 Si MAS NMR data, the experimental 29 Si 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: Q n (mAl) with n and m capable of varying from 0 to 4. The de-convolution criterion followed at best published 29 Si 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 29 Si MAS NMR spectra.
The de-convolution of the 29 Si 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. The NMR line at −89.2 ppm corresponded well to the chemical shift of Si atoms in Q 4 (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 27 Al atoms ranging from 3 to 1, respectively (Q 4 (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 Q 4 (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 (Q 4 ) silicon nuclei surrounded by m AlO 4 tetrahedra and associated with the chemical 29 Si environments in alkali-activated metakaolin. Indeed, according to Fernandez-Jiménez and co-workers [40], the 29 Si 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 Q 4 (4Al), Q 4 (3Al) and Q 4 (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 Q 4 (4Al) units in zeolites NaA and NaP, was assigned to the resonance of 27 Al 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 NH 4 + ions onto alkali brick, the 29 Si sites were somewhat better distinguished and, hence, the different MAS NMR lines could be more easily quantified (Figure 4 and Table 1). 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.
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 29 Si MAS NMR spectrum of alkali brick ( Figure 4) revealed Q 4 (0Al) units in quartz [47,48]. The low intensity of this signal in our 29 Si 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 29 Si resonances. In addition to the Q 4 (0Al) signal corresponding to the quartz mineral, other Q 4 (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 Q 4 (0Al) [45,49].
From 29 Si NMR data (see Table 1), it was possible to estimate the silica contribution of un-reacted aluminosilicate gel, Si(gel)%, to the overall 29 Si MAS NMR peak of "zeolitized" brick-metakaolinite from the expression (in atomic %): where I%(gel) and I%(zeolites) represent the intensity percentages of 29 Si resonances ascribed respectively to aluminosilicate gel and zeolites in the 29 Si 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 Q 4 (4Al), Q 4 (2Al), Q 4 (2-1Al) units in aluminosilicate gel(s) [41,43]; and (ii) we did not take into account polymerized amorphous silica structures with Q 4 (0Al) (i.e., the 29 Si NMR peaks at −113.3 ppm and −119.5 ppm, see Table 1). We found at least 70 atomic % of 29 Si nuclei present in the "aluminosilicate" gel(s) of the partially zeolitized brick. From the quantification of the different Q 4 (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].
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) 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 Figures 1 and 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.  Figure S3A; b see Figure S3B; c see Figure 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 1 H 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 δ 1 H range of~1-3 ppm, they were either nonacidic or weakly acidic hydroxyl groups. These were assigned to isolated silanol/aluminol protons in brick minerals.
Ceramics2023, 6, FOR PEER REVIEW 11 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 1 H 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 δ 1 H range of ∼1-3 ppm, they were either nonacidic 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 1 H 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, 23 Na 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): Note that the residual 23 Na signals observed in Figure 6B might be due to structural sodium bound to other brick minerals. After the addition of an NH 4 Cl solution ([NH 4 Cl] = 0.5 mol·L −1 ) into a treated-brick suspension, the 1 H 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 NH 4 + ions which were adsorbed onto the Brønsted sites of the treated brick, leading to NH 4 + -zeolite complexes [51][52][53]. As for Na + ions adsorbed onto treated brick, 23 Na MAS NMR analysis indicated their near disappearance after NH 4 Cl 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 + + NH 4 + → >SO − NH 4 + + Na + Note that the residual 23 Na 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 1 H and 23 Na MAS NMR spectroscopy. The 1 H MAS NMR spectrum of the acidified brick revealed the presence of a sharp peak at δ = 4.7 ppm ascribed to H 3 O + 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]. 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 (pHPZC = 5.85) or the pH at the iso-electric point (pHIEP = 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 1 H and 23 Na MAS NMR spectroscopy. The 1 H 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 nSi/nAl = 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 δ1H = 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].   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 (pHPZC = 5.85) or the pH at the iso-electric point (pHIEP = 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 1 H and 23 Na MAS NMR spectroscopy. The 1 H 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 nSi/nAl = 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 δ1H = 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 NH 4 Cl solution ([NH 4 Cl] = 0.5 mol·L −1 ). The 1 H MAS NMR spectrum of the recovered sample displayed three new proton resonances ( Figure 7B). The peaks at δ 1 ∼ = 7.3 ppm and δ 1 ' = 7.0 ppm were assigned both to NH 4 + ions adsorbed onto Brønsted sites of framework zeolites: LTA and NaP (note that the δ 1 and δ 1 ' peaks were not yet assigned to specific zeolites). Whereas the peak at δ 2 = 5.3 ppm was attributed to ammonium species involved in fast exchanges with NH 4 + (δ 1 and δ 1 ') 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 1 H MAS NMR spectrum then exhibited a line broadening and 1 H 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 1 H nuclei along bridging Si−O − −Al chains. The resulting merging signal was centered at δ 3 = 5.07 ppm. Note that the weak resonance detected at the lower field in the 1 H MAS NMR spectrum (δ 1 = 6.75 ppm) was still ascribed to the own protons of ammonium ions directly bound to bridging Si−O − −Al sites (as Si−O − NH 4 + −Al).
NMR spectrum then exhibited a line broadening and 1 H 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 1 H 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 1 H 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 1 H 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 1 H Solidstate 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 1 H, 23 Na and 27 Al MAS NMR analyses permitted to reveal a progressive decomposition of brick zeolites as the medium pH decreased below 5 (see Figure S2).  It could then be contended that upon H 3 O + /NH 4 + exchange, the resonance of zeolitic 1 H nuclei were low-field-shifted by ∆δ 1 = 2.45 ppm (from 4.70 to~7.15) and ∆δ 2 = 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 1 H Solidstate 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 1 H, 23 Na and 27 Al MAS NMR analyses permitted to reveal a progressive decomposition of brick zeolites as the medium pH decreased below 5 (see Figure S2).  In the 27 Al MAS NMR spectrum of an alkali-brick sample, most aluminum atoms are tetrahedrally coordinated (Al IV ), which results in three 27 Al 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 29 Si 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 δ 27 Al 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 29  In the 27 Al MAS NMR spectrum of an alkali-brick sample, most aluminum atoms are tetrahedrally coordinated (Al IV ), which results in three 27 Al 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 29 Si atoms to give Al(OSi) 4 units) of both NaA and NaP zeolites [34,[59][60][61][62][63][64][65].

27 Al MAS NMR
distinct aluminosilicate frameworks and would be differentially affected by structura tors, i.e., bond angles/lengths in T-O-T, with T = Si or Al [14,58]. In the 27 Al MAS NMR spectrum of an alkali-brick sample, most aluminum atom tetrahedrally coordinated (Al IV ), which results in three 27 Al MAS NMR signals, see Fi 10A. The particular NMR resonances detected in the chemical shift range 57-61 ppm Figure 10A) were partially attributed to tetrahedrally coordinated aluminum (boun oxygen to four 29 Si atoms to give Al(OSi)4 units) of both NaA and NaP zeolites [34,59 According to the literature [34,[63][64][65], a slightly lower chemical shift (about 1 p in the δ 27 Al range 57.6-59.3 ppm was more expected for NaP than for NaA. As fo NMR resonance at 63.2 ppm (see Figure 10A), it was ascribed to tetrahedrally coordin aluminum in geo-polymeric gels, which were not transformed into zeolites [66].  According to the literature [34,[63][64][65], a slightly lower chemical shift (about 1 ppm) in the δ 27 Al 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] As for 29 Si 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 Q 4 (3Al), Q 4 (2Al) and Q 4 (1Al) units, respectively; According to Ramdas-Klinowski's works [69], the 29 Si chemical shifts of brick gels could be correlated linearly with the structural parameter, ∑d TT , according to: δ( 29 Si) = 143.03 − 20.34·( ∑ d TT /Å) (7) in which ∑d TT is defined as: 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 Q 4 (3Al), Q 4 (2Al) and Q 4 (1Al) units of brick gel(s), see Table 3. Table 3. Chemical shifts, δ( 29 Si) of 29 Si nuclei of the different Q 4 (mAl) units present in modifiedbrick frameworks. From these angle values, it was possible to assess the corresponding chemical shifts of 27 Al nuclei present in the different Q 4 (mAl) units of brick gel(s) by employing Equation (6). We found δ( 27 Al) = 61.4 ppm, 61.8 ppm and 61.0 ppm for Q 4 (3Al), Q 4 (2Al) and Q 4 (1Al) units of brick gel(s), respectively. These δ( 27 Al) values were consistent with the broad NMR resonances ranging from~59 to~61.5 ppm, which were observed in the 27 Al 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 Q 4 (4Al) units (see Table 3); the corresponding chemical shift of 27 27 Al nuclei (δ( 27 Al) = 58.3 ppm) was found to be slightly higher than that determined above. δ( 29 Si) = −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 ). 27 Al 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 NH 4 Cl treatment, the 27 Al MAS NMR spectrum of the recovered material changed noticeably in comparison with that of the alkali-brick (Figures 9 and 10). The incorporation of NH 4 + 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 + /NH 4 + exchange at the brick surface). As observed by 27 Al 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 27 AlO 4 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 NH 4 + ions with Q 4 AlO 4 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 NH 4 + ions onto alkali-brick resulted in acidification of brick particles, as evidenced by the significant shifting towards lower field observed by 1 H 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 − H 3 O + NH 3 , as previously suggested [51][52][53]56,57]. In this assumption, the mean Si-O-Al angles of NH 4 + -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 • .

2D MAS NMR ( 27 Al-{ 1 H} D-HMQC)
In order to observe 27 Al-1 H dipolar-mediated correlations, 2D 27 Al( 1 H) 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 27 Al( 1 H) D-HMQC spectrum: one between 1 H nuclei of NH 4 + (δ 1H = 7.0 ppm) and Al IV (δ 27Al ∼ = 64 ppm); and the other between 1 H nuclei of H 3 O + (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 1 H NMR signal, as already observed previously for different types of zeolites [52,53,56,57]. Moreover, the 1 H 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.: δ1H = 7.0 ppm). This could be explained by the fact that the hydrogen-exchange rate between 1 H 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 1 H 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 paragraph 3.3.2 and Figure 8).

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 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). The observation of an Al IV −NH 4 + 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 1 H NMR signal, as already observed previously for different types of zeolites [52,53,56,57]. Moreover, the 1 H MAS NMR signal reached a maximal intensity at about 5 ppm. At this resonance maximum, some of the NH 4 + 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 NH 4 + /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 1 H 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 1 H 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).

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 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 H 3 O + 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 Cd 2+ , Fe 2+ , Ni 2+ and Pb 2+ (Me 2+ ), were previously studied thermodynamically on the basis of both the general properties of divalent metals in water and the global (zeolitic and geopolymeric) 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 + /Me 2+ ion-exchange reactions revealed that the replacement of the exchangeable cation (Na + ) on the composite by Me 2+ 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). 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 Cd 2+ , Fe 2+ , Ni 2+ and Pb 2+ (Me 2+ ), 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 + /Me 2+ ion-exchange reactions revealed that the replacement of the exchangeable cation (Na + ) on the composite by Me 2+ 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(H 2 O) 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.
of both the general properties of divalent metals in water and the global (zeolitic and geopolymeric) 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 + /Me 2+ ion-exchange reactions revealed that the replacement of the exchangeable cation (Na + ) on the composite by Me 2+ 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 Figure 13. Histograms representing the evolution of the amounts of Si, Al, Na and Cs (or Rb) released in the leachates during the acid leaching (at pH 2 and room temperature) of the composite no-calcined and calcined at 600 °C and 1000 °C for 4 h.
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 innersphere "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.

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- No calcined 600°C 1000°C Figure 13. Histograms representing the evolution of the amounts of Si, Al, Na and Cs (or Rb) released in the leachates during the acid leaching (at pH 2 and room temperature) of the composite no-calcined and calcined at 600 • C and 1000 • C for 4 h.

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 geopolymeric 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 29 Si, 27 Al and 1 H nuclei environments existing in brick-composite networks. 29 Si and 27 Al 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. 1 H MAS NMR studies showed the existence of bridging Si−O − H 3 O + −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 27 Al( 1 H) D-HMQC NMR. Detailed spectral analyses of 1 H, 29 Si and 27 Al 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 + , H 3 O + and NH 4 + ) 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).

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 + , Sr 2+ and Co 2+ ) 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.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/ceramics6030108/s1, Figure S1: determination of the point of zero charge and isoelectric point of alkali brick; Figure S2

Data Availability Statement:
The authors confirm that the data supporting the findings of this study are available within the article.