Municipal Solid Waste Incineration Bottom Ash as Sole Precursor in the Alkali-Activated Binder Formulation

The concern about the large amount of weathered bottom ash (WBA) produced in waste-to-energy plants (WtE) has caused an increased search for alternatives to reduce their environmental impact. The present study aims to provide an added value through the WBA valorization from municipal solid waste incineration (MSWI) for its use as a sole precursor for developing alkali-activated binders (AABs). Alkali-activated weathered bottom ash binders (AA-WBA) were formulated with a liquid-to-solid ratio of 1.0 and using sodium silicate (80 wt.%) and NaOH (20 wt.%) at different concentrations (2, 4, 6, and 8M) as alkali-activator solutions. AA-WBA were cured at room temperature to extend their applicability. The effect of the alkali-activator solution molarity on the final properties of the AA-WBA was evaluated. The physicochemical characterization by XRD, FTIR, and SEM evidenced the presence of the typical phases (calcium silicate hydrate and gehlenite) of C-(A)-S-H gel. Leaching concentrations of As, Cu, and Mo exceed the acceptance in landfills for inert waste, while the leaching concentration of Sb exceeds the one for non-hazardous waste. The structure of the binders depends on the alkalinity of the activator, obtaining better results using NaOH 6M in terms of microstructure and compressive strength (6.7 MPa). The present study revealed that AA-WBA for non-structural purposes can be obtained. The AA-WBA formulation contributes to the WBA valorization and development of low-carbon cements; therefore, it is an encouraged alternative to ordinary Portland cement (OPC). Considering the amounts and costs of the WBA, sodium silicate, NaOH, and water, the total cost of the developed formulations is comprised in a range between 137.6 and 153.9 €/Tn.


Introduction
Municipal solid waste incineration (MSWI) in waste-to-energy (WtE) plants is one of the main processes for municipal solid waste (MSW) management [1]. MSWI contributes to the energy recovery from MSW, and the reduction of its volume (by 90%) and weight (by 75%) [2]. It is extensively applied in many European countries. According to the information provided by the Confederation of European Waste-to-EnergyPlants (CEWEP) [3], 492 WtE plants are operating in Europe [4]. In 2017, around 70 Mt of MSW were incinerated in the EU (EU-28) [5]. The main by-product produced during the MSW treatment in WtE plants is the incinerator bottom ash (IBA). IBA represents 85% of the solid resulting from combustion [6] and it is classified as a hazardous (EWC 19 01 11*-asterisk means hazardous) or non-hazardous waste (EWC 19 01 12) depending on its concentration of hazardous and increases the AA-WBA applicability. This work focuses on the effect of NaOH concentration from a physicochemical, physical, mechanical, and environmental point of view on the resultant binders.

Materials and Methods
The WBA sample was provided by the VECSA company, which is responsible for valorizing the IBA collected from the WtE plant located in Tarragona (Catalonia, Spain), with a capacity of 140 kt·y −1 . The feed stream treated in this incineration plant is mainly composed of household rubbish, with a small input from commercial sources. Around 32 kt·y −1 of fresh IBA is obtained in this WtE plant. The combustion temperature is 950 • C [42]. After combustion, fresh IBA is further processed in a conditioning plant to recover some valuable materials (ferrous and non-ferrous metals) and to obtain a homogenized granular material. Finally, the resultant IBA is subjected to natural weathering outdoors for at least 3 months to stabilize heavy metal(loid)s and to obtain WBA, which is reused as secondary aggregate. Because of the heterogeneity of WBA, 100 kg were collected from stockpiles and then homogenized and stored in 30 L plastic containers. The alkali activator used consisted of a commercial Na 2 SiO 3 solution (Scharlab S.L.; 26.44 wt.% of SiO 2 and 8.21 wt.% of Na 2 O; ρ = 1.37 g·cm −3 ) and NaOH solutions prepared by using NaOH pearls (Labbox Labware S.L.; purity > 98%) dissolved in distilled water: 2 (ρ = 1.08 g·cm −3 ), 4 (ρ = 1.16 g·cm −3 ), 6 (ρ = 1.20 g·cm −3 ), and 8M (ρ = 1.24 g·cm −3 ).
The preparation to obtain the WBA powder started by quartering the whole 100 kg sample to acquire a representative sub-sample of 10 kg. The quartered sample was dried in a stove at 105 • C for 24 h. Then, the dried sample was sieved (by using standard sieves: 32, 16, 8, 4, 2, 1, 0.5, 0.125, 0.063 mm) to determine the particle size distribution (PSD), as shown in Figure 1. Subsequently, the WBA was crushed and milled until a powder below 80 µm in particle size was obtained. Finally, a metal magnet (Nd; 0.485 T) was passed over the ground sample to remove magnetic particles.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 3 of 16 from a physicochemical, physical, mechanical, and environmental point of view on the resultant binders.

Materials and Methods
The WBA sample was provided by the VECSA company, which is responsible for valorizing the IBA collected from the WtE plant located in Tarragona (Catalonia, Spain), with a capacity of 140 kt·y −1 . The feed stream treated in this incineration plant is mainly composed of household rubbish, with a small input from commercial sources. Around 32 kt·y −1 of fresh IBA is obtained in this WtE plant. The combustion temperature is 950 °C [42]. After combustion, fresh IBA is further processed in a conditioning plant to recover some valuable materials (ferrous and non-ferrous metals) and to obtain a homogenized granular material. Finally, the resultant IBA is subjected to natural weathering outdoors for at least 3 months to stabilize heavy metal(loid)s and to obtain WBA, which is reused as secondary aggregate. Because of the heterogeneity of WBA, 100 kg were collected from stockpiles and then homogenized and stored in 30 L plastic containers. The alkali activator used consisted of a commercial Na2SiO3 solution (Scharlab S.L.; 26.44 wt.% of SiO2 and 8.21 wt.% of Na2O; ρ = 1.37 g·cm −3 ) and NaOH solutions prepared by using NaOH pearls (Labbox Labware S.L.; purity > 98%) dissolved in distilled water: 2 (ρ = 1.08 g·cm −3 ), 4 (ρ = 1.16 g·cm −3 ), 6 (ρ = 1.20 g·cm −3 ), and 8M (ρ = 1.24 g·cm −3 ).
The preparation to obtain the WBA powder started by quartering the whole 100 kg sample to acquire a representative sub-sample of 10 kg. The quartered sample was dried in a stove at 105 °C for 24 h. Then, the dried sample was sieved (by using standard sieves: 32, 16, 8, 4, 2, 1, 0.5, 0.125, 0.063 mm) to determine the particle size distribution (PSD), as shown in Figure 1. Subsequently, the WBA was crushed and milled until a powder below 80 µm in particle size was obtained. Finally, a metal magnet (Nd; 0.485 T) was passed over the ground sample to remove magnetic particles. The elemental oxide composition of WBA was assessed by X-ray fluorescence (XRF) analysis with a spectrophotometer Panalytical Philips PW 2400 sequential X-ray equipped with the software UniQuant ® V5.0. Major oxides are given in Table 1. There was a high content of SiO2, CaO, Al2O3, and Na2O, which are the main compounds used to obtain AACs [43]. The high SiO2 content was due to the presence of primary and secondary glass in WBA [6,44]. The CaO and Al2O3 content was because WBA contains cementitious materials based on OPC, as well as synthetic ceramics from small domestic works. It is important to highlight that the percentage shown in Table 1 represents the total The elemental oxide composition of WBA was assessed by X-ray fluorescence (XRF) analysis with a spectrophotometer Panalytical Philips PW 2400 sequential X-ray equipped with the software UniQuant ® V5.0. Major oxides are given in Table 1. There was a high content of SiO 2 , CaO, Al 2 O 3 , and Na 2 O, which are the main compounds used to obtain AACs [43]. The high SiO 2 content was due to the presence of primary and secondary glass in WBA [6,44]. The CaO and Al 2 O 3 content was because WBA contains cementitious materials based on OPC, as well as synthetic ceramics from small domestic works. It is important to highlight that the percentage shown in Table 1 represents the total content of Appl. Sci. 2020, 10, 4129 4 of 15 the oxides considering both their amorphous and crystalline phases. The SiO 2 and Al 2 O 3 availability (reactive phases) of the WBA sample are reported in previous studies by the authors [37].   The proportions of the alkali-activator solution, alkali dosage, and silicate modulus (Ms) are given in Table 2. The formulation's optimal stoichiometry was determined after carrying out a preliminary study to delimit the proportion of raw materials and to obtain the best mechanical performance of AA-WBA binders cured at room temperature. The same amount of WBA, NaOH, and Na2SiO3 solutions (referred as liquid; L) were used in each formulation. The concentration of the NaOH solutions (2, 4, 6, and 8M) was varied to evaluate its effect on the AA-WBA final properties and structure of the cementitious matrix. The concentrations of NaOH were chosen considering the previous studies carried out by the authors [41].  The proportions of the alkali-activator solution, alkali dosage, and silicate modulus (Ms) are given in Table 2. The formulation's optimal stoichiometry was determined after carrying out a preliminary study to delimit the proportion of raw materials and to obtain the best mechanical performance of AA-WBA binders cured at room temperature. The same amount of WBA, NaOH, and Na 2 SiO 3 solutions (referred as liquid; L) were used in each formulation. The concentration of the NaOH solutions (2, 4, 6, and 8M) was varied to evaluate its effect on the AA-WBA final properties and structure of the cementitious matrix. The concentrations of NaOH were chosen considering the previous studies carried out by the authors [41]. The AA-WBA binders' preparation procedure began by mixing the NaOH and Na 2 SiO 3 (ratio of 1:4 by weight). This ratio was selected because a higher proportion of NaOH increases the porosity and reduces the mechanical performance of the binders [33]. Then, the WBA and the alkali-activator solution were mixed (liquid-to-solid ratio of 1:1 by weight) and mechanically stirred in a plastic beaker for 5 min. Fresh AA-WBA binders were poured into 25 mm cubic molds and sealed in plastic bags for 3 days in a climate chamber (25 ± 1 • C and relative humidity of 95% ± 5%). After 3 days, the samples were unmolded and removed from the plastic sealed bag, and the AA-WBA specimens continued curing in the climate chamber under the same above-mentioned conditions. Nine cubic-shaped specimens were prepared for each formulation. The AA-WBA specimens' characterization was carried out after 28 curing days.
The hydrolytic stability of the AA-WBA binders was evaluated by introducing a dried cubic specimen of each formulation in boiling water for 20 min [39]. Afterwards, the specimen remained at room temperature until reaching a constant weight in a desiccator with silica gel. Then, before and after the test, the sample was weighed to determine the mass loss and to verify the chemical stability and resistance to the dissolution of the AA-WBA binders. The integrity of AA-WBA was assessed by introducing a cubic specimen in deionized water (liquid-to-specimen ratio of 10.0 by weight) under stirring for 2 days. Conductivity (k) and pH were measured (after 1, 15, 30, 60, 120, 240, 600, 1440, 2160, and 2880 min) to evaluate pH variation and ion diffusion over time and the effect that may have on the formation of cementitious phases.
The main reaction product of alkali activation of calcium-rich precursors such as WBA [41] is calcium silicate hydrate (C-(A)-S-H) gel [39,45]. C-(A)-S-H gel coexists in the AAC's microstructures along with the N-A-S-H gel [46]. It has even been reported that a mixture of both gels improves the durability of AACs as compared to the single N-A-S-H gel [47]. However, when both formed gels coexist, it is difficult to conduct a proper elucidation of C-(A)-S-H using classical characterization techniques such as X-ray diffraction (XRD) or Fourier transform infrared spectroscopy (FTIR). For this reason, selective dissolution characterization is gaining popularity in order to eliminate any ambiguity between C-(A)-S-H and N-A-S-H gels [46]. AA-WBA binders were subjected to the salicylic acid/methanol (SAM) extraction [43] and HCl extraction [44] to determine C-(A)-S-H and N-A-S-H phases' content. The calcium-containing phases were dissolved in the SAM medium after the attack, while phases without calcium remained in the insoluble residue. The SAM extraction procedure consisted of mixing 1 g of grounded sample with salicylic acid (6 g) and methanol (40 mL) for 1 h. The mixture was subsequently filtered (Whatman filter with 20 µm pore size) to obtain the insoluble residue. The HCl extraction consisted of stirring 1 g of AA-WBA binder with 250 mL of HCl (1:20) solution for 3 h, followed by filtration (Whatman filter with 20 µm pore size). Then, the insoluble residues were washed with deionized water and dried in a desiccator until reaching a constant weight. The percentage of weight loss due to SAM and HCl extraction was calculated by weighing the insoluble residue.
XRD analysis was performed to determine the crystalline phases of AA-WBA binders. FTIR was used to determine the reaction products of alkali-activated WBA by means of various Si-O-X (X=Al, Si) stretching peaks' identification, which are assigned to different chemical structures of silicate phases. Spectrum Two™ equipment from Perkin Elmer was used. The insoluble residue obtained in SAM and HCl extraction was also analyzed by FTIR to identify possible changes in the AA-WBA binders' composition and elucidate the co-existence of both C-(A)-S-H and N-A-S-H gels. A scanning electron microscopy (SEM) technique was conducted by an ESEM FEI Quanta 200 to evaluate the AA-WBA binders' microstructure. A planar sample (1.5 mm thickness) of each formulation was obtained by means of a diamond disc (140 rpm) cutter, and coated with graphite.
Bulk density and open porosity were studied by using 3 cubic shape specimens for each formulation. The values were determined following the standard EN 1936:2006. After 28 curing days of the AA-WBA specimens, compressive strength (σ c ) tests were performed using Incotecnic MULTI-R1 equipment. Three tests were performed for each formulation. A progressive load until fracture was applied with a loading rate of 240 kg·s −1 .
Leaching tests in deionized water for 24 h were conducted according to European standard EN 12457-2 to evaluate the potential release of heavy metal(loid)s from WBA and AA-WBA binders. A PerkinElmer ELAN 6000 ICP mass spectrometry (ICP-MS) device was used to analyze heavy metal(loid)s (As, Ba, Cd, Cr, Cu, Hg, Mo, Pb, Ni, Sb, and Zn) in the obtained eluates.
The cost of AA-WBA binders (€·t −1 ) considering the cost of the raw materials (Table 3) used for their formulation was calculated.

Hydrolytic Stability Test and Integrity Test
The boiling water test demonstrated the AA-WBA binders' resistance to hydrolytic degradation. The specimen mass loss percentage was lower than 3% in all formulations. The AA-WBA binders also remain consolidated after the integrity test. Figure 3a shows the pH results of AA-WBA binders. There was a significant pH increase at the beginning of the integrity test due to the alkalinity of all samples. The pH values were stabilized after 120 min around a range between 10.75 and 11.25. The conductivity values ( Figure 3b) increased gradually during the test because of the contribution of Na + and OH − ions found in unreacted NaOH and Na 2 SiO 3 . The presence of Cl − ions in WBA also led to an increase of conductivity. As expected in both cases, the higher the concentration used to formulate the AA-WBA binder, the higher the pH and conductivity values obtained. These results demonstrate the chemical stability of AA-WBA binders.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 6 of 16 R1 equipment. Three tests were performed for each formulation. A progressive load until fracture was applied with a loading rate of 240 kg·s −1 . Leaching tests in deionized water for 24 h were conducted according to European standard EN 12457-2 to evaluate the potential release of heavy metal(loid)s from WBA and AA-WBA binders. A PerkinElmer ELAN 6000 ICP mass spectrometry (ICP-MS) device was used to analyze heavy metal(loid)s (As, Ba, Cd, Cr, Cu, Hg, Mo, Pb, Ni, Sb, and Zn) in the obtained eluates.
The cost of AA-WBA binders (€·t −1 ) considering the cost of the raw materials (Table 3) used for their formulation was calculated.

Hydrolytic Stability Test and Integrity Test
The boiling water test demonstrated the AA-WBA binders' resistance to hydrolytic degradation. The specimen mass loss percentage was lower than 3% in all formulations. The AA-WBA binders also remain consolidated after the integrity test. Figure 3a shows the pH results of AA-WBA binders. There was a significant pH increase at the beginning of the integrity test due to the alkalinity of all samples. The pH values were stabilized after 120 min around a range between 10.75 and 11.25. The conductivity values (Figure 3b) increased gradually during the test because of the contribution of Na + and OHions found in unreacted NaOH and Na2SiO3. The presence of Clions in WBA also led to an increase of conductivity. As expected in both cases, the higher the concentration used to formulate the AA-WBA binder, the higher the pH and conductivity values obtained. These results demonstrate the chemical stability of AA-WBA binders. conductivity during the integrity test.

Selective Chemical Extractions
The SAM and HCl extraction results are given in Table 4. The percentage of mass dissolved by SAM extraction in OPC paste compared with that of AA-WBA binders was higher due to its composition based in C-S-H phases. The main contribution of C-(A)-S-H phases dissolved in WBA

Selective Chemical Extractions
The SAM and HCl extraction results are given in Table 4. The percentage of mass dissolved by SAM extraction in OPC paste compared with that of AA-WBA binders was higher due to its composition based in C-S-H phases. The main contribution of C-(A)-S-H phases dissolved in WBA powder probably came from the residues of small construction works in the domestic sphere. The most C-(A)-S-H formation in AA-WBA occurred in the AA-WBA-6M sample, as it showed the maximum value of dissolved mass by SAM extraction. A slight reduction of dissolved mass by SAM extraction in the AA-WBA-8M sample compared with that of the AA-WBA-6M sample was also observed. These results reveal that the NaOH 6M was the maximum NaOH concentration from which the precipitation of Ca 2+ in form of C-(A)-S-H was disadvantaged against the formation of aluminosilicate gels [49]. The HCl extraction results show that increases in the NaOH concentration of AA-WBA binders resulted in an increase in the dissolved mass. The selective chemical extraction by HCl led to the dissolution of calcium and sodium carbonate phases [50], sodium aluminosilicate gels, and zeolites [44], as well as C-(A)-S-H phases leading to the removal of Ca 2+ and leaving silica gel as an insoluble residue [46]. The results of HCl extraction indicate that the higher the NaOH concentration, the higher mass dissolved. This fact is due to the formation of C-A-S-H gels as demonstrated in SAM extraction, as well as the increased presence of sodium carbonate phases as the alkali dosage was increased ( Table 2).

Physicochemical Characterization
The analysis of the XRD patterns of the WBA and AA-WBA binders shows the presence of new crystalline phases. The appearance of new peaks or variations demonstrates the formation of new crystalline phases in the AA-WBA binders. This can be observed in the shaded areas of Figure 4, where the XRD pattern of WBA compared with that of the AA-WBA-6M sample is shown.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 7 of 16 which the precipitation of Ca 2+ in form of C-(A)-S-H was disadvantaged against the formation of aluminosilicate gels [49]. The HCl extraction results show that increases in the NaOH concentration of AA-WBA binders resulted in an increase in the dissolved mass. The selective chemical extraction by HCl led to the dissolution of calcium and sodium carbonate phases [50], sodium aluminosilicate gels, and zeolites [44], as well as C-(A)-S-H phases leading to the removal of Ca 2+ and leaving silica gel as an insoluble residue [46]. The results of HCl extraction indicate that the higher the NaOH concentration, the higher mass dissolved. This fact is due to the formation of C-A-S-H gels as demonstrated in SAM extraction, as well as the increased presence of sodium carbonate phases as the alkali dosage was increased ( Table 2).

Physicochemical Characterization
The analysis of the XRD patterns of the WBA and AA-WBA binders shows the presence of new crystalline phases. The appearance of new peaks or variations demonstrates the formation of new crystalline phases in the AA-WBA binders. This can be observed in the shaded areas of Figure 4, where the XRD pattern of WBA compared with that of the AA-WBA-6M sample is shown. In the shaded areas of Figure 5a,b, some variation on the peaks' intensity between the AA-WBA binders is appreciated, therefore indicating that the amount of these new crystalline phases increases as the NaOH concentration increases. In the shaded areas of Figure 5a,b, some variation on the peaks' intensity between the AA-WBA binders is appreciated, therefore indicating that the amount of these new crystalline phases increases as the NaOH concentration increases.  PDF# 029-0380). In the AA-WBA-8M sample, gismondine (CaAl2Si2O8·4H2O; PDF# 020-0452) and gaylussite ((Na2Ca(CO3)2·5H2O; PDF# 020-1088) were also identified. All the aforementioned compounds are related with C-(A)-S-H phases excluding gaylussite, which is a sodium carbonate compound that reveals the cation exchange between precursor and alkali-activator solution [51]. It is important to note that XRD results agree with the results obtained in the selective chemical extractions. Depicted in Figure 7 is the WBA FTIR spectrum compared to the FTIR spectra of the AA-WBA binders, which was used to find evidence of the alkali-activation of the WBA. The broad band at 984 cm −1 ascribed to T-O stretching vibrations (where T=Si or Al) was displaced towards higher frequencies, indicating the formation of C-(A)-S-H gel [52]. With respect to the AA-WBA spectra, a shift of the main broad band towards lower frequencies was observed. This fact is probably due to both the inclusion of aluminum on the C-S-H gel [53], which is consistent with chemical extraction and XRD results. There was a weak peak at 780 cm −1 associated with Si-O-Si bridging bonds in quartz (SiO2) [54]. The peak mentioned above was indiscernible in AA-WBA-6M and AA-WBA-8M samples.  PDF# 029-0380). In the AA-WBA-8M sample, gismondine (CaAl 2 Si 2 O 8 ·4H 2 O; PDF# 020-0452) and gaylussite ((Na 2 Ca(CO 3 ) 2 ·5H 2 O; PDF# 020-1088) were also identified. All the aforementioned compounds are related with C-(A)-S-H phases excluding gaylussite, which is a sodium carbonate compound that reveals the cation exchange between precursor and alkali-activator solution [51]. It is important to note that XRD results agree with the results obtained in the selective chemical extractions.  PDF# 029-0380). In the AA-WBA-8M sample, gismondine (CaAl2Si2O8·4H2O; PDF# 020-0452) and gaylussite ((Na2Ca(CO3)2·5H2O; PDF# 020-1088) were also identified. All the aforementioned compounds are related with C-(A)-S-H phases excluding gaylussite, which is a sodium carbonate compound that reveals the cation exchange between precursor and alkali-activator solution [51]. It is important to note that XRD results agree with the results obtained in the selective chemical extractions. Depicted in Figure 7 is the WBA FTIR spectrum compared to the FTIR spectra of the AA-WBA binders, which was used to find evidence of the alkali-activation of the WBA. The broad band at 984 cm −1 ascribed to T-O stretching vibrations (where T=Si or Al) was displaced towards higher frequencies, indicating the formation of C-(A)-S-H gel [52]. With respect to the AA-WBA spectra, a shift of the main broad band towards lower frequencies was observed. This fact is probably due to both the inclusion of aluminum on the C-S-H gel [53], which is consistent with chemical extraction and XRD results. There was a weak peak at 780 cm −1 associated with Si-O-Si bridging bonds in quartz (SiO2) [54]. The peak mentioned above was indiscernible in AA-WBA-6M and AA-WBA-8M samples. Depicted in Figure 7 is the WBA FTIR spectrum compared to the FTIR spectra of the AA-WBA binders, which was used to find evidence of the alkali-activation of the WBA. The broad band at 984 cm −1 ascribed to T-O stretching vibrations (where T=Si or Al) was displaced towards higher frequencies, indicating the formation of C-(A)-S-H gel [52]. With respect to the AA-WBA spectra, a shift of the main broad band towards lower frequencies was observed. This fact is probably due to both the inclusion of aluminum on the C-S-H gel [53], which is consistent with chemical extraction and XRD results. There was a weak peak at 780 cm −1 associated with Si-O-Si bridging bonds in quartz (SiO 2 ) [54]. The peak mentioned above was indiscernible in AA-WBA-6M and AA-WBA-8M samples. This fact probably means that a large proportion of quartz had reacted to form new phases due to the higher alkalinity of the alkali-activator solution used. The strong band at 1429 cm −1 , assigned to the stretching mode of carbonates, as well as sharp peaks at 875 cm −1 and 713 cm −1 which were related to the bending mode of carbonates, were less intense in AA-WBA binders [54].
Appl. Sci. 2020, 10, x FOR PEER REVIEW 9 of 16 This fact probably means that a large proportion of quartz had reacted to form new phases due to the higher alkalinity of the alkali-activator solution used. The strong band at 1429 cm −1 , assigned to the stretching mode of carbonates, as well as sharp peaks at 875 cm −1 and 713 cm −1 which were related to the bending mode of carbonates, were less intense in AA-WBA binders [54]. The FTIR spectra of WBA before and after SAM is shown in Figure 8a. The band at 984 cm −1 associated with T-O stretching vibrations (where T=Si or Al) was shifted to 1003 cm −1 , which means that C-(A)-S-H phases were dissolved by SAM extraction [55]. The strong band at 1429 cm −1 , assigned to the stretching mode of carbonates, as well as sharp peaks at 875 cm −1 and 713 cm −1 related to the bending mode of carbonates, did not present any changes. Figure 8b shows the FTIR spectra of the AA-WBA-6M sample before and after SAM and HCl extraction. This sample was chosen because it was the one that would potentially present better results. A more pronounced shifting (995 to ≈1065 cm −1 ) in the band associated with T-O stretching vibrations (where T=Si or Al) could be observed because of the large amount of C-(A)-S-H phases in the AA-WBA-6M sample. Both in SAM extraction and HCl extraction spectra, three peaks (≈790, ≈930, ≈1050 cm −1 ) and a shoulder (≈1162 cm −1 ) were also observed, which are characteristic of silica gels [56,57].  The FTIR spectra of WBA before and after SAM is shown in Figure 8a. The band at 984 cm −1 associated with T-O stretching vibrations (where T=Si or Al) was shifted to 1003 cm −1 , which means that C-(A)-S-H phases were dissolved by SAM extraction [55]. The strong band at 1429 cm −1 , assigned to the stretching mode of carbonates, as well as sharp peaks at 875 cm −1 and 713 cm −1 related to the bending mode of carbonates, did not present any changes. Figure 8b shows the FTIR spectra of the AA-WBA-6M sample before and after SAM and HCl extraction. This sample was chosen because it was the one that would potentially present better results. A more pronounced shifting (995 to ≈1065 cm −1 ) in the band associated with T-O stretching vibrations (where T=Si or Al) could be observed because of the large amount of C-(A)-S-H phases in the AA-WBA-6M sample. Both in SAM extraction and HCl extraction spectra, three peaks (≈790, ≈930, ≈1050 cm −1 ) and a shoulder (≈1162 cm −1 ) were also observed, which are characteristic of silica gels [56,57].
Appl. Sci. 2020, 10, x FOR PEER REVIEW 9 of 16 This fact probably means that a large proportion of quartz had reacted to form new phases due to the higher alkalinity of the alkali-activator solution used. The strong band at 1429 cm −1 , assigned to the stretching mode of carbonates, as well as sharp peaks at 875 cm −1 and 713 cm −1 which were related to the bending mode of carbonates, were less intense in AA-WBA binders [54]. The FTIR spectra of WBA before and after SAM is shown in Figure 8a. The band at 984 cm −1 associated with T-O stretching vibrations (where T=Si or Al) was shifted to 1003 cm −1 , which means that C-(A)-S-H phases were dissolved by SAM extraction [55]. The strong band at 1429 cm −1 , assigned to the stretching mode of carbonates, as well as sharp peaks at 875 cm −1 and 713 cm −1 related to the bending mode of carbonates, did not present any changes. Figure 8b shows the FTIR spectra of the AA-WBA-6M sample before and after SAM and HCl extraction. This sample was chosen because it was the one that would potentially present better results. A more pronounced shifting (995 to ≈1065 cm −1 ) in the band associated with T-O stretching vibrations (where T=Si or Al) could be observed because of the large amount of C-(A)-S-H phases in the AA-WBA-6M sample. Both in SAM extraction and HCl extraction spectra, three peaks (≈790, ≈930, ≈1050 cm −1 ) and a shoulder (≈1162 cm −1 ) were also observed, which are characteristic of silica gels [56,57].   Figure 9 shows the SEM images of AA-WBA binders in backscattering electron (BSE) mode. The effect of alkali-activator concentration in the compactness and microstructure of the material is demonstrated. Two differentiated areas in all samples can be observed; a light grayish compact area which was essentially attributed to the C-(A)-S-H gel [52]; and a dark grayish disaggregated area composed of unreacted WBA. In the AA-WBA-2M sample (Figure 9a), it can be observed that the unreacted WBA area is prominent compared with the compact area. The same appearance on the AA-WBA-4M microstructure is shown in Figure 9b. However, there is a predominance of the C-(A)-S-H gel in the cement matrix in contrast with Figure 9a. The microstructure of the AA-WBA-6M (Figure 9c) shows a prominent lighter grayish homogeneous-colored matrix with two distinguished areas corresponding to C-(A)-S-H gel. The compact area is associated with the C-(A)-S-H gel formed by sodium silicate [52], while the rough compact area is ascribed to the C-(A)-S-H gel formed by NaOH [58]. The microstructure of AA-WBA-8M sample (Figure 9d) shows three distinguished areas: The same two light grayish areas of Figure 9c and a darker grayish area corresponding to the unreacted WBA. The formation of C-(A)-S-H gel to a greater or lesser extent, depending on NaOH in the alkali-activator solution used, was demonstrated. The SEM images reveals that the use of NaOH 6M led to a greater formation of the C-(A)-S-H gel in agreement with the selective extraction results mentioned above (Section 3.2).
Appl. Sci. 2020, 10, x FOR PEER REVIEW 10 of 16 Figure 9 shows the SEM images of AA-WBA binders in backscattering electron (BSE) mode. The effect of alkali-activator concentration in the compactness and microstructure of the material is demonstrated. Two differentiated areas in all samples can be observed; a light grayish compact area which was essentially attributed to the C-(A)-S-H gel [52]; and a dark grayish disaggregated area composed of unreacted WBA. In the AA-WBA-2M sample (Figure 9a), it can be observed that the unreacted WBA area is prominent compared with the compact area. The same appearance on the AA-WBA-4M microstructure is shown in Figure 9b. However, there is a predominance of the C-(A)-S-H gel in the cement matrix in contrast with Figure 9a. The microstructure of the AA-WBA-6M ( Figure 9c) shows a prominent lighter grayish homogeneous-colored matrix with two distinguished areas corresponding to C-(A)-S-H gel. The compact area is associated with the C-(A)-S-H gel formed by sodium silicate [52], while the rough compact area is ascribed to the C-(A)-S-H gel formed by NaOH [58]. The microstructure of AA-WBA-8M sample (Figure 9d) shows three distinguished areas: The same two light grayish areas of Figure 9c and a darker grayish area corresponding to the unreacted WBA. The formation of C-(A)-S-H gel to a greater or lesser extent, depending on NaOH in the alkali-activator solution used, was demonstrated. The SEM images reveals that the use of NaOH 6M led to a greater formation of the C-(A)-S-H gel in agreement with the selective extraction results mentioned above (Section 3.2).

Physical and Mechanical Characterization
The density and open porosity of the AA-WBA binders were determined to analyze the mechanical behavior because these properties are extremely linked. The bulk density (1.18 ± 0.1 g·cm −3 ) and open porosity (38 ± 0.5%) results demonstrated the material formation with low density and porosity, probably due to the reaction between metallic aluminum and the alkaline activator to generate hydrogen gas [33,34]. Figure 10 depicts the compressive strength and bulk density results where the same trend can be observed in both cases. A maximum value of compressive strength and bulk density was obtained in the AA-WBA-6M sample, in congruence with selective chemical

Physical and Mechanical Characterization
The density and open porosity of the AA-WBA binders were determined to analyze the mechanical behavior because these properties are extremely linked. The bulk density (1.18 ± 0.1 g·cm −3 ) and open porosity (38 ± 0.5%) results demonstrated the material formation with low density and porosity, probably due to the reaction between metallic aluminum and the alkaline activator to generate hydrogen gas [33,34]. Figure 10

Environmental Characterization
The environmental characterization through leaching tests in deionized water according EN 12457-2 was performed to assess the potential release of heavy metal(loid)s in the AA-WBA binders. In the case of the WBA, the test was carried out for both a non-milled sample and milled sample to evaluate if the process of particle size reduction increased the potential release of heavy metals and metal(loid)s on leachates. In the AA-WBA, the test was carried out on crushed particles below 4 mm, aiming to simulate the worst possible scenario for the AA-WBA leachates, which is at the end of their life cycle when they are demolished. The leaching concentrations of heavy metal(loid)s in WBA and AA-WBA binders are given in Table 5. In WBA, the leaching concentrations of Cu, Cr (only for WBA powder), Mo, and Sb exceeded the limit for acceptance at landfills for inert waste [59]. Most of the heavy metal(loid)s were more leached (As, Pb, Ni, Sb, and Zn) in AA-WBA binders when compared to WBA due to pH increase and the greater pH-dependent metals' mobility. Leaching concentration of Cu and Mo exceeded the acceptance at landfills for inert waste. However, Sb and As were the main causes of concern, since the former exceeded the limit marked for acceptance at landfills for nonhazardous waste, whilst As was very close to this limit. The presence of these two heavy metal(loid)s was because of the substantial amount of primary and secondary glass in the WBA [6]. Arsenic (As2O3) and antimony (Sb2O3) oxides are extensively used in the glass industry as fining agent, to lighten glass and to remove air bubbles [60]. It is important to highlight that the AA-WBA binders were subjected to an aggressive leaching test. The authors expect better results can be obtained with other, less aggressive leaching tests (such as the monolithic test). The environmental characterization reveals that the higher Na2O content in alkali-activator solution, the higher the heavy metal(loid)s' activation.

Environmental Characterization
The environmental characterization through leaching tests in deionized water according EN 12457-2 was performed to assess the potential release of heavy metal(loid)s in the AA-WBA binders. In the case of the WBA, the test was carried out for both a non-milled sample and milled sample to evaluate if the process of particle size reduction increased the potential release of heavy metals and metal(loid)s on leachates. In the AA-WBA, the test was carried out on crushed particles below 4 mm, aiming to simulate the worst possible scenario for the AA-WBA leachates, which is at the end of their life cycle when they are demolished. The leaching concentrations of heavy metal(loid)s in WBA and AA-WBA binders are given in Table 5. In WBA, the leaching concentrations of Cu, Cr (only for WBA powder), Mo, and Sb exceeded the limit for acceptance at landfills for inert waste [59]. Most of the heavy metal(loid)s were more leached (As, Pb, Ni, Sb, and Zn) in AA-WBA binders when compared to WBA due to pH increase and the greater pH-dependent metals' mobility. Leaching concentration of Cu and Mo exceeded the acceptance at landfills for inert waste. However, Sb and As were the main causes of concern, since the former exceeded the limit marked for acceptance at landfills for non-hazardous waste, whilst As was very close to this limit. The presence of these two heavy metal(loid)s was because of the substantial amount of primary and secondary glass in the WBA [6]. Arsenic (As 2 O 3 ) and antimony (Sb 2 O 3 ) oxides are extensively used in the glass industry as fining agent, to lighten glass and to remove air bubbles [60]. It is important to highlight that the AA-WBA binders were subjected to an aggressive leaching test. The authors expect better results can be obtained with other, less aggressive leaching tests (such as the monolithic test). The environmental characterization reveals that the higher Na 2 O content in alkali-activator solution, the higher the heavy metal(loid)s' activation.

AA-WBA Formulation Costs
The costs of AA-WBA formulations compared to OPC costs are shown in Table 6. An increase in the cost of AA-WBA binders in a range between 29 to 45% with respect to OPC can be observed. As expected, the higher the NaOH concentration, the higher the cost of the formulation due to the high cost of NaOH pearls. It is important to note that the electricity for the WBA grinding process was not considered; therefore, the formulation's final cost would be slightly higher. However, the key to produce a competitive AA-WBA binder in cost terms is to reduce the cost of sodium silicate.

Conclusions
The valorization of the WBA is a challenge to be solved by the WtE plants due to the continuous growth of MSW around the world. The work reported herein validates the potential of WBA as precursor in alkali-activated cements. The use of WBA as precursor promotes the zero-waste principle and contributes to the development of new low-carbon cements. Moreover, the novelty and uniqueness of this research was based on reducing the curing temperature of the AA-WBA. The curing at room temperature allowed for the extension and facilitation of the applicability of the AACs. The hydrolytic stability and integrity tests demonstrated that AA-WBA were properly bonded. Selective chemical extraction and physicochemical characterization revealed the formation of C-(A)-S-H gel in the AA-WBA. It also proved the influence of the alkali-activator concentration in their microstructure. The selective chemical extraction and physical characterization results were in accordance with the compressive strength results, where it was demonstrated that the use of AA-WBA is only for non-structural purposes. Environmental characterization showed a concerning activation of As and Sb, which entails a setback for the waste management of the material once its cycle of life is finalized. The cost of the AA-WBA formulations demonstrate that it is not possible to compete with OPC in economic terms due to the high cost of sodium silicate. It would be necessary to search alternatives to produce a sodium silicate from waste glass, making AA-WBA binders more economical and sustainable.
The authors will focus their future line investigation on the formulation of AA-WBA using the least polluted fractions of WBA (coarser fractions). The main aim is using fractions that contain greater availability of SiO 2 , hence favoring the formation of C-(A)-S-H and N-A-S-H gels to enhance the mechanical properties of AA-WBA. The environmental properties of AA-WBA binders would also improve given the low content of heavy metal(loid)s in coarse fractions.  Acknowledgments: The authors would like to thank the Catalan Government for the quality accreditation given to their research groups DIOPMA (2017 SGR 118). The authors also want to thank SIRUSA and VECSA for supplying MSWI Bottom Ash and Karla Montes for the linguistic revision. Alex Maldonado-Alameda is grateful to the Government of Catalonia for the research Grant (FI-DGR 2017).

Conflicts of Interest:
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.