Design of Geopolymers Based on Greek CDWs Using the Taguchi Method

Abstract

This study explored an alternative approach to managing construction and demolition waste (CDW) in Greece by repurposing waste bricks and tiles as secondary raw materials for geopolymer synthesis. Alkali dissolution tests indicated that waste brick is more susceptible to alkaline attack than tile waste. The Taguchi method was employed as a design of experiments (DoE) approach to optimize synthesis and address CDW mineralogical variability, targeting maximum compressive strength. The primary influencing factors were alkali content (64%) and soluble silicon (33%). Geopolymers with a compressive strength of 42.8 MPa were synthesized at 90 °C for 3 days under optimal conditions: a soluble silicon-to-alkali molar ratio of 0.5, an alkali-to-aluminum molar ratio of 1, and a 50:50 sodium–potassium ion mixture.

1. Introduction

Construction and demolition wastes (CDWs) are a considerable environmental challenge globally. They contribute significantly to total waste generation. By numbers, CDW volume is estimated to exceed 2 billion tons annually on a global scale, as per the United Nations Environment Programme (UNEP) [1]. Moreover, CDW accounts for around 35 percent of the total waste production in the European Union, which corresponds to over 374 million tons annually [2]. In Greece alone, construction, renovation, and demolition activities generated 10 million tons of CDWs in 2022, excluding excavated soils. Of this waste, only 53% was recovered, while the rest was disposed of in landfills, according to the Hellenic Statistical Authority [3].

The current inadequate management of CDWs contributes to serious environmental challenges such as resource depletion, increased landfill burden, and greenhouse gas emissions [4,5]. Moreover, the expanding construction sector escalates the need for raw materials, exerting additional strain on natural resources. The enormous volume of this waste, along with its mishandling, highlights the critical need for novel approaches to divert CDW from landfills and turn it into valuable resources.

A substantial portion of CDW, approximately 35% [6], consists of ceramics, including concrete, masonry, and mortars. These materials include aluminosilicate and calcium silicate, which can be used to create new, environmentally friendly building materials called geopolymers [7,8,9,10]. Geopolymers are mainly produced from aluminosilicate industrial wastes and by-products where Si and Al are predominantly in amorphous phases after mixing with a condensed alkali solution [11,12]. In comparison to conventional building materials, geopolymers exhibit a reduced amount of embodied energy and carbon dioxide emissions. They rapidly acquire mechanical strength and can withstand severe conditions without deterioration [13]. Moreover, they are advantageous due to their versatility, which is contingent upon the selection of the raw materials and the processing methods employed. The adaptability of geopolymer synthesis is crucial when there is a need for products with distinctive characteristics. This versatility is reflected in a wide range of applications reported in the literature, such as geopolymers with enhanced thermal performance at elevated temperatures [14], their use as alternative concrete materials [15,16], and the development of geopolymer composites reinforced with various types of fibers to improve mechanical properties [17]. In addition, geopolymers have been explored as absorbents for toxic elements [18], in the production of lightweight and thermally insulating products [19], and in “just-add-water” formulations for ease of application [20]. Their potential for carbon capture and environmental remediation has also been highlighted in recent studies [21].

The use of CDWs as resources in the geopolymer technology is an emerging scientific field. Recently, several studies explored the use of CDW ceramic parts (concrete, brick, and tile wastes) as secondary raw materials in the geopolymer synthesis of pastes and mortars [22,23,24]. The primary research on CDW geopolymer synthesis involved the mixing of CDWs with supplementary cementitious materials (metakaolin [25], ground granulated blast furnace slag [26], fly ash [27,28], silica fume [29], OPC [30], etc.) to achieve products with improved properties due to their relatively low geopolymerization potential. Furthermore, synthesis optimization involved the investigation of a number of key parameters, such as the selection of CDW precursors, the processing prior to synthesis, the curing conditions, and the alkali activator’s type and concentration [31,32,33]. These parameters have been evaluated for their impact on microstructure, fresh-state properties, and mechanical performance.

However, the mineralogical variability of CDWs across the globe presents a significant challenge in achieving uniform and high-performance geopolymer compositions. This inconsistency in raw material properties complicates the synthesis process, highlighting the need for a rapid and efficient optimization tool to enhance geopolymer production. Such a tool is essential for adapting processes to local material characteristics, ensuring scalability, and maintaining economic viability. Addressing this challenge can bridge the gap between waste material variability and the effective utilization of geopolymers, ultimately improving resource efficiency and advancing circular economy principles in the construction industry.

To tackle these challenges, a multifactorial design approach based on the Taguchi technique has proven particularly effective [9,34,35]. The Taguchi method is a systematic and efficient approach for optimizing complex processes and identifying key parameters and their interactions with minimal experimental effort [36,37]. Its ability to manage variability makes it highly suitable for geopolymer synthesis, where the diverse mineralogical composition of CDWs requires precise adjustments to factors such as activator concentration, curing conditions, and mix design. By leveraging this approach, researchers can rapidly determine optimal geopolymer production parameters, ensuring consistent performance despite fluctuations in raw material quality [35,38]. Beyond enhancing the reliability of geopolymers, this strategy accelerates the development of sustainable construction solutions.

This paper outlines a systematic approach for the design and production of mechanically robust geopolymers by valorizing CDWs originating from Greece as the sole aluminosilicate precursors. Specifically, two CDW sources (bricks, tiles) collected from demolition activities and appropriately separated-were utilized in geopolymer synthesis. The Taguchi method, an efficient design of experiments (DoE) approach for assessing the interdependent effects of multiple parameters on a given process, was employed to optimize the synthesis. The investigation involved the examination of the following synthesis parameters: soluble silicon content of activation solution, alkali content of the activation solution, and type of alkali. To check the results of the Taguchi experiments, alkali leaching tests were performed on CDWs. Additionally, the prepared geopolymers were characterized by XRD, FTIR, and SEM measurements.

2. Materials and Methods

2.1. Materials

We tested CDW bricks and tiles originating from demolition activities in Greece as alternative aluminosilicate sources for geopolymer synthesis. The two CDW sources were appropriately processed through crushing and grinding to achieve a specific fineness (d₅₀ ~ 20 μm) before their use. Figure 1 displays the particle size distribution of the brick wastes (BW) and tile wastes (TW) determined by laser granulometry measurements. The mineralogical and chemical compositions of BW and TW are presented in Figure 2 and

Table 1. Precursors’ chemical composition (% wt.). The amorphous content and specific surface area (SSA) are also given.

Source	BW	TW
SiO2	52.3	56.7
Al2O3	15.1	16.3
Fe2O3	7.9	6.8
CaO	6.1	7.1
MgO	8.3	3.7
K2O	1.6	3.2
Na2O	0.8	0.5
SO3	0.3	0.2
TiO2	0.5	0.6
P2O5	0.2	0.1
Cl	0.02	0.03
L.O.I.	6.9	4.8
Amorphous content (%)	43.3	32.5
SSA (m2/g)	1.5667	1.6093

, respectively.

2.2. Leachability Measurements

To examine the geopolymerization potential of the BW and TW precursors, aluminum and silicon leachability tests were performed in two different alkaline media of sodium and potassium hydroxides, maintaining a 10 M concentration. The experimental procedure for each sample included the following steps [39]. (i) A precise amount of the precursor (1.0000 ± 0.0001 g) was mixed with 40 mL of an alkaline hydroxide solution in a plastic container. (ii) The mixture was stirred continuously for 24 h. (iii) After that, the solution was filtered with a Whatman grade 589/3 filter (110 mm wide) under vacuum. (iv) The filtered liquid was then diluted with distilled water to make a total of 250 mL and to obtain a solution of a measurable concentration. (v) A portion of this solution was treated with concentrated HCl (37% wt.) to lower the pH below 1, preventing any solid from forming. (vi) Finally, the solution was analyzed with atomic absorption spectroscopy (AAS) to measure the amounts of silicon and aluminum. The solid precipitate was also examined through XRD analysis to evaluate the impact of leaching in the mineralogical content of the samples.

2.3. Geopolymer Preparation

The alkaline solutions were prepared by dissolving dry pellets of sodium (NH, CAS: 1310-73-2) and/or potassium (KH, CAS: 1013-58-3) hydroxides in tap water. A silicon oxide solution in the form of colloidal dispersion (Si sol, 50% in H₂O, CAS: 7631-86-9) was then added to promote the nucleation process. The alkaline solution was stirred for 1 h to ensure that a chemical equilibrium was achieved. A regular mortar mixer (Controls 65-L0005) was used to mechanically mix the aluminosilicate precursors (BW or TW) and the alkaline solution to obtain a homogeneous and workable slurry. The slurries were then casted in cubic molds (50 × 50 × 50 mm), kept for 2 h in room temperature, and stored inside plastic bags to avoid loss of humidity. Finally, the slurries were cured at various temperatures and for various durations to specify the optimum curing conditions. We performed compression tests seven days after specimen preparation. The synthesis procedure followed was similar to that used for the preparation of OPC pastes and mortars according to [40,41].

We conducted initial experiments to explore the geopolymerization potential of the BW and TW precursors (

Table 2. Geopolymer mixtures prepared in the initial experiments (%wt.).

ID	Synthesis Parameters			Precursor	Activator		H2O	Curing Conditions
	Si/Na	Na/Al	L/S		Si Sol	NH		T (°C)	t (d)
BWi1	0.0	1.0	0.25	71.5	0.0	8.5	20	50	2
BWi2	0.5	1.0	0.25	66.2	11.8	7.8	14.1	50	2
BWi3	1.0	1.0	0.25	61.7	22.0	7.3	9.0	50	2
BWi4	0.5	0.5	0.25	72.5	6.5	4.3	16.8	50	2
BWi5	0.5	1.5	0.25	61.0	16.3	10.8	11.8	50	2
BWi6	0.5	2.0	0.25	56.5	20.2	13.4	9.9	50	2
BWi7	0.5	1.0	0.35	61.3	10.9	7.3	20.5	50	2
BWi8	0.5	1.0	0.30	63.5	11.3	7.5	17.6	50	2
BWi9	0.5	1.0	0.20	69.0	12.3	8.2	10.5	50	2
BWi10	0.5	1.0	0.25	66.2	11.8	7.8	14.1	70	1
BWi11	0.5	1.0	0.25	66.2	11.8	7.8	14.1	70	3
BWi12	0.5	1.0	0.25	66.2	11.8	7.8	14.1	25	2
BWi13	0.5	1.0	0.25	66.2	11.8	7.8	14.1	70	2
BWi14	0.5	1.0	0.25	66.2	11.8	7.8	14.1	90	2
TWi1	0.0	1.0	0.25	70.9	0.0	9.1	20.0	50	2
TWi2	0.5	1.0	0.25	65.4	12.5	8.4	13.7	50	2
TWi3	1.0	1.0	0.25	60.6	23.3	7.8	8.4	50	2
TWi4	2.0	1.0	0.25	52.9	40.6	6.8	0.0	50	2
TWi5	1.0	0.5	0.25	69.0	13.2	4.4	13.4	50	2
TWi6	1.0	1.5	0.25	54.1	31.1	10.4	4.4	50	2
TWi7	1.0	2.0	0.25	48.8	37.5	12.5	1.3	50	2
TWi8	1.0	1.0	0.20	62.9	24.1	8.0	4.9	50	2
TWi9	1.0	1.0	0.30	58.1	22.3	7.4	12.1	50	2

). Additionally, to determine a rough value range of the key factors of the geopolymer synthesis, such as: (i) the soluble silicon content (Si/Na) defined as the molar ratio of added silicon to sodium content in alkaline solution (0–4), (ii) the alkalinity of the starting solutions (Na/Al), shown by the molar ratio of alkali to aluminum (0.5–2.0), (iii) the workability of the slurries (L/S), specified as the mass ratio of total water to solids, including precursor and solid components of the reagents used (0.25–0.35), and (iv) the curing conditions in terms of temperature (25–90 °C) and duration (1–3 days).

The results from the initial experiments were used to adjust the range of the synthesis parameters. This helped to improve the production of the CDW-based geopolymer using the Taguchi method. The Taguchi method belongs to the fractional multivariate design of experiments that allows the investigation of the combined effect of selected parameters by conducting the minimum number of experiments [9]. Three synthesis parameters were chosen for investigation, and the goal of the experimental design was to maximize the unconfined compressive strength (UCS) values of the produced geopolymers. The examined synthesis parameters were the following: (i) the amount of added silicon compared to the amount of alkali of the activation solution (Si/R, where R is sodium + potassium); (ii) the alkalinity of the activation solutions in terms of alkali hydroxides’ amount compared to the amount of aluminum of the precursor (R/Al); and (iii) the type of alkali based on the ratio of sodium to the total amount of sodium and potassium (Na/R). The experimental plan included changing the abovementioned parameters at three different levels using an L₉ (3³) orthogonal array (

Table 3. Geopolymer mixtures prepared according to the Taguchi method.

ID	Investigated Parameters			Precursor	Activator			H2O
	Si/Na	Na/Al	Na/R		Si Sol	NH	KH
BWt1	0.0	0.4	0.0	75.0	-	-	5.0	20.0
BWt2	0.0	0.7	0.5	72.7	-	3.0	4.2	20.0
BWt3	0.0	1.0	1.0	71.5	-	8.5	-	20.0
BWt4	0.5	0.4	0.5	73.2	5.2	1.7	2.4	17.4
BWt5	0.5	0.7	1.0	69.8	8.7	5.8	-	15.6
BWt6	0.5	1.0	0.0	63.7	11.4	-	10.6	14.3
BWt7	1.0	0.4	1.0	71.5	10.2	3.4	-	14.9
BWt8	1.0	0.7	0.0	64.5	16.1	-	7.5	12.0
BWt9	1.0	1.0	0.5	60.6	21.6	3.6	5.0	9.2

). A complete factorial design necessitates the execution of 27 experiments (3 factors, 3 levels per factor; experiments = #levels^#factors = 3³ = 27). The Taguchi method, using an L₉ orthogonal array, reduced the number of experimental trials to nine. The exported data were then analyzed through ANOVA (analysis of variance) statistical analysis to find the percentage contribution of each examined parameter in the development of UCS in each geopolymer sample.

2.4. Characterization Methods

The mineralogical content of the CDW precursors and final products was studied through X-ray diffraction (XRD) using a Bruker D8 ADVANCE X-ray diffractometer. (Bruker, Billerica, MA, USA). The measurements were conducted to a 20–70° 2θ range employing a step size of 0.01° and a duration of 1 second for each step. The XRD patterns were then analyzed utilizing Diffrac.Eva v3.1 software.

Fourier transform infrared (FTIR) measurements were conducted using a Jasco 4200 type A spectrophotometer (JASCO Europe s.r.l., Cremella, Italy). The KBr pellet method was used to obtain the FTIR spectra, which had a range of 400 to 4000 cm⁻³ and a resolution of 4 cm⁻¹. The pellets were fabricated by compressing a blend of the sample and desiccated KBr (sample–KBr ratio roughly 1:200) at a pressure of 10 t/cm².

The morphology and phase stoichiometry of the synthesized products were investigated using a JEOL JSM-5600 scanning electron microscope (JEOL Ltd., Tokio, Japan) coupled with an Oxford Link ISIS 300 energy-dispersive X-ray (EDX) spectrometer (Oxford Instruments, Oxford, UK). Prior to analysis, the samples were prepared in fragment form and coated with a thin layer of gold to enhance conductivity.

2.5. Mechanical Strength Tests

A Toni Technik uniaxial testing press was used to test the UCS of the produced geopolymers, ensuring a load rate of 1.5 kN/s, in compliance with the [42]. The UCS was measured as the mean strength value of three tested specimens 7 days after the synthesis according to [6].

3. Results and Discussion

3.1. Leachability Tests

The leachability tests were performed in a way to simulate the first step of the geopolymer synthesis, which includes the alkaline dissolution of the aluminosilicate source. This method can be useful to facilitate the evaluation of the precursor susceptibility to alkaline attack and therefore the accumulation of reactive Si and Al species that will form the structural units of the geopolymer network.

Table 4. AAS results.

Precursor	Silicon (mg/L)		Aluminum (mg/L)
	NH (10M)	KH (10M)	NH (10M)	KH (10M)
BW	39.3	28.9	15.1	8.6
TW	25.2	19.3	10.7	5.9

presents the results of AAS measurements in terms of Si and Al concentrations (mg/L) that were leached from the examined aluminosilicate wastes (BW and TW) when diluted in concentrated alkaline solutions (10M NH and KH). The results of AAS are also depicted in Figure 3, after normalization of their values, representing the % Si and Al amounts dissolved from the total Si and Al quantities of the aluminosilicate precursors.

As is evident, both BW and TW precursors are somewhat susceptible to alkaline dissolution with the dissolved Si and Al species: 3.0–7.5% and 3.5–8.5%, respectively. BW exhibited higher leachability in relation to the TW source. Furthermore, the NaOH solution gave better results since Na ions have a smaller ionic radius than K, thus more efficiently penetrating the aluminosilicate network [43]. The leachability of BW and TW was compared with well-established aluminosilicate precursors such as metakaolin (MK) and fly ash (FA), indicating that the leached values of Greek BW and TW were significantly lower [39]. Nevertheless, it is important to note that the leaching behavior of Si and Al alone is not sufficient to determine the overall quality of the prepared geopolymers. Therefore, further experiments were necessary to evaluate the synthesis process and the properties of geopolymers derived from these precursors.

The solid residues of the filtration, after drying, were examined by XRD in order to identify possible structural changes in the mineralogical composition of the materials. Figure 4 shows the XRD patterns of the BW and TW precursors and their solid residues after the leaching process in 10 M NH for 24 h. As is evident from the XRD patterns, the crystallographic phases of the precursors are still detectable in the solid residues, albeit in lower intensities. This fact supports the low leachability of the BW and TW precursors. However, BW solid residue exhibits greater amorphousness compared to the TW residue, suggesting that a larger amount of the precursor is leached during alkaline dissolution. This observation aligns with the AAS results.

3.2. Initial Experiments on Geopolymer Synthesis

Initial experiments on the geopolymer synthesis of BW and TW were carried out by varying one synthesis factor at a time. Specifically,

Table 5. UCS values for BW and TW geopolymers according to initial experiments.

ID	UCS (MPa)
	1	2	3	Average	SD
BWi1	10.9	11.5	11.1	11.2	0.3
BWi2	17.5	16.4	16.8	16.9	0.6
BWi3	13.4	12.3	11.5	12.4	1.0
BWi4	6.3	7.4	6.9	6.9	0.6
BWi5	1.9	2.1	2.2	2.1	0.2
BWi6	0.7	0.8	0.8	0.8	0.1
BWi7	11.6	11.9	12.2	11.9	0.3
BWi8	15.4	15.9	15.5	15.6	0.3
BWi9	12.8	13.9	14.6	13.8	0.9
TWi1	2.3	2.7	2.4	2.5	0.2
TWi2	4.8	4.1	4.0	4.3	0.4
TWi3	6.2	6.9	7.2	6.8	0.5
TWi4	5.8	5.1	4.9	5.3	0.5
TWi5	3.6	3.6	4.1	3.8	0.4
TWi6	5.5	5.5	5.6	5.5	0.1
TWi7	0.3	0.4	0.4	0.4	0.1
TWi8	4.8	5.6	4.7	5.0	0.5
TWi9	5.2	6.1	6.4	5.9	0.6

displays these experiments along with the UCS values of the resulting samples.

The UCS values from the experiments demonstrated good dispersion (1–17 MPa), indicating that the choice of the investigated synthesis parameters and their variability influenced the geopolymer properties. Of the two waste precursors, BW demonstrated higher geopolymerization capabilities, corroborating the findings of Si and Al leachability tests. Furthermore, BW is the main aluminosilicate material in Greek CDWs, offering the opportunity of repurposing higher amounts of wastes through geopolymer technology [44]. Tile waste can be employed in two possible ways in geopolymer technology: as the sole precursor for the production of low-strength building materials or in combination with the remaining ceramic fraction of Greek CDWs to develop high-quality building materials.

Figure 5 depicts the UCS values of BW and TW geopolymers in relation to the variation in synthesis parameters Si/Na, Na/Al, and L/S. Concerning the Si/Na molar ratio, the existence of extra silicon in the geopolymer synthesis is beneficial for preparing samples of higher strength. This silicon reacts with the readily available aluminum dissolved from the precursor at the early stages of the dissolution, thus promoting the formation of oligomers that will then take place in the polycondensation reaction of the geopolymer synthesis [45]. However, the beneficial impact of extra silicon on the mechanical properties has a maximum threshold (Si/Na = 0.5 and 1.0 for BW and TW, respectively). Higher Si/Na values result in a decrease in the UCS values. Significantly high extra silicon content saturates the alkaline solution, negatively affecting the precursor dissolution rate and subsequently the geopolymerization reactions.

The Na/Al molar ratio is associated with the alkalinity of the activation solutions. In general, the existence of alkali metals promotes the dissolution of the geopolymer precursors and at the same time compensate for charge imbalances in the geopolymer network [45]. On the other hand, high alkali content can cause deterioration of the geopolymer structures, since the excess of these metals reacts with the atmosphere, producing carbonates. As can be seen in Figure 5, a value of Na/Al close to 1 achieves the highest UCS values for both BW and TW precursors (UCS = 16.4 and 6.8 MPa).

The impact of water content was also assessed by investigating the L/S mass ratio. It was obvious that low (L/S = 0.20) or high (L/S = 0.35) water content had a negative effect on the UCS values of the CDW geopolymers. This observation indicates that low water content affects the workability of the pastes, resulting in poor mixing of the reactants, while high content lowers the alkalinity of the activation solution and thus the effective dissolution of the CDW precursor. An L/S of 0.25 achieves the best mechanical performance (UCS = 16.9 and 6.8 MPa for BW and TW, respectively).

Figure 6 shows how curing temperature (25, 50, 70, and 90 °C) and time (1, 2, and 3 days) affect the UCS values of the BW geopolymers. As is evident, an increase in both temperature and time had a positive effect on the development of UCS values. Among the two curing parameters, temperature has a greater effect, since an increase from 25 to 90 °C led to an approximately 90% enhancement of strength. On the other hand, time also increased the UCS values, but to a much lesser degree (31%). Considering the results of curing condition experiments, the curing at 90 °C for 3 days was selected to be applied for the rest of the experiments.

3.3. Optimization of Geopolymer Synthesis

We optimized the geopolymer synthesis of BW using the Taguchi multifactorial design based on the results of the initial experiments.

Table 6. UCS values of BW geopolymers based on the Taguchi method.

ID	UCS (MPa)
	1	2	3	Average	SD
BWt1	4.1	4.4	4.2	4.2	0.2
BWt2	19.1	22.3	19.8	20.4	1.7
BWt3	16.5	14.1	15.1	15.2	1.2
BWt4	13.7	14.9	13.5	14.0	0.8
BWt5	31.7	28.7	29.1	29.8	1.6
BWt6	38.4	37.9	39.1	38.5	0.6
BWt7	13.1	12.1	12.6	12.6	0.5
BWt8	19.9	21.2	22.6	21.2	1.4
BWt9	39.5	36.4	33.2	36.4	3.2

sums up the designed experiments along with their synthesis parameters and the corresponding UCS.

Figure 7 illustrates the influence of each synthesis parameter on the UCS of BW geopolymers, as derived from the data processing of Table 6 through ANOVA analysis. To be more descriptive, the UCS value corresponding to a Si/Na molar ratio of 1.0 (Figure 7) represents the average UCS value of all samples listed in Table 6 that were prepared using a Si/Na ratio of 1.0.

The alkalinity of the activation solutions (R/Al) was found to be the most significant synthesis parameter that influenced the mechanical behavior of the BW geopolymers (influence level of 64%). Indeed, an escalation in this molar ratio (0.4–1.0) led to the formation of geopolymer structures of enhanced UCS values by 66%. Higher values of R/Al molar ratios were not examined following the results of initial experiments. The extra silicon content of the alkaline solution (Si/R) follows in impact, exhibiting a 33% contribution on the UCS values. The maximum strength (27.4 MPa) was achieved on samples prepared by applying a Si/R ratio equal to 0.5. These observations confirmed the preliminary investigation on Si/R molar ratio (Figure 5). The type of alkali in the alkaline solutions has a minimal effect on the samples’ mechanical strength, contributing just 3%. Nevertheless, the incorporation of a 1:1 sodium and potassium mixture in the alkaline solutions gives better mechanical strength.

From the above analysis, the best synthesis conditions of BW geopolymers are the following: Si/R = 0.5, R/Al = 1.0 and Na/R = 0.5 when cured at 90 °C for 3 d. A 95% confidence interval suggests an ideal compressive strength of 39.4 ± 4.9 MPa. To verify the model’s predictions, we synthesized geopolymer specimens under optimal conditions, cured them adequately for seven days, and performed UCS measurements. The average UCS calculated from three specimens was 42.8 MPa. This value aligns with the projected range, affirming the reliability of the employed experimental design approach.

3.4. XRD Analysis

Figure 8 displays the XRD patterns of the BW precursor and selected BW geopolymer samples. The mineralogical profile of the BW precursor is typical, containing silicate phases (quartz, albite, diopside, muscovite and microcline), carbonates (calcite and buetschliite), and oxide (maghemite), with quartz being the main phase. The geopolymer samples retain the same mineralogical phases as the BW precursor, indicating that its crystalline components remain largely unaffected during alkaline activation. Therefore, it is the amorphous part of BW precursor that actively participates in the formation of the geopolymer network of the final products. Furthermore, a slight increase in the background intensity within the 20–40° 2θ range in the XRD patterns of BW geopolymers indicates the growth of the amorphous geopolymer matrix. Samples with low Si/R values (BWt2 and BWt4) favor the formation of crystallographic phases with a Si/Al close to 1, as in the case of certain zeolitic phases (phillipsite). The presence of zeolites in geopolymers has been linked to the formation of products with lower mechanical strength [46]. Conversely, BW geopolymers prepared with increased Si/R and R/Al molar ratio values form more stable geopolymer matrices with an absence of zeolitic phases, resulting in the improvement in mechanical strength (BWt6 and BWt9).

3.5. FTIR Analysis

The FTIR spectra (Figure 9) show no significant differences among the synthesized geopolymers. The BW precursor spectrum appears broad and featureless due to its glassy, heterogeneous nature. Between 800 and 1300 cm⁻¹, we see overlapping peaks from both crystalline and non-crystalline aluminosilicate materials. There is also a wide peak associated with the stretching of Si–O–T bonds (where T can be either silicon or aluminum in a tetrahedral shape). The hump reaches its highest point at about 1110 cm⁻¹ in the BW precursor, but it gets narrower and moves to lower values in geopolymers. This shows that more silicon is being replaced by tetrahedral aluminum, which is in line with the formation of an aluminosilicate network [47,48,49]. The Si–O–Si and Si–O–Al asymmetric bending vibrations (~460 cm⁻¹) are less sensitive to network changes [50]. The ~1635 cm⁻¹ band is related to the bending of O–H in water. The 875 cm⁻¹ band is linked to the stretching of Al–O and indicates the process of geopolymerization [50]. The peaks at 800 and 780 cm⁻¹, which are associated with the Si–O–Si stretching in quartz, are strong in BW precursor, but almost vanish in geopolymers. A 620 cm⁻¹ peak, indicating hexa-silicate ring vibrations, suggests nanocrystalline zeolitic phase formation [48]. All samples contain carbonate species, which can be seen from a band around 1450 cm⁻¹ that comes from the vibrations of CO₃²⁻ ions.

3.6. Microstructural Analysis

Figure 10 presents SEM images of the optimum synthesis (Si/R = 0.5, R/Al = 1.0 and Na/R = 0.5), which achieved the highest UCS value (42.8 MPa). The binding matrix appears relatively dense; however, it contains pores ranging from 80 to 300 μm. This porosity suggests that either improved compaction is required during specimen casting or that water trapped within the matrix evaporated during curing, contributing to the observed porosity in the geopolymer. Furthermore, the material appears to consist of a binding matrix, representing the geopolymer network, which encapsulates unreacted portions of the BW precursor. The matrix composition obtained from the EDX analysis (Figure 10c) is Si:Al:R = 4.1:1.2:1.8. The polygonal particles within the matrix (Figure 10b, particles indicated by the red arrows) exhibit a high concentration of Si, indicating the presence of quartz, which remains detectable (Figure 8).

4. Conclusions

This study led to the following conclusions.

• The ceramic part of Greek CDWs can be valorized as secondary raw materials in the development of greener building materials called geopolymers. Indeed, products with a wide variety of compressive strength values (2–43 MPa) were prepared depending the synthesis conditions.
• The leachability tests provided insight into the reactivity of CDWs to alkali activation. Brick waste exhibited greater sensitivity to alkali attack compared to tile waste, suggesting a higher potential for geopolymerization. This was further confirmed by synthesis experiments, which demonstrated that tile waste can be utilized either for producing low-strength building materials or for blending with the remaining ceramic fraction of Greek CDWs.
• The Taguchi method successfully identified the synthesis parameters that majorly affect the CDW geopolymer synthesis by targeting the maximum compressive strength of the prepared samples. Alkalinity of the activation solution plays the most important role in the CDW geopolymer synthesis, since it promotes CDW dissolution and offers charge balance in the newly formed geopolymer network.
• The optimum synthesis conditions to obtain CDW geopolymers of maximum compressive strength (42.8 MPa) are the following: Si/R = 0.5, R/Al = 1.0, Na/R = 0.5 (R = Na + K) at 90 °C for 48 h.
• Materials’ characterization showed that the successful preparation of CDW geopolymers is linked with the active participation of the amorphous aluminosilicate part of the precursor in the geopolymer synthesis, resulting in a new network where aluminum ions have substituted silicon in tetrahedral orientation. This network works as a binding matrix that encapsulates the unreacted part of the CDW precursor.