One-part alkali-activated mortars based on clay brick waste, natural pozzolan waste, and marble powder waste

0.27. It was determined that a 7-day curing period between

65 and 80 °C resulted in the highest mechanical properties. The 7-day CS of WCB and natural pozzolan-based materials was determined to be 37 and 26 MPa, respectively. Ulugöl et al. [13] reported the effects of different curing temperatures (50–125 °C), curing periods (24–72 h), and NaOH concentrations (10–15%) on the properties of AAM produced from brick, tile, and glass-based construction and demolition waste. Using WCB and 12% NaOH, a CS of more than 45 MPa was achieved after a 24-hour curing period at 115 °C. The larger particle sizes in the glass explained the lower CS of glass-based materials. Another study [14] produced AAM by replacing NaOH-activated GGBS with 10–50% WCB. As the WCB ratio in the mixture increased, the mechanical properties decreased. However, the CS of the samples with a 10–50% WCB ratio (approximately 25–45 MPa) was sufficient and acceptable for many applications. In recent years, Pommer et al. [15] determined that decreasing the particle size of the WCB resulted in an increased geopolymerization rate and improved mechanical properties. On the other hand, the coarser particles led to an increase in porosity, which influenced the physical properties of the AAM. As a precursor, WCB was substituted for FA by Liang et al. [16], who found pastes containing WCB set faster than the reference paste. The energy used to produce AAM decreased with the addition of WCB. Compared to a reference sample, the total energy savings of the 40 wt% WCB-containing sample were approximately 55%. The effects of alkali activator concentration (8–12 M), alkali solution ratio (1.5–2.5), and curing temperatures (40–60 °C) on the mechanical and durability properties of WCB-based geopolymer bricks were researched by Maaze et al. [17]. The results showed that the optimum molarity, alkali solution ratio, and curing temperature were 10 M, 2.5, and 60 °C, respectively. The samples were considered suitable for high-temperature use because of their thermal stability. Their CS reduced only 3.1–5.8% after exposure to 400 °C and 7.1–8% after exposure to 1000 °C. In another study, Roy et al. [18] replaced GGBS with WCB at 5%, 10%, 15%, and 20% ratios. They found that a 10% replacement ratio with a 6 M alkali concentration gave the maximum CS of 27 MPa. Wang et al. [19] produced an eco-friendly engineered geopolymer composite by using GGBS, FA, and WCB precursors. Three WCB proportions of 20%, 50%, and 80% were used in the precursors. 18 mm polyethylene fibers were included in the samples to increase the ductility. 80% WCB incorporation resulted in the highest values of CS (approximately 50 MPa), tensile strength (higher than

2.1. Raw Materials

WCB, an urban transformation waste, was the aluminosilicate precursor for producing one-part alkali-activated mortar. WCB, with a specific gravity of 2.40, was first broken into small pieces and then dried in an oven at 90 ºC for 48 hours until it reached a constant weight. Then, it was ground in a ball mill to a maximum particle size of 90 µm. The mineralogical composition of WCB consists of kaolinite, muscovite, quartz, gypsum, and carbonates. Calcination is applied at 850–950 °C for the production of bricks. While kaolinite and gypsum disappear at these calcination temperatures, quartz and muscovite continue to exist in the structure in a stable state [2]. The mineralogical composition of the raw materials was determined by X-ray diffraction (XRD) analysis using a Shimadzu XRD-6000 energy dispersive diffractometer (Fig. 1). According to the XRD analysis, the main mineralogical composition of WCB consisted of quartz and albite. The chemical composition of the raw materials was obtained by X-ray fluorescence spectroscopy using a Philips 71 PW-2404 Spectrometer (Table 1). WCB had a high content of SiO2 (61.11%), as well as 14.05% Al2O3. During the calcination stage of the clay, the bound water in the clay minerals evaporates, the crystalline structure of the clay breaks down, and irregular amorphous phases of SiO2 and Al2O3 are formed [22]. These amorphous phases increase the pozzolanic activity of the material [23]. The flexural and compressive strength results (1.12 MPa and

4.56. MPa, respectively) from the pozzolanic activity test

carried out according to TS 25 [24] showed that WCB has pozzolanic activity.

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 391–401, December 2024

Figure 2. Raw materials of the study: (a) WCB; (b) NP; (c) MP; (d) NaOH; (e) SS; (f) SK. WCB: Waste clay bricks; NP: Nevşehir pozzolan; MP: Marble powder; SS: Sodium silicate; SK: Standart Sand.

Table 2. Mix design for WCB-based one-part alkali-activated mortars Sample code

WCB: Waste clay bricks; NP: Nevşehir pozzolan; MP: Marble powder; SS: Sodium silicate; SK: Standart sand.

WCB (Fig. 2a) was substituted by two various natural stone wastes with low- and high-calcium contents to prevent strength losses due to the relatively low reactivity of WCB in environmental conditions and to produce mortar with higher mechanical properties. NP (Fig. 2b) was waste obtained from the waste storage area of a volcanic quarry in the Cappadocia Region (Nevşehir) of Türkiye. The specific gravity of NP was 2.55, the maximum particle size was 90 µm, and the specific surface area determined according to TS EN 196–6 [25] was 7648 cm2/g. As shown in Table 1, the sum of SiO2, Al2O3, and Fe2O3 contents was more than 70%, revealing that NP could satisfy the chemical requirements for its use in alkali activation. In addition, NP was a low-calcium source because its CaO content (0.33%) was less than 10%. Albite and anorthite, found in NP (Fig. 1), were additional indicators of its pozzolanic activity. Following the pozzolanic activity test [24], NP exhibited pozzolan-

ic activity, as shown by its flexural and compressive strength results (1.86 and 6.81 MPa, respectively). MP was supplied from a marble-cutting factory in Izmit (Fig. 2c). The bright white MP was dried in an oven at 90 °C for 48 h until it reached a constant weight. Then, it was ground in a ball mill to a maximum particle size of 90 µm. With 55.7% CaO content, MP was a high-calcium source. The mineralogical composition of MP mainly consisted of calcite and a small amount of quartz (Fig. 1). In the production of AAM, an alkali activator should be used to increase the pH of the reaction medium to ensure the dissolution of the aluminosilicate precursor and the development of subsequent polycondensation reactions [26]. As alkali activators, solid sodium hydroxide (NaOH) with 98% purity (Fig. 2d) and anhydrous sodium metasilicate (Na2SiO3=SS) with 50–52% SiO2 and 50–48% Na2O ratios (Fig. 2e) were supplied by Sigma-Aldrich. Standard sand

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 391–401, December 2024

(SK) with a specific gravity of 2.58 and a maximum particle size of 2 mm (Fig. 2f) in compliance with TS EN 196-1 [27] was used as the aggregate.

2.2. Mixing Ratios

The mixture produced only from WCB (100WCB) was accepted as the reference sample. The study attempted to determine the optimum NP and MP substitution ratios to increase the mechanical properties of WCB-based one-part alkali-activated mortar. In this context, the amount of WCB was gradually replaced by NP or MP in proportions of 25%, 50%, and 75% wt. The total alkali activator: aluminosilicate precursor, NaOH: SS, and binder: SK ratios remained constant at 1:4, 1:2, and 1:3, by wt., respectively (Table 2). No chemical additives were used in the mixtures to prevent possible interactions with the alkali solution. The optimum water: binder ratio that allows producing a workable mortar was determined through trials during the preliminary testing process. The optimum water: binder ratio of WCB-based two-part alkali-activated mortars should be 0.4–0.5 [28, 29], and lower water: binder ratio causes low workability [30]. However, in this study, the optimum water: binder ratio that ensures the workability of WCB-based onepart alkali-activated mortar was determined to be 0.65:1. It is possible that the increased water needs results from solid-state alkali activators, which raise mixing temperatures as soon as they come into contact with water and continue to do so throughout the dissolution process. The resulting heat caused a more significant amount of water to evaporate during mixing, thereby increasing the need for mixing water. On the other hand, in two-part alkali-activated mortars, mixing alkali activators with water for 24 hours before adding them to the mixture does not cause an increase in temperature, so there is no need for more water during mixing.

2.3. Production and Curing Conditions

WCB, NP/MP, solid alkali activators (NaOH and SS), and SK were mixed in dry form for 3 min using the onepart mixing method (Fig. 3a). Then, deionized water was gradually incorporated into the mixture and mixed for another 5 min (Fig. 3b). The fresh mixture was placed into two layers in prismatic molds measuring 40x40x160 mm (Fig. 3c). After half of the mixture was slowly poured into the mold, mechanical vibration was applied for 30 seconds to eliminate air bubbles and ensure complete settling into the mold. Then, the remaining mixture was poured into the mold, and vibration was applied again. The polyethylene-coated samples were kept in the mold at 23 °C for 24 h. The samples removed from the molds after 24 h were cured under ambient conditions of 23 °C and 55% RH for 7, 28, 90, 180, and 365 days (Fig. 3d).

2.4. Experimental Process

A dynamic UV test was applied to the samples by TS EN 14579 [31]. The UV (in km/s) is calculated using the following Eq. (1): UV=l / t(1) Where (l) is the length of the sample and (t) is the time the pulse takes to transverse that length.

Figure 3. Production stages of WCB-based one-part alkali-activated mortars: (a) mixing of dry raw materials; (b) incorporation of deionized water to the dry mixture; (c) molding the fresh mixture; (d) curing the samples under ambient conditions of 23 °C and 55% RH up to 365 days. WCB: Waste clay bricks.

The CS test was performed using an MFL 100 kN machine at a constant speed of 0.5 N/mm2 according to TS EN 196-1 [27]. The CS (in MPa) was calculated using the following Eq. (2): CS=Fc / A(2) Where (Fc) is the maximum load at fracture (in N), and (A) is the cross-section of the area of the sample resisting the load (in mm2). In an experiment with N measurements, one or more may deviate significantly from the others. These "bad data" may be easily discarded if an error in the experimental procedure or an unrelated effect can be found. A statistical criterion should be applied to determine which results can be rejected. Chauvenet's Criterion is an approach that establishes an acceptable statistical scattering about the mean value of a given sample of "N" measurements. According to the Criterion, every data point that falls in a range around the mean representing a probability of 1-1/(2N) ought to be kept. Data points may be rejected when the data point's deviation from the mean is smaller than 1/(2N) [32]. The current study conducted six measurements using UV and CS tests. The arithmetic means, and standard deviations of the first results were calculated. Then, "Chauvenet's Criterion table" was considered to determine whether any data point should be rejected. According to Chauvenet's Criterion for eliminating a data point, the maximum acceptable deviation was 1.73 for six measurements. All measurement results that deviate more than this value (1.73) were eliminated. Afterward, the arithmetic mean of the remaining results was calculated again [33] and declared the "test result."

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 391–401, December 2024

Figure 4. Effect of the WCB: NP ratio on the UV and CS of the samples. WCB: Waste clay bricks; NP: Nevşehir pozzolan; UV: Ultrasound velocity; CS: Compressive strength.

3.1. Effect of Mixing Ratios on Mechanical Properties

The effects of the change in the WCB: NP ratio on the UV and CS of the samples are shown in Figure 4. The reference sample's 7- and 28-day UVs (100WCB) were 2.51 and 2.88 km/s, and the CSs were 3.61 and 5.92 MPa, respectively. The mechanical properties increased as the curing period increased. However, the effect of the longterm curing period on the properties was discussed in the next section. The lower CS of the reference sample at 7 and 28 days revealed that the reference sample was not sufficient to provide high strength when cured at room temperature. Similar to this finding, Migunthanna et al. [21] found the 7and 28-day CSs of WCB-based mortar activated with only solid SS to be 3.3 and 5.4 MPa, respectively. Although the same aluminosilicate precursor (WCB) type was used, the CS was lower than the values obtained in this study. This may be due to the use of NaOH in addition to SS as an alkali activator. Similarly, Zawrah et al. [9], Tuyan et al. [11], and Robayo-Salazar et al. [34] found that WCB-based two-part alkali-activated mortars activated with SS and NaOH had low CS in the early period under ambient conditions. Because of the low degree of reaction of WCB under ambient conditions [11], the limitations in mechanical properties may be addressed by increasing the curing temperature [12]. However, in this study, only ambient curing was applied, considering that curing at high temperatures increases energy consumption in the production process. By adding NP to the reference sample, enhanced mechanical properties were obtained in the early period. According to Figure 4, due to substituting WCB with 25% NP, the 7- and 28-day UVs increased to 1.17 and 1.13 times their original value, and the CSs increased to 1.21 and 1.33 times their original value, respectively. The highest mechanical properties were achieved at 50% NP substitution. The 7- and 28-day UVs of the 50WCB:50NP sample increased by 1.29 and 1.23 times compared to the reference sample, reaching 3.25 and 3.56 km/s, respectively. The same samples' 7- and 28-day CS increased by 2.27 and 2.10 times, reaching 8.21 and 12.43 MPa, respectively. For alkali activation to occur to a high degree, the SiO2, Al2O3, Fe2O3, and CaO content in the reaction medium

Figure 5. Effect of the WCB: MP ratio on the UV and CS of the samples. must be sufficient. In particular, SiO2 and Al2O3 are primary components that affect the formation of the three-dimensional amorphous network. Mixtures containing high amounts of amorphous SiO2 and Al2O3 allow relatively higher mechanical properties because of the more complete alkali activation reactions consisting of three stages (dissolution, rearrangement, and polycondensation) [35]. The NP's total SiO2 and Al2O3 content (95.77%) was higher than that of the WCB (75.16%). This means that when NP was added to the mixture, it increased the amounts of reactive components and made it easier for the reactions to progress. Thus, higher mechanical properties were achieved. On the other hand, the effect of increasing the NP ratio by more than 50% on strength development was negligible. Similarly, in the literature, when the FA content was more than 40% in WCB+FA-based one-part alkali-activated mortars [21] and more than 50% in two-part alkali-activated mortars [10], the strengths did not increase. Considering that NP has chemical composition properties similar to FA, these results are consistent with the findings of this study. The effects of the change in the WCB: MP ratio on the UV and CS of the samples are shown in Figure 5. Adding MP to the reference sample obtained the highest mechanical properties early. As a result of substituting WCB with 25% MP, the 7- and 28-day UVs increased by 1.24 and

1.16. times, and the CSs increased by 2.49 and 2.33 times,

respectively. Gradually increasing the MP ratio to 50% and 75% increased mechanical properties. The highest mechanical properties were detected in the 25WCB:75MP sample. This sample's 7- and 28-day UVs increased by 1.37 and 1.28 times, compared with the reference sample's, reaching 3.44 and 3.70 km/s, respectively. The same sample's 7- and 28day CSs increased by 4.12 and 3.64 times, reaching 14.88 and 21.58 MPa, respectively. In addition, the mechanical properties of WCB+MP-based samples were higher than those of WCB+NP-based samples at the same curing period. This result might be due to the high calcium-containing components in the MP. As shown in Table 1, WCB and NP contain only 8.76% and 0.33% CaO, respectively. MP, which includes 55.7% CaO, is a calcium-rich raw material. A three-dimensional amorphous sodium aluminosilicate hydrate (N-A-S-H) gel forms during alkali activation [1], and this gel gives the mortar its main properties. In this study, the SiO2 and Al2O3 components required for forming

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 391–401, December 2024

Figure 6. CS of WCB-based one-part alkali-activated mortars as a function of UV. WCB: Waste clay bricks; NP: Nevşehir pozzolan; UV: Ultrasound velocity; CS: Compressive strength.

the N-A-S-H gel in WCB+MP-based samples were primarily supplied by WCB. During the polycondensation process, when calcium-rich components from MP were added to the reaction medium, Na+ ions in the N-A-S-H gel and dissolved Ca+2 ions were replaced, and a solid (N, C)-A-S-H gel was formed. In addition, a calcium silicate hydrate (C-S-H) gel was formed, similar to that in Portland cement. The coexistence of these gels within the aluminosilicate network resulted in higher strength development [9, 36, 37]. By measuring the UV of a material, the early reaction stages, the microstructure development, improvement of rheological properties, setting, hardening, strength development, and even durability properties can be successfully determined [38–40]. It is possible to state that materials with higher UV have a more compact microstructure, as no gaps, cracks, or damage would prevent sound from propagating through the material [41]. A more compact microstructure allows the production of materials with higher mechanical properties. Therefore, regardless of the curing period, the CS-UV relationship of the samples is shown in Figure 6. The lowest CS was determined in the reference sample, whereas the highest was in the 25WCB:75MP sample. All samples showed a similar trend in terms of the CS-UV relationship. The CS changed in direct proportion to the UV. Because the regression lines' correlation coefficients (R2) were in the range of 0.97–0.99, the relationship between CS and UV was strong, indicating that CS of the one-part alkali-activated mortars could be estimated through UV. The Eq. (3) below represented the linear relationship between CS and UV:

Figure 7. XRD results of one-part alkali-activated mortars. XRD: X-Ray diffraction.

CS=a x UV + b(3) The slope of the regression lines is referred to as the value "a." The CS-UV relationship of the samples at each mixture ratio was expressed according to the following Eqs. (4–10): (100WCB) CS=8.17 x UV – 17.06(4) (75WCB:25NP) CS=11.74 x UV – 30.05(5) (50WCB:50NP) CS=11.98 x UV – 30.56(6) (25WCB:75NP) CS=11.50 x UV – 28.85(7) (75WCB:25MP) CS=16.13 x UV – 40.87(8) (50WCB:50MP) CS=19.45 x UV – 52.66(9) (25WCB:75MP) CS=21.62 x UV – 59.15(10)

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 391–401, December 2024

Figure 8. Effects of the curing period on the CS of the samples. WCB: Waste clay bricks; NP: Nevşehir pozzolan; UV: Ultrasound velocity; CS: Compressive strength.

3.2. Effect of Mixing Ratios on Mineralogical Properties

Figure 7 shows the XRD results of the reference sample and the samples containing NP/MP with the highest mechanical properties (50WCB:50NP and 25WCB:75MP) after ambient curing for 28 days. Quartz mineral (Q) was the main crystalline phase in the reference sample. After alkali activation reactions, no new phase formation was observed. This might be because WCB had a relatively low reactivity, i.e., a low amorphous character, and the reactions could not develop sufficiently under ambient conditions. The low amorphous character of WCB was due to the bricks being allowed to cool slowly after calcination [11]. This finding obtained from the XRD analysis was parallel to the low mechanical properties exhibited by the reference sample after a 28-day curing period under ambient conditions. The XRD pattern of geopolymer mortars containing various amounts of NP and MP had a similar XRD pattern to the reference sample; however, new amorphous and semi-crystalline phases were present. After the alkali activation of 50WCB:50NP, a mixture high in SiO2 and Al2O3 formed an alkali three-dimensional amorphous aluminosilicate (N-A-S-H) gel called geopolymer gel. In addition, a sodium calcium aluminosilicate hydrate ((N, C)-A-S-H) phase at 2θ=39.8° was present in the mortar. These newly formed gel types provide a better-developed internal structure and, therefore, higher mechanical properties than the reference sample. On the other hand, hydration products in the 25WCB:75MP sample are composed of semi-crystalline phases of calcium silicate hydrate (C-S-H), which have in-

tense peaks centered around 2θ=32.2°, 50.1°, and 55.1°, and (N, C)-A-S-H in amorphous geopolymer structure. This semi-crystalline phase improved the mechanical properties of the mortar by serving as nucleation sites for the development and accumulation of the amorphous geopolymer products [37]. In other words, the coexistence of both gel phases within the same network structure promoted alkali activation and ensured the internal structure development with higher CS of the mortar.

3.3. Effect of Curing Period

The effects of the curing period on the CS of the samples are given in Figure 8. A tendency to increase in the CS of all samples was observed as the curing period increased. The highest slope of the curve, between 7 and 28 days, indicated that the mortar structure developed rapidly during the early period. During the early curing period, reactions occurred rapidly because the raw materials, water, and sufficient ambient temperature required for alkali activation were present in the mortar. The three-dimensional amorphous network structure produced as a result of the reactions enabled the filling of the voids in the mortar, the development of a more compact internal structure, and thus a rapid increase in mechanical properties [42]. The slope of the CS-curing period curve of the reference sample was relatively low, leading to slower strength development. The fastest strength development in the early period was detected in samples containing WCB and MP. The 25WCB:75MP samples showed the highest UV (3.44 km/s) and

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 391–401, December 2024

CS (14.88 MPa) after 7 days. These properties improved quickly over time; therefore, the samples retained their highest mechanical properties during the subsequent curing periods. In other words, the development of alkali activation in the first 7 days later contributed to the material's performance. On the other hand, regardless of the mixing ratio, the increase in CS of all samples decreased gradually starting from the 90th day and tended to remain constant, especially after the 180th day. In particular, the rate of increase between 90 and 365 days was negligible compared with the growth rate between 7 and 90 days. This finding could be attributed to the fact that the reaction products (both the (N, C)-A-S-H and C-S-H gels) that occurred during the early curing period have already filled the voids of the one-part alkali-activated mortar structure, causing the mortar structure to become more compact. The tendency for mechanical properties to remain constant in the later curing period is also compatible with the results of other experimental research in the literature. The strength increment of FA-based one-part AAM between 180 and 365 days was also stated as "negligible" by Haruna et al. [42]. The production of geopolymers from WCB, NP, and MP wastes would lower the cost of storing these wastes, the land demand required for storage, and the requirement for raw materials. Lowering the demand for raw materials would help to decrease greenhouse gas emissions and energy requirements. In addition, using these wastes to synthesize one-part alkali-activated mortars makes them a more environmentally friendly alternative to the traditional ones. The results of this experimental study demonstrate that WCB, NP, and MP wastes can be efficiently used to produce one-part alkali-activated mortars at the laboratory scale. However, the potential challenges for large-scale implementation should be identified to make the one-part alkali-activated mortars available in real-world industrial applications. For instance, the continuity of the waste raw material sources must be ensured. With the use of raw materials on an industrial scale, there might be a shortage of these sources in the foreseeable future. Since the chemical composition of raw materials (SiO2, Al2O3, Fe2O3, and CaO contents) significantly influences the development of geopolymerization reactions and the properties of the resulting product in both fresh and hardened states, raw materials with constant chemical composition should be used. It should be considered that the greater transportation distances of the source materials would reduce the environmental benefits of the onepart alkali-activated mortars. In addition, in laboratory-scale applications, mixture preparation, molding, and curing can be easily implemented. However, for this production process to be carried out on an industrial scale, equipment and labor requirements should be defined correctly, and practical application methods should be developed. The production cost of the material on an industrial scale could be another challenge due to the higher cost of alkali activators. Therefore, a detailed cost analysis of WCB-based one-part alkali-activated mortars is required.