The Water treatment hematite-rich sludge (WTHs) is collected from Camwater Co. Cameroon. The kaolin is from Kribi, locality of South Cameroon. The sludge is left to settle for 24 hours in a plastic container, then dried in open air (24˚C ± 3˚C), crushed and sieved to a particle size of 75 µm. Meanwhile, the kaolinitic clay was enriched in clay fraction by wet sieving to pass through a 75 µm mesh and then dried in an oven for 24 hours at 105˚C. The dry product obtained is heated in an electric furnace of type NGS-SRL for 4 hours at 700˚C with a temperature rate of 5˚/min. The thermal treatment allowed the kaolin to be transformed into metakaolin which is more amorphous and reactive [4][13]. The activating solution was prepared by mixing equal volumes of the aqueous solution of sodium hydroxide (8M) and sodium silicate (SiO₂: 28.7%; Na₂O: 8.9% and H₂O: 62.4%).

The formulation of geopolymer cement is made by mixing metakaolin, the WTHs powder and activating solution in the ratios shown in Table 1. The ratio is selected as the best consistency after several tests using the Vicat apparatus. Three formulations are then retained; the first one, labelled MKS0 was made by mixing only the solid precursor, namely metakaolin with the activator solution and was used as reference cement. Two other formulations where 5% and 10% by mass of metakaolin were substituted with sludge powder and denoted MKS5 and MKS10 were done respectively. Above 10% substitution, the paste loses its homogeneity and is transformed into small agglomerates. The mixture was stirred using an M&O brand mixer, model N50-G until a homogeneous paste is obtained. One part of the fresh paste was used to determine the setting times and the other poured into cylindrical molds (diameters = 20 mm and heights = 40 mm) in PVC for further analyses. After casting, the specimens obtained were covered with a thin polyethylene film then left at room temperature in the laboratory (24˚C ± 3˚C) for 24 hours. The mechanical measurements were carried out on cylindrical specimens on day 28 according to standard EN 196-1 [14][15].

Table 1. Formulations of geopolymer cement pastes.

The starting materials (MK and WTHS) were characterized using inductive Couple Plama (ICP) for chemical composition, XRD, amorphous phase content, Humidity and dryness, FTIR spectroscopy for structural composition and the BET specific surface area (SSA).

Geopolymer cement performance were evaluated via the measurement of the setting time according to the EN 196-3 using the Vicat device [16], linear shrinkage using a digital caliper on cylindrical test discs, Archimedes measure of apparent density and open porosity. The compressive strength is measured according to standard EN 1961 [17][18]. Optical microscopy was carried out using a Ceramics Instruments Model 101T-M7 optical microscope equipped with a stereomicroscope, a binocular head and a tablet with integrated camera.

The dryness rate and water content of the sludge powder were found to be 93.89% and 6.49% respectively. These results indicate that the sludge used is classified as dry because its dryness is greater than 85% [19].

The chemical composition of metakaolin is given in Table 2.

Table 2. Chemical composition of metakaolin.

*LOI: Loss of Ignition.

Metakaolin is mainly composed of silica (SiO₂) and alumina (Al₂O₃), the main oxides for geopolymerization. The combination of Na₂O and K₂O values suggests the presence of micaceous minerals and feldspars. The data also reveals a low LOI value, which therefore suggests a low quantity of organic matter. The Si/Al molar ratio of 2.1 is in quite good and is in accordance with other work specially that of Fernàndez-Jimènez et al. (2005) who demonstrated that an aluminosilicate precursor reacts easily in the presence of an alkaline solution when Si/Al molar ratio is less than 4 [20][21].

Table 3 presents the values of the specific surface area (SSA) of the starting materials. The specific surface area of MK is greater than that of WTHs powder, which may lead to greater reactivity of the aluminosilicate powder compared to the sludge.

Table 3. SSA values of metakaolin and WTHs.

Figure 1 presents the structural phases of starting kaolin (Figure 1(a)) in comparison with those of metakaolin (Figure 1(b)).

Figure 1(a) highlights the presence of kaolinite [Si₂O₅Al₂(OH)₄], illite [ɤ-FeO(OH)], lepidocrocite [K_x+y[(Si_4-xAl_x)O₁₀(Al_2-yFe(II)_Y)(OH)₂]], and quartz (SiO₂). After the amorphization of the clay fraction at 700˚C, the disappearance of the kaolinite was observed (Figure 1(b)). The presence of a dome around 2θ =15˚ - 25˚ on the MK curve was also noted, which expressed the existence of an amorphous phase in this material [11]. This amorphous phase is obviously that of metakaolinite following dehydroxylation. This observation corroborates the rate of amorphous phase obtained which is 65.15%. The rate of 50.13% for the sludge evidences that it is mainly made up of amorphous. This is in agreement with the appearance of the diffractogram of Figure 1(c) which presents several halos characterizing amorphous phases. The only crystalline phase observed is hematite, a coloring oxide which could explain the reddish color of WTHs.

Figure 1.XRD patterns of (a) Kaolin, (b) Metakaolin and (c) WTHs: Kaolinite (K), Illite (I), Lepidocrocite (L), Quartz (Q) and Hematite (H).

Figure 2 presents the FTIR spectrum of the sludge. A broad band is observed at 3288 cm⁻¹, a band not assigned until now. The absorption bands around 1366 - 1603 cm⁻¹ are attributed to the stretching vibrations of the C=O, C=C, and C=O groups which belong to organic matter [22]. The band observed at 906 cm⁻¹ which expresses the deformation vibration of the Al-OH bonds and that at 999 cm⁻¹ which reflects the Si-O-Si bond vibration [23][24]. Finally, the absorption band at 515 cm⁻¹ indicates the vibration of Fe-O attributed to the hematite in accordance with its XRD diffractogram. This high concentration of hematite also explains the strong red color of the WTHs [25][26].

Figure 2. FTIR spectrum of the WTHs.

3.2.1. Setting Time

The initial and final setting times shown in Figure 3 exhibit the decreasing of setting times with increasing WTHS contents from 69 - 105 min for 5% and 58 - 90 min for 10%. The formation of the geopolymer gel strongly depends on the availability of silicates and the aluminates ions which have a tendency to condense to form Si-O-Si and Al-O-Al bonds respectively [27]. The sludge may promote the dissolution of the aluminosilicates and reduce the alkaline cations by entrapping them in the medium. It also promotes the process of polycondensation of the silicate and aluminate monomers, hence the reduction in the setting time. In terms of setting acceleration, the reduction in setting time observed in this study (58 - 90 min for MKS10) aligns with findings by Sekizkades et al. (2025) for transition metal oxide-modified geopolymers, where Fe₂O₃ reduced setting times by 40-60%. This setting accelerating effect could be advantageous for numerous applications where the use of the paste requires a reduced setting time for the work.

3.2.2. Fourier Transformed Infrared Spectra

Figure 4 shows the FT-IR spectra of the geopolymer cements. The bands at 3400 - 3450 cm⁻¹ and 1640 - 1650 cm⁻¹ correspond to deformation vibration of the H-O-H bond and the elongation vibration of the O-H groups respectively [28]. The absorption bands around 976 - 987 cm⁻¹ reflect the presence of aluminosilicate gel in the products obtained [24]. The absorption bands around 773 - 776 cm⁻¹ correspond to the vibrations of the connections Al(VI)-OH and Al(VI)-O where aluminum in VI coordination. Those around 666 - 692 cm⁻¹ express the symmetrical elongation vibration of the Si-O and Al-O bonds, which shows that the aluminum initially in coordination (VI) in metakaolin occupies the coordination (IV) in geopolymers [29]. The shoulder around 547 - 566 cm⁻¹ expresses the deformation vibrations of the Si-O-Si and Si-O-Al bonds of the network [24]. Finally, the absorption bands around 408 - 428 cm⁻¹ and 382 - 392 cm⁻¹, characteristic of vibrations of Ti-O and Fe-O bonds. No new link was established due to the presence of sludge.

Figure 3. Initial and final setting times.

Figure 4. FTIR spectra of (a) MKS0, (b) MKS5 and (c) (MKS10).

3.2.3. Linear Shrinkage

Figure 5 indicates the changes of the linear shrinkage of the geopolymers as a function of age. It increases very slightly with age for formulations without or with 5% sludge and it increases rapidly with the addition of sludge to 10%. In all cases, the linear shrinkage remains very low, 0.5% for MKS0, 0.6% for MKS5 and 1.7% for MKS10. The increase in shrinkage with the addition of sludge is related to its humidity which is a measure of its water retention capacity in its structure. That sludge used is quite high because its water content is 6.5%. In fact, the shrinkage is due to the departure of the moisture water, a physicochemical process of diffusion on the surface leaving voids to be filled, which generates an internal readjustment of the system which evolves towards a more consolidated and stable state. In general, these shrinkages remain relatively within the acceptable limits of dimensional shrinkages for geopolymer binders reported by Elimbi et al., 2011 [4].

Figure 5.Plots of linear shrinkage versus curing time.

3.2.4. Apparent Density

Figure 6 shows the evolution of the apparent density of the geopolymer specimens as a function of sludge content and curing age. A decrease in apparent density is observed, obviously with age but an increase with the addition of sludge is also observed. This increase is more pronounced on the 21st day: 2.19 g/cm³ for MKS0, 2.49 g/cm³ for MKS5 and 2.51 g/cm³ for MKS10, and on the 28th day: 1.83 g/cm³ for MKS0, 1.90 g/cm³ for MKS5 and 1.97 g/cm³ for MKS10. This means that sludge promotes the densification. Indeed, the sludge seems to promote the agglomeration of particles and their rearrangement which results in a more compact structure during the geopolymerization process, due to the progressive reduction of water either by its integration into the poly-sialate network or by release due to drying.

Figure 6. Plots of apparent density versus WTHs content at day 21 and 28.

Figure 7 exhibits the open porosity of the cements as a function of the sludge dosage and curing duration. There is an increase in porosity with increase in sludge. This indicates that sludge, despite its effect of agglomeration and packaging of particles, also promotes the creation of pores. Such a dense structure which integrates pores indicates the presence of a three-dimensional interconnection where the particles in grains are distributed inside the matrix. The whole giving a dense network with pores which are close either to the matrix or to a grain and others which are open [30]-[32]. Obtaining one or the other type of porosity depends on several parameters including the addition of materials such as sludge in the case of geopolymer formulations. In the present work, the sludge used was directly introduced into the geopolymer paste. It has been demonstrated in previous works that this method generates open macroporosity characterized by pronounced interconnectivity.

The porosity-density relationship in this study (1.83 - 1.97 g/cm³ with 67% - 77% porosity) shows interesting parallels with S. Ma et al. (2023) hierarchical porous geopolymers. However, while their systems achieved porosity through templating methods, the WTHs naturally induces interconnected porosity, potentially offering manufacturing advantages.

Figure 7. Open porosity at day 21 and 28.

3.2.5. Compressive Strength of Geopolymer Cements

The mechanical resistance of the specimens as a function of the sludge content on the 28^th day is shown on Figure 8. The compressive strength values obtained (14.36 - 20.04 MPa) are comparable to those reported by G. Liang et al. (2022) for lightweight geopolymer foams (8 - 15 MPa) but lower than optimized iron-rich geopolymers such as those by reported by S. Ahmed et al. (2020) achieving 25 - 35 MPa with red mud incorporation. This difference can be attributed to the higher porosity (67% - 77%) in our WTHs-modified systems versus 45% - 60% in red mud-geopolymers. Considering those results, the sludge acts globally as a non-reactive filler, but with a noticeable impact on the compressive strength of the geopolymers.

Figure 9 presents the micrographs of MKS0, MKS5 and MKS10. For a magnification of 0.3X, pores are observed in all three geopolymer formulation while for a magnification of 0.7X, pores are observed only in MKS0. This suggests that MKS0 has macropores while MKS5 and MKS10 have micropores. The decrease in pore size of MKS5 and MKS10 could be explained by the addition of sludge which occupies the voids and makes the material dense. Furthermore, it appears that in all cases, the microstructure is more or less homogeneous with a slight difference in the connectivity of the particles in the matrix. Macropores appear on the matrix surface; this observation is correlated with the very high open porosity [33]. This observation confirms the interconnected microstructure which results in a three-dimensional structure, a configuration which is accentuated with the addition of sludge.

Figure 8. Compressive strength versus WTHs content at day 28.

Figure 9. Optical microscopy images of the cement pieces MKS10 ((a) and (b)), MKS5 ((c) and (d)) and MKS0 ((e) and (f)) at 0.7X and 0.3X magnifications.

3.2.6. Physical Appearance

Figure 10 shows the physical appearance of the specimens. MKS0, the specimen with no sludge had pale yellow color while the others are reddish due to the presence of hematite (containing Fe³⁺) in the sludge.

The sample MKS0 on the 28^th day, presented some whitish areas with a saline appearance. This observation is generally due to efflorescence favored by the fairly slow rate of the geopolymerization process [34][35]. Indeed, when dissolution is slow, CO₂ from the air penetrates the geopolymer matrix and reacts with the alkaline cation to form bicarbonate. The salts formed during efflorescence are generally sodium or potassium bicarbonate or hydroxycarbonate. Regarding the source of K⁺ or Na⁺ cations, they come from the solid and liquid precursors and constitute the part of the initial mixture which is not involved in the geopolymer gel network. Not only does it reduce the aesthetic appearance but also is a defect for the durability of the product. This phenomenon did not appear on MKS5 and MKS10 formulations which contain the sludge. This is very interesting given the fact that efflorescence is a degrading effect. Considering that sludge contains very little or no sodium ions to create excess of alkaline ions, hence the absence of efflorescence. Therefore, the role of the sludge is highlighted since it changed the typical behavior of aluminosilicate precursors by limiting the excess of ions in the system. The absence of efflorescence in WTHs-containing formulations contrasts with many waste-incorporated geopolymers and suggests that the hematite-rich sludge may act as an alkaline buffer [36], a phenomenon requiring further investigation through leaching studies and pore solution analysis.

Figure 10. Physical appearances of the samples at (a) day 1 (b) day 7 and (c) day 28.

A Comparison has been done between some properties of the present geopolymer cements and other works in the literature. Table 4reports the comparison of compressive strength, apparent density, and porosity values with some research works.

Table 4. Comparison of geopolymer cements properties after 28 days.