*4.2. Precursors: Rich in Aluminosilicates*

Metakaolin is one of the main precursors rich in aluminosilicates according to some authors [233,234,242]. This is due to its high reactivity and the way in which the material is obtained, which can be originated from ceramic wastes [243], or more commonly, due to calcination of clays rich in kaolinite mineral, known as kaolin [244,245]. The commercial production of metakaolin occurs with the calcination of kaolin at temperatures ranging from 500 to 800 ◦C, depending on the degree of crystallization and purity of the material [246]. The kaolinite present in the material undergoes a dehydroxylation reaction at around 550 ◦C, becoming metakaolinite. Figure 9 presents a scheme for producing metakaolin from traditional kaolin [233,247]. Burning at temperatures below 400 ◦C is not suitable for producing the precursor. Likewise, burning at temperatures above 950 ◦C is not suitable either, due to the formation of mullite that does not have the ability to be alkali activated owing to its high crystallinity and because it is not soluble in an alkaline medium [248,249].

**Figure 9.** Metakaolin manufacturing process through kaolin calcination [247].

It is interesting to note that the use of metakaolin is environmentally advantageous because metakaolin synthesis emits about 5 to 6 times less CO2 than the OPC production process [250]. Furthermore, kaolin can be extracted not only from mineral sources but depending on the composition it can be obtained from industrial paper waste [233]. The different sources will influence the reactivity of the metakaolin obtained, but they are ecologically more viable solutions. Another advantage is that, due to the high reactivity of metakaolin, the structure formed by alkaline activation is extremely resistant, forming a better-defined gel microstructure [233].

Regarding the disadvantages, the price of metakaolin is emphasized. Although lower than the price of OPC, it is superior to other options of precursors rich in aluminosilicates, such as fly ash [249]. Another disadvantage is related to the tendency of efflorescence of the compounds obtained by the alkaline activation of metakaolin, in general, related to the inefficiency of the chemical reaction, which can generate a whitish appearance in the concrete obtained [251,252]. The high retraction tendency of metakaolin is also mentioned, due to the chemical composition of the material, with higher levels of aluminum oxide [253].

Some examples of metakaolin-based alkali-activated cement are now presented. Hasnaoui et al. (2021) [254] evaluated the behavior of geopolymeric concretes produced with metakaolin activated by hydroxide and sodium silicate with recycled fine and coarse aggregates. The authors obtained compressive strength above 50 MPa at 28 days. Gomes et al. (2020) [255] carried out the evaluation of the mechanical properties of geopolymeric concretes produced with metakaolin activated by silicate and sodium hydroxide, containing conventional fine and coarse aggregate and steel fibers. The authors analyzed the results obtained through concepts of fracture mechanics, obtaining values compatible with application in HPC. Dias and Silva (2019) [256] evaluated the effects of the mass ratio of Na2O/SiO2 and K2O/SiO2 on the compressive strength of geopolymers produced with metakaolin without the use of silicates as activators and without the use of coarse aggregates. Albidah et al. (2020) [257] reported on the application of geopolymer concrete produced from metakaolin activated by sodium hydroxide. The authors used two types of coarse aggregates of different particle sizes, in addition to sand as a fine aggregate. The authors also used steel fiber to reinforce the geopolymer matrix, obtaining compressive strength results of approximately 58 MPa at 28 days of cure. Finally, Rocha et al. (2018) [258] evaluated the mechanical properties of geopolymers produced with metakaolin activated by hydroxide and silicate, both sodium and potassium. The authors used river washed sand as fine aggregate. The compositions evaluated by the authors reached approximately

40 MPa with 1 day of cure, showing compressive strength higher than 80 MPa after 28 days of evaluation. These results are compatible with the stipulated values for HPC.

After metakaolin, the second most used precursor rich in aluminosilicates is fly ash. As highlighted in Section 2.2, this material is obtained through the burning of mineral coal for energy production in thermoelectric plants, being one of the most produced alternative binders around the world, with an annual production of more than 900 million tons in 2019 [249]. Its characteristic is fine spherical particles, with a chemical composition based on aluminum, silicon, calcium, iron, magnesium, and carbon residues [259]. Another relevant feature is the particle size in the fly ash, ranging from <1 mm to more than 100 mm, which indicates a high specific surface area and void filling capacity. The main disadvantage of using fly ash in alkali-activated cement is the difficulty in synthesizing hardened products, due to the material curing, which must be done in thermal curing at a temperature of 65 to 90 ◦C to increase the material's reactivity, which it is very low at ambient temperatures [246,260–262].

In the works by Luo et al. (2021) [263], the authors compared the interfacial transition zone (ITZ) behavior of two different types of paste, one based on OPC, and another based on alkali-activated cement produced with fly ash. The results obtained showed that the ITZ of the alkali-activated paste is more strongly adhered to the aggregates than the cementitious paste, contributing to the high mechanical strength of the material. Moghaddam et al. (2021) [264] evaluated the mechanical behavior of geopolymeric concrete produced with fly ash, using rubber as aggregates and incorporating steel fibers. The strength values obtained were equivalent with HPC applications. Finally, the work by Pasupathy et al. (2021) [265], investigated the durability performance of geopolymeric concretes produced with fly ash in saline environments. The authors used two particle sizes of coarse aggregate and sand as fine aggregate. As an activator solution, a combination of sodium hydroxide and silicate was used. In addition to the greater durability of geopolymer concrete, when compared to OPC-based concrete, the authors observed that the compressive strength values were compatible with HPC applications.

Other precursors rich in aluminosilicates with high application potential are calcined illite-smectite clays [266] or calcined feldspars [267,268]. In addition to these, there are volcanic ash, natural pozzolans, and metallurgical slag with low amounts of calcium [269–272]. Recently, some authors have proposed the application of industrial waste, such as chamotte, waste from the ceramic industry [243], and magnesium phosphate from the chemical industry for the production of ammonium [273]. In the studies mentioned above, the alternative precursors shown compatible performance when compared with conventional precursors (e.g., metakaolin), indicating its potential applications in HPC and UHPC. However, further investigations are required to confirm their use for these applications.

## *4.3. Precursors: Rich in Calcium*

Regarding calcium-rich precursors, the main materials used are steel residues or by-products, mainly blast furnace slag. The alkaline activation of these materials results in products similar to those obtained during the hydration of OPC such as C-S-H, but with the potential to be even stronger [236,274]. Regarding calcium-rich precursors, the main materials used are steel residues or by-products, mainly blast furnace slag. The alkaline activation of these materials results in products similar to those obtained during the hydration of OPC such as C-S-H, but with the potential to be even more resistant [275]. In fact, blast furnace slag is already reactive in water, but with very low kinetics. Thus, the presence of alkaline compounds only accelerates the material's hardening reaction [276]. This, in fact, is what motivates the use of slag as a substitute for clinker. Indeed, in the presence of this binder, which during the hydration step forms a solution rich in calcium, the slag can be activated.

The products obtained by the alkaline activation of blast furnace slag in the presence of sodium hydroxide or sodium silicate are composed of hydrated calcium silicates, however, showing substitution of chemical species Si by Al. Figure 10 allows us to understand the

alkaline activation process of the slag, comparing it to the hydration process of OPC. It is observed that the OPC hydration reaction produces large calcium chains linked to Si tetrahedrons and with the presence of chemically linked interstitial water. This structure is called the Dreiketten structure [232,233,275]. In the case of alkaline activation of blast furnace slag, very similar calcium chains are formed. However, some Si tetrahedra, attached to the calcium structure, are replaced by Al tetrahedrons, which allows the formation of cross-links between different Dreiketten chains, giving greater rigidity and strength to the formed compounds. The occurrence of cross-links chemically unbalances the compounds formed by the alkaline activation of the slag, which is why the presence of alkaline ions, preferable metals such as Na+1 and K+1, is necessary to promote the balance of charges. Other metallic ions can also occupy the interstice of the chains, as is the case of Al+3 e Ca+2. In addition to these ions, there is the presence of chemically bound interstitial water in the structure of this material, which is called C-A-S-H or tobermorite [277,278], to differentiate it from hydrated OPC products, usually nomenclated as C-S-H. The microscopic appearance of tobermorite is illustrated in Figure 11.

The main disadvantages of the application of blast furnace slag as a precursor in alkali-activated cement are related to the loss of workability in the material in the fresh state. Marvila et al. (2021) [236] evaluated the rheological aspects of HPC produced with alkali cement activated on the basis of blast furnace slag and sodium hydroxide solution. The authors observed that in smaller amounts of sodium, up to 7.5% Na2O, the obtained material behaved rheologically similar to materials based on OPC. However, at levels above 10% of Na2O, the materials behaved rheologically as fluids with high initial yield stress and dynamic viscosity, impairing the applicability of the material. However, this characteristic is not so relevant, because, as aforementioned, blast furnace slag is reactive to lower levels of alkalinity, enabling the application of the material with lower amounts of sodium.

It is worth highlighting some recent studies that proved the viability of using blast furnace slag as a precursor of alkali-activated cement. For example, Chen et al. (2021) [279] verified the effects of alkaline solution dosage on the properties of materials produced with alkali-activated blast furnace slag, obtaining compressive strength values above 60 MPa after 1 day of cure and above 100 MPa at 28 days, with results equivalent to the production of UHPC. He et al. (2021) [280] also produced materials based on alkali-activated cement from blast furnace slag, evaluating the influence of hydrated lime as an activated agent for the material. The results obtained were above 40 MPa after three days of curing and above 70 MPa after 28 days, with respect to compressive strength. Therefore, the feasibility of applying blast furnace slag as a precursor for alkali-activated materials is proven. Other examples of calcium-rich precursors with potential for application in alkali-activated cement are gypsum desulfurization waste (FGD), cellulosic paper sludge residues and marble residues, which are also amorphous and fine.

Another possibility of alkali-activated cement widely used by some authors is the combined application of calcium-rich and aluminosilicate-rich precursors, as is the case reported by Neupane and Hadigheh (2021) [281]. These authors produced an HPC with alkali-activated cement using silica fume and blast furnace slag, also considering two types of fine aggregate (medium and fine sand) and two types of coarse aggregate, with different particle sizes. The compressive strength results obtained were higher than 30 MPa after 7 days of curing and higher than 50 MPa after 28 days, proving the application of the material as an HPC. Kotop et al. (2021) [282] evaluated the engineering properties of concrete obtained from alkali-activated cement of fly ash and calcium-rich slag, using the application of nanoclays and carbon nanotubes. The compressive strength results obtained reached 60 MPa at 28 days, proving the viability of this type of material for HPC application. The work by Mahmood et al. (2020) [283] also evaluated the mechanical properties of concrete produced with alkali-activated cement-based on fly ash and blast furnace slag, using coarse and fine aggregates and alkaline solution based on hydroxide and sodium silicate. The results obtained for compressive strength at 28 days are above 50 MPa.

**Figure 10.** (**a**) Structure of the C-S-H; (**b**) Structure of C-A-S-H.
