Strength and Durability Performance of Hybrid Alkaline Clay Brick Waste –Coconut Shell Ash Cement

2.1. Blaine test and Preparation of Mortar Prisms

The specific surface area of the binders was determined according to the Blaine test protocol outlined in the K S EAS 148:-6:1 (2000) standard. Mortar prisms were prepared for Testing by KS EAS 148: 1 [15] using 1,350 g of standard sand complying with EN 196. A mixture of 450 g of test cement and standard sand was prepared using a PAN Analytica mixer machine for 1 minute. Subsequently, an alkaline solution of 0.5 M or 2 M Na2SO4, was added to achieve the desired consistency. The mixture was trowel-mixed for another 4 minutes to ensure uniformity in the cement mortar. The resulting slurry was then transferred into a three-pong mold, measuring 40×40×160 mm for compressive strength tests and 100×100×100 mm for durability studies. The molds were clamped onto a vibrating machine for 2 minutes to compact the mortar. Upon removal from vibration, the mortar prisms were immediately identified by their casting date and the type of cement used, denoted as 5-1-4, 5-4-1, 3-1-6, and 3-6-1, representing the mix mass ratio of OPC:CSA: CBW as detailed in Table 1. To safeguard against moisture loss caused by evaporation, a polythene sheet that was impermeable and flat was used to cover the prisms. These prisms were then placed in a temperature-controlled room with relative humidity exceeding 90% for 24 hours±30 minutes, after which they were demolded and air-cured for 28 days. Similarly, PC specimens were also prepared for reference purposes, but potable water was used as the mixing liquid, and they were cured in water for 28 days.

2.2. Compressive Strength Testing

The 28-day compressive strength was determined using an automatic compression testing machine (model YAW300) housed at Mombasa Cement Limited, a prominent cement manufacturing company in Kilifi County, Kenya. The testing procedure strictly followed the protocol outlined in KS EAS 148: 1 (2000).

2.3. Differential Thermogravimetric Analysis (DTA)

A Stanton Redcroft STA-781 device was used for the thermal analysis of raw clay, coconut shell powder, and hydrated cement. 25 g each of the clay, coconut shell powder, and hardened cement were placed in a platinum crucible and the thermal analyzer. The measurements were between

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 374–390, December 2024

Figure 1. Preparation of specimens from the cube [16]. 25 ºC and 1000 ºC, and the heating rate was set to 10 ºC per minute. Thermal analysis of clay and powder samples was performed under atmospheric air conditions, whereas measurements for hydrated cement samples were conducted under a nitrogen atmosphere.

2.4. Durability Test

The experimental materials' permeability, sorptivity, Porosity, and durability were assessed by following the durability testing manual recommended by [16]. Two methods are available to evaluate the level of connectivity of pores in a concrete matrix: oxygen permeability index (OPI) and sorptivity [17]. The durability properties' quantification involves measuring the flow rate derived from Darcy's formula for laminar flow through a porous medium. 2.4.1. Preparation of Specimens for Durability Test Durability tests were conducted following the protocol outlined by Alexander and coworkers [16]. Cement blocks of dimensions 100 millimeters that were allowed to harden for 28 days were bored perpendicular to the pouring direction and divided into slices, as illustrated in Figure 1, to obtain two specimens, each with a diameter of 70±2 millimeters and a depth of 30±2 millimeters. The samples were labeled with an identifying number on their exterior using a permanent ink marker. Before experimentation, the labeled samples were permitted to dehydrate for seven days in a drying chamber regulated to 50 oC. After the samples were extracted from the drying chamber, they were cooled in an airtight container for 120 minutes until achieving a regulated indoor temperature of ±23 oC. 2.4.2. Oxygen Permeability Index Test The South African standard [18], which prescribes the use of the oxygen permeability index (OPI), was used to assess oxygen permeability. The test setup consisted of a 70 mm diameter core fixed in a rubber sleeve and attached to a permeability cell, with a 30 mm thick piece as a cover con-

crete. The aim was to observe the oxygen pressure drop in this section. The method involved measuring the pressure decay of oxygen passed through a 30 mm thick slice (representing the cover concrete) of a 70 mm diameter core placed in a rubber collar secured on top of a permeability cell. Then, the permeability cells were placed in the test apparatus. The cell was prepared for insulation by exposing it to an oxygen pressure of 100±5 kPa. The pressure and time factors were then recorded accordingly. The test was terminated 6 hours ±15 minutes after the start, or at a pressure of approximately 50 kPa, with pressure lags recorded at intervals of approximately 5 kPa. Using Equation 1, we calculated the Darcy permeability coefficient, K, in meters per second. (1) Where K is the permeability coefficient in m/s; ω is the molecular mass of oxygen, 32 g/mol; V is the volume of oxygen under pressure in permeameter (m3), g is the gravitational acceleration whose value is 9.81 m/s2, R is the universal gas constant, 8.313 Nm/Kmol. A is the superficial cross-sectional area of the sample (m3), d is the average sample thickness (m), 𝛉 is the absolute temperature (K), t is time (s) for pressure to decrease from Po to P, Po is the pressure at the beginning of the test (kPa), and P is the pressure at the end of the test. Subsequently, the oxygen permeability index (OPI) was determined as in Equation 2. Oxygen permeability index=-log (k) (2) 2.5.3. Testing for Sorptivity and Porosity The disc samples were subjected to permeability and water sorptivity tests according to the protocols described in the durability handbook [16]. Samples were tapped up to 5 mm above the test surface to ensure the unidirectional flow of the caliper. They were then placed in a dish of calcium hydroxide solution, ensuring that the solution level was just above the edge of the sample. The samples were weighed at 3, 5, 7, 9, 12, 16, 20 and 25 minutes, respectively.

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 374–390, December 2024

This was followed by vacuum saturation of specimens by applying 75 kPa suction. Finally, the specimens were submerged in water saturated with Ca(OH)2 for 18 hours and fifteen minutes. The samples were re-weighed, and their Porosity was calculated using Equation 3. (3) Where: Msv is the vacuum saturated mass of the samples to the nearest 0.01; Mso is the initial mass of the specimen to the nearest 0.01 g at a time to; A is the cross-sectional area of the specimens to the nearest 0.02 m2; d is the average specimens' thickness to the nearest 0.02 mm; Pw is the density of water, 10-3 g/mm3. The mass of the water absorbed at each weighing period is calculated using Equation 4: Mwt=Mst – Mso(4) Where: Mst is the mass to the nearest 0.01 g of the specimen at time t. Sorptivity was calculated from the slope of the graph of water absorbed (Mwt) versus the square root of time (an hour) using Equation 5. (5) Where: F is the slope of the best-of-fit line obtained by plotting Mwt against t½; S is sorptivity

2.5. Correlation Analysis for Permeability Test

Parameters The intercorrelations between the results of the permeability tests were compared with the compressive strength performances for all cement mixes produced. The factors correlating permeability included oxygen permeability index (OPI), Porosity, and sorptivity. Principal component analysis (PCA) was used to determine the correlation. A scree plot for the eigenvalues was plotted to determine the number of significant components. The "elbow" for the eigenvalues (Fig. 2) falls on eigenvalue 2, indicating two principal components. The first two factors explain a variance of 94.82%. The selected principal components were used to plot the loading, biplots, and score plots and analyze the existing correlations. Chemometrics Agile Tool (CAT) was used to perform the PCA, which works with the programming language R and the algorithm NAPLES (Nonlinear Algorithm for Partial Least Squares).

2.6. Migration Test

Following the procedures described in ASTM C1202-97 (2001), three samples of the 50% OPC replacement and the reference cement were allowed to cure in water for 28 days before being tested for accelerated chloride ion migration. In an electrochemical test setup, 500 cm3 of a 3.5% NaCl solution was added to the cathode compartment, while an equal volume of a 0.3 M NaOH solution was added to the anode compartment. Epoxy-coated mortar prisms were positioned in the test cell for each cement category. Stainless steel electrodes were attached, and the arrangement

Figure 2. Scree plot for the eigenvalues. was clamped in place. A potential difference of 12±0.1 V was applied for 36 hours, and the current was monitored every 30 minutes. The containers were covered with polyethylene paper to prevent evaporation, and glass rods were used to shake the solutions continuously. After exposure, the prisms were removed, the surfaces abraded, and cut into layers. Depending on the distance from the aggressive media, the samples were dried, pulverized, and packaged before being subjected to chloride analysis. 2.6.1. Chloride Analysis 10 g of ground samples were mixed with 75 ml of deionized water for one hour using a shaker. After a settling time of 30 minutes, the sample was treated with 25 ml of strong nitric acid to dissolve the cement lumps. The mixture was stirred, rinsed into a beaker, and allowed to stand for 4 minutes. The filtrate was heated and filtered. It was then poured into a volumetric flask. A sodium hydrogen carbonate solution was introduced, and water was added until the total volume reached 100 ml. The resulting solution was split into three equal parts, each titrated with a 0.1 M silver nitrate solution. This titration was performed three times on each part. A chloride-free CaCO3 blank was run through the same procedure. Using Equation 6, the concentration of chloride ions in the original sample was determined based on the titration results. (6) Where V1 is the final volume of the extract, V2 is the volume of the blank, V is the volume of silver nitrate, and M is the molarity of silver nitrate used. 2.6.2. Determination of Diffusion Coefficients The standard procedures outlined in Kenyan standards [19] were adopted to determine the chloride concentration in the tested samples. These results were represented in chloride profiles. From the graphs, chloride diffusion

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 374–390, December 2024

Figure 4. Particle size distribution results for CBW and CSA.

coefficients, Dmig, were derived from error fitting curves in Equation (7) using the method of least squares [20, 21].

60 minutes at the desired temperature, two thermocouples were used to measure the surface and internal temperature differentials. After allowing the samples to cool, they were weighed, and the thermal expansion was measured with a vernier caliper. Mass loss due to the high temperatures was measured with an electronic scale. The residual compressive strength was documented, and the thermal resistivity was calculated using Equations 10 and 11, considering both weight loss and strength loss as a measure of thermal stress.

(7) C is the concentration of the Cl- ions at a depth x, and at the moment, t and D are the diffusion coefficients. The diffusion of chloride ions in concrete is approximated from the solution to Fick's second law equation [22] under non-steady-state conditions assuming boundary conditions boundaries C(x, t)=0 at t=0, 0 < x < ∞ and C(x, t)=Cs at X= 0, 0 < t <∞, constant ion interactions, linear chloride binding, and one-dimensional diffusion into a semi-infinite solid. Crank's solution to Fick's second law of diffusion is given by Equation 8. (8) The surface concentration (C(S)), migration diffusion coefficient (Dmig), and the chloride ion concentration (C(x,t)) at any depth (x) within the mortar bulk at time t constitute the equation. The Gaussian error function is obtained in computer spreadsheets as the error correction function, abbreviated as erf. Apparent diffusion coefficients, Dapp, are calculated from the following equation [20, 21]. (9) In this context, R stands for the gas constant, F represents the Faraday constant, T indicates the electrolyte temperature measured in Kelvin, Zi denotes the valency of the ion ΔE refers to the effective applied voltage measured in volts, and t represents the duration of the test or exposure measured in seconds.

2.7. Determination of Thermal Resistance

The prepared mortar prisms (40 mm x 40 mm x 160 mm) of 5-1-4 and OPC cement paste were cured for 28 days and pressure tested after exposure to an electric furnace at 5.0 °C per minute. To maintain thermal stability for

(10) W1 and W2 are the specimen weights before and after heating, respectively, and WL is the loss in weight. (11) CL is the loss of strength, while C1 and C2 are the compressive strength of the mortar before and after heating, respectively.

3.1. Blaine Test of the Precursor Materials, Blended

and Commercial Types of Cement The specific surface area (SSA) is an essential factor in determining the reactivity of binders. Figure 3 represents the SSA for the binders. In this figure, the particles' specific surface area (in g/cm²) of the test materials is displayed as measured by the Blaine air permeability analysis All the blended cement had higher SSA values above the minimum 3401 cm2/g of the standard OPC. The mechanically activated precursor (CSA/CBW) materials show high SSA of 3902 and 4504 cm2/g and are responsible for the overall high SSA in the binders. Increased CBW substitution resulted in a higher surface area on the blended cement than the addition of CSA. As it reacts, a layer of hydration product forms around the outside of the particle, separating the unreacted core of the particle from the surrounding water. As this layer grows

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 374–390, December 2024

Figure 6. TG-DTA analysis for the raw clay sample. thicker, the rate of hydration slows down [23]. This implies that a small particle will react faster than a large particle. Thus, CBW is mechanically more reactive than CSA.

3.2. Particle Size Distribution of the Precursor Materials

The particle size distribution of the mechanically activated CSA and CBW are represented in Figure 4. The particle size distribution of any cementitious material significantly affects its performance. The graph provides a detailed visual representation of the granular composition of these materials after mechanical activation. The particle size distribution of any cementitious material significantly affects its performance. The CBW had a slightly higher particle size distribution compared to CSA, with more than 90% of its particles being below 45 microns in size, as shown in Figure 4. Thus, the CSA contains larger particles compared to CBW. This could result from the unburned carbon particles as heat in the furnace was not controlled. It is known that the material passing the sieve 45 µm is more reactive [24]. This indicates that both precursor materials are potentially mechanically reactive [25]. The reaction rate of the active precursor ma-

terials/cement particles primarily depends on their sizes. [26] worked on the microstructural and physical characterization of different solid wastes from clay bricks for reuse with cement. They observed that the clay brick waste sample particle with a small radius (large surface area) was more pozzolanic. Smaller particles hydrate more quickly in water than larger particles [27]. Typically, particles with a diameter of 1 µm react entirely in about one day, while one with a diameter of 10 µm reacts completely in about one month. It is assumed that one with a diameter larger than 50 µm never becomes fully reacted, even in the presence of a sufficient water source [27]. Arslan [28], in their study on the dissolution kinetics of iron oxide leaching from clays by oxalic acid, reported that small powdered particles [29] had a high dissolution.

3.3. Thermal analysis of the precursor

The thermal analysis carried out on the precursors used in this study is shown in Figures 5 and 6. The graphs represent the various phase transformations that occur when coconut shell powder and raw clay are heated at different temperatures.

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 374–390, December 2024

Figure 7. Comparison of the XRD pattern for the raw clay (top) and CBW (bottom) clay. I: Illite; Mg: Magnetite; Q: Quartz; H: Hematite; W: Wollastonite; N: Nepheline; K: Kaolinite; F: Feldpar.

The TG/DTA analysis of the raw coconut shell powder shows a smaller region of mass loss at lower temperatures (below 250 oC), accompanied by an endothermic peak up to a temperature of about 150 oC. It could be associated with evaporation of the moisture [30]. This is preceded by a region of thermal decomposition and mass loss, followed by a sharp exothermic peak near 400 °C, further associated with the formation of coconut shell activated charcoal [31]. Additionally, it could be related to the combustion of the organic lignin/cellulose fraction. A broad endothermic peak with gradual mass loss is observed at temperatures above 800 oC. This was related to the thermal decomposition of hemicellulose. These findings agree well with previous work reported by [30], who obtained 38.44% and 47.54% mass loss between 200–280 °C and 280–365 °C, which was linked to the decomposition of cellulose and hemicellulose. The TG-DTA graph of the clay sample is represented in Figure 6. An endothermic peak below 200°C is observed. This peak is associated with the thermal loss of physisorbed water within the clay interlayer, which is often mobile and loosely bound. At temperatures between 250 °C and 450 °C, an endothermic peak representing the dissolution of water molecules in the interlayer ions' initial strongly bonded coordination sphere is observed. This could further result from the hydroxylation of the functional groups connected to the clay minerals being removed at this temperature range [32]. According to Arslan [28], the condensation and dehydration of structural hydroxyl groups are indicated by an exothermic peak at 450–470 °C in the DTA curve. An endothermic peak related to removing hydroxyl groups from kaolinite is observed at 550 °C. Similarly, at approximately 573 °C, a smaller peak is observed, indicating a transition from quartz-α to quartz-β [33]. The results also show a broad endothermic peak spanning from 600 °C to almost 940 °C, which is attributed to the dehydroxylation of the silica-alumina content present in illite, forming an amorphous substance. At temperatures above 950 °C, a sharp exothermic peak is detected, indicating the formation

of crystalline silica-alumina phases. These findings suggest that the optimal temperature for calcination of kaolinite clays should be below this threshold [34].

3.4. Mineralogical Analysis of CBW and CSA

The mineralogical analysis of the precursor materials of the blend cements is shown in Figures 7 and 8. The figures show various minerals present in raw and calcined clays. The XRD pattern for the raw and CBW clay (Fig. 7) reveals the presence of various minerals, including illite, magnetite, quartz, hematite, wollastonite, nepheline, kaolinite, and feldspars. The calcined clay brick waste shows the disappearance of the kaolinite upon thermal treatment, whereas illite loss of crystallinity is evident. Furthermore, a collapse of the illite interlayer is observed, with the basal peak shifting to a higher 2θ value. Additionally, hematite and magnetite (Fe oxides) formation is noted, along with the disappearance of feldspars and the emergence of wollastonite and nepheline. The results in Figure 8 show an XRD analysis of the CSA, revealing a series of peaks associated with crystalline SiO2 (quartz). These features may indicate burning temperatures exceeding 700 °C [35]. Peaks related to K-chlorate, a phase previously observed in vegetable ashes burned at temperatures above 400 °C, are also identified [36]. The two broad scattering haloes observed around 20° and 27° 2θ angles are related to an unburned organic fraction [37]. Figures 7 and 8 show silica-alumina minerals' presence in the two precursors. Calcined clay, with its high alumina and silica content, promotes the formation of calcium-silicate-hydrate (C-S-H) and calcium-aluminate-silicate-hydrate (C-A-S-H) gels, enhancing the mechanical properties and durability of HACs. The mineralogical compositions of pozzolanic precursor materials, such as CBW and CSA, play a pivotal role in the hydration process and microstructure development in hybrid alkaline cement [38]. Rich in silica, the ash from coconut shells aids in the creation of C-S-H and refines the pore structure, creating a matrix that

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 374–390, December 2024

Figure 8. XRD pattern of coconut shell ash (CSA). is more robust and compact [39]. Recent studies further highlight the importance of understanding the mineralogical properties of these precursors to optimize the performance of hybrid alkaline cement [40, 41].

3.5. Compressive Strength Test for Cement Mortars

The compressive strengths development of the binders (3-6-1, 3-1-6, 5-4-1, and 5-1-4) activated by different concentrations of Na2SO4-alkaline activators (0.0, 0.5, and 2 M) and different OPC:CSA:CBW ratio at 28 days of curing are shown in Figure 9. They are compared to conventional cement. Generally, all blended cement prepared using water exhibited low compressive strength compared to alkali-activated blends and OPC (Fig. 9) after 28 days. This could be attributed to the slow strength development due to the incorporated pozzolana's slow reaction nature. The OPC substitution effect is demonstrated by decreased compressive strength across all the blends with OPC substitution from 50% to 70%. The strength of OPC was almost double that of 70% replacement, with 3-6-1 giving the lowest strength. This can be explained by more proportionate unreacted filler materials (CSA/CBW), hence weak bonding among them [42]. The results show that each formulation (50 and 70 OPC replacement) with high CBW replacement exhibited high compressive strength. Fine particles of calcined clay can explain this and are bound to fill the pores more effectively, resulting in a denser and stronger matrix. Grinding calcined clay brick waste reduced accumulation and decreased the average particle diameter, improving its reactivity and performance in the cement matrix. The optimal particle size distribution ensures better packing density and enhances the mechanical properties of the cement, particularly its compressive strength [43].

Figure 9. 28-day compressive strength (𝑓𝑐𝑢) results for 0.5 w/c binder mortars compared to Portland cement. The compressive strength of the activated blended cement achieved about 30% and 50% increase with the addition of the 0.5 M and 2 M Na2SO4 in each blend. A considerably significant increase in compressive strength with alkali activation addition in H2O<0.5 M<2 M across all blends is observed. The 50% OPC replacement blends displayed better compressive strength performance than OPC at 2 M Na2SO4, with 5-14 having the highest compressive strength. The presence of SO42- ions in the activator solutions within the cement mortar matrix may explain why chemically activated mortars achieve greater compressive strength values. The amount of alkaline activation products (gels) formed is responsible for the mechanical development of the alkaline hybrid binders [44].

3.6. Durability Index (DI) Test Results

Based on the details provided by Bakera and Alexander [45], this is a synopsis of the findings of the DI tests for oxygen permeability and water sorptivity using CBW and

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 374–390, December 2024

Table 2. Judgement criteria for the quality of concrete durability index (DI) tests Qualitative description

CSA. The quality of the concrete was evaluated using the standards laid out in Table 2. Sorptivity measures the rate of fluid movement through the concrete under capillary suction, while OPI measures the degree of interconnectivity of the pores in the matrix. The Porosity of the cement mortar and the sorptivity are influenced by the degree of hydration and the ratio of weight to cement [46]. The permeability of the sample decreases with increasing OPI value and vice versa. Conversely, high sorptivity/porosity values of cement matrix indicate high permeability. Permeability determines how easily liquids can penetrate and flow through the concrete when exposed to a pressure gradient. 3.6.1. Binder Oxygen Permeability Index The oxygen permeability test results for different binder types cured for 28 days are given in Figure 10. The figure illustrates the variation in permeability based on precursor content and alkali concentration. The values provide a quantitative measure of the concrete's resistance to oxygen ingress, with higher OPI indicating lower permeability and enhanced durability. The results, based on Durability Index (DI) judgment criteria in Table 2, show that all blended cement mortars prepared using water had OPI values below 9.5 (log scale), signifying "poor" quality. This value was lower compared to PPC and OPC mortar. As can be observed in Figure 10, there was a notable decrease in OPI in blended cement in the order 5-1-4>5-4-1>3-1-6>3-6-1. This order was replicated for the blends prepared with the activator at the two different concentrations. Hybrid alkali-activated mortars exhibited higher OPI compared to the respective non-activated mortars. They had values above 9.5 (log scale), which signifies "good" quality. There was also an increasing trend in OPI values with an increase in the concentration of NaSO4, the alkali activator, for each cement mix. The 2 M 2 Na2SO4 HAC showed higher OPI values than those of 0.5 M Na2SO4, with all having OPI values above 10 (log scale), which signifies "excellent" quality. This could be attributed to the presence of SO42- ions in the activator solutions in the cement mortar matrix. SO42- in mortars has been reported to form ettringite crystals, which fill the hardened cement mortar cracks and pores, lowering the permeability [47]. The ettringite formation is as shown in Equation 12. 6Ca2++2[Al(OH)4]-+3SO42-+26H2O→Ca6Al2(SO4)3(OH)12.26H2O(12)

The formation of this product in the early days of curing causes densification of the hardened cement matrix.

Figure 10. Oxygen permeability index (OPI) results for 0.5 w/c binder mortars cured for 28 days. It was clear that the substitution effect was a significant factor in the development of the cement matrix. A negative OPI response was pronounced on the increased substitution of CSA compared to that of CBW. This could be linked to the powder's particle size distribution, which is not uniform. The pozzolana "micro-filler effect," in the fine and ultrafine size range in combination with Portland cement, can also enable higher packing densities of the mortar or cement mixture [48]. Heterogeneous nucleation can also be attributed to this effect as a second physical phenomenon that could be significantly related to small particle size distribution. The pozzolana fine particles settle between the clinker crystals, lowering the energy barrier to allow hydrates to form on foreign fine particles. The finer the particles, the faster the reaction. This aspect could contribute positively to the quality of permeability property of 5-1-4 compared to 5-4-1 and 3-1-6 compared to 3-6-1 mortar cement. Generally, all the HAC blends offered reduced permeability. This could be linked to refined pore structure and the ability to bind chloride ions chemically, limiting chloride ingress and protecting the concrete. Additionally, their high alkalinity maintains a protective passive layer on the steel. These properties make hybrid alkaline cements particularly durable in chloride-rich environments like marine settings and roads treated with de-icing salts [49]. After 28 days of curing, hybrid alkaline cement blends exhibit superior resistance to chloride ion penetration compared to traditional Portland cement, highlighting their enhanced durability in aggressive environments [50]. Continuous curing after 28 days [51] reported significantly improved flexural strength, compressive strength, and resistance to chloride penetration of alkali-activated slag mortar with 2–3% alkali contents. 3.6.2. Porosity of Different Mortars Tested The Porosity of the different mortars tested is given in Figure 11. The graph with a high porosity indicates a high chance of fluid penetration into the mortar matrix. This highlights the variation in Porosity across the samples,

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 374–390, December 2024

Figure 11. Porosity (%) results for 0.5 w/c binder mortars cured for 28 days. which directly impacts their mechanical and structural properties and, thus, their performance under different aggressive environments. All the blended binders in water exhibited higher Porosity compared with OPC mortars (Fig. 11). This trend is more pronounced at a higher precursor replacement of 70%. The blend containing 60% CSA displayed the highest Porosity of 51% higher than the standard OPC mortar. Based on these findings, it appears that the pore structure of the OPC mortar is more complex and intricate than that of mortars made with pozzolana cement. The presence of more visibly enlarged macropores in the pozzolana mortar during its early curing stages could be the reason behind this [52]. The reaction of pozzolanic precursor materials with Ca(OH)2 is a slow process, forming calcium-silica-alumina (C-S-A) gel, an insoluble cementitious material [47]. The formed insoluble crystalline particles may fill the pores of the hardening concrete, making it dense with refined pores at a later age. Compared to non-activated mortars, alkali-activated hybrid mortars reduced Porosity by more than 10% in all cement categories. The 5-1-4 hybrid mixture of 0.5 M and 2 M Na2SO4 showed the highest reduction in Porosity at 19% and 36%, respectively, while the 3-6-1 mixture showed the lowest reduction at 6% and 26%, respectively. Increasing the dosage of Na2SO4, an alkali activator, also showed a similar trend. The corresponding 2 M Na2SO4 HAC showed lower Porosity than 0.5 M Na2SO4 HAC. This could be due to the low water absorption in the OPC hybrid copolymer due to the presence of chemical activators that increase the dissolution and gel formation of silica and alumina, resulting in the formation of more C-S-H gels and AFm-type phases as well as C–A–S–H with longer chains. The dense C-S-H gels formed by the chemical activation of aluminosilicate materials such as calcined clay brick waste/coconut shell ash lead to reduced Porosity in the properties of the chemically activated materials [51]. The C-S-H gels formed during the dissolution and gelation of silica and alumina are structurally similar to those formed during the hydration of cement [53].

Figure 12. Sorptivity results for 0.5 w/c binder mortars cured for 28 days. According to Angulo-Ramírez et al. [54], the low Porosity of HB mortar indicates that the pores have been refined and the pore network in this system has become more tortuous. Compared to OPC, hybrid cement mixtures exhibit exceptional pore refinement [55]. This behavior can be associated with the potential of the blended cement structure to reduce Porosity and compactness, which were caused by the existence and production of a strongly cross-linked gel (C-A-S-H) with longer chains than the C-S-H gel commonly seen in Portland cement. The good behavior of alkali-activated CBW- CSA mortar is generally explained by the coexistence of aluminosilicate gels, such as C-A-S-H. 3.6.3. Water Sorptivity Test of Different Mortars Tested The water sorptivity test results for various binder types, all cured for 28 days, are given in Figure 12. The different binders include OPC, PPC, and CSA/CBW blended cement prepared in water and activated with alkali. This highlights differences in water absorption rates, directly influencing the material's durability and long-term performance. The results provide the suitability of each binder type for use in environments where moisture exposure is a concern. Hybrid cement prepared with 2 M Na2SO4 had lower sorption capacity than those prepared with 0.5 M Na2SO4, suggesting that the Porosity of the chemically activated cement decreases with increasing Na2SO4 concentration. In alkali-activated cement, chemical activators promote the dissolution of silica and alumina, which favors C-A-S-H gel formation; this could be the reason for the low water absorption of the cement. Since aluminosilicate materials such as CBW-CSA form dense C-A-S-H gels, chemically activated cement has a lower sorption capacity [56]. The architectures of the C-A-S-H gels are similar when the cement is hydrated or activated with alkali [53]. An alkali activator ranging from 0 M to 4 M NaOH was used by [57] to activate phosphogypsum binders and ground granulated blast furnace slag. Their data showed that low molarity was the most porosity-promoting factor.

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 374–390, December 2024

Figure 13. 2D factors loading plots and biplots. (a) Loading plot (94.8% of total variance). (b)Biplot (94.8% of total variance). OPI: Oxygen Permeability Index; Cs: Compressive strength; PTY: Porosity; SPY: Sorptivity; 361H, 316H, 541H, and 514H are blends prepared with water, 361W, 316W, 541W, and 514W are blends with 0.5M Na2SO4 and 361S, 316S, 541S, and 514S are blended with 2M Na2SO4.

In this study, sorptivity reduced as the CBW/CSA ratio increased, according to experiments with varying ratios (1:4 to 4:1). Sample 5-1-4 showed decreased sorptivity independent of the alkali concentration. This could be associated with the high specific surface area of CBW (Fig. 3) in hybrid-blend cement that has been found to enhance its durability significantly. Nano-sized particles, such as nano-silica, improve cement's mechanical properties and durability by filling micro-pores and refining the pore structure. This optimization leads to a denser microstructure, which enhances resistance to water and chloride ingress, ultimately increasing the lifespan of the cement. Additionally, these nanoparticles promote secondary hydration reactions, further contributing to improved durability by creating additional C-S-H gels, which are crucial for the structural integrity of the cement [58]. The activation conditions and precursor characteristics significantly affect the hybrid binder's effectiveness, as stated by Balun and Karataş [59]. The results obtained by Bernal and colleagues [60] from their investigation into the engineering and durability properties of a combination of slag and alkali-activated metakaolin were similar. The results showed that water absorption and permeability were reduced when activator concentrations were increased and metakaolin was substituted within a specific range. According to several researchers, chemically activated cements have shown mixed results in water absorption and permeability compared to OPC concretes. While using slag as the precursor, Mithun and Narasimhan [61] found that an alkali-activated binder had lower total Porosity, less water absorption, and less chloride ion penetration than an OPC. Air permeability and sorptivity are two areas where AACs made of slag typically underperform regular concrete for a given strength grade [62]. Similarly, Albitar et al. [63] noted that compared to

geopolymer concrete, OPC concrete exhibited reduced sorptivity and water absorption. Precursors, activators, specimen maturity, and testing procedures are potential causes of the observed discrepancy between hybrid binders and OPC [64]. 3.6.4. Correlation Analysis for the Permeability Test

3.7. Chloride Ingress

3.7.1. Chloride Profiling The chloride profiles against the depth of cover of each category of the mortar cubes are shown in Figure 15. These involved determining the concentration of the chloride ions (×10−10 g/g) at different depths of the test binders' mortar bulk at 0.5 w/c ratios.

Figure 15. Chloride ions concentration at different depths of penetration for 0.5 w/c binder mortars cured for 28 days.

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 374–390, December 2024

Table 3. Dapp, R2, and Cs (%) values for different cement mortars

Figure 16. Chloride error function fitting for 5-1-4 (2 M). Dapp=4.20 x 10-10 m2/s, and Cs=1.89% (R2=0.98).

3.7.2. Apparent Diffusion Coefficients for Chloride Ions Figure 16 displays the error fitting curve for calculating the apparent chloride diffusion coefficient (Dapp) for the test binder, 514 (2 M) mortar. The chloride apparent diffusion coefficients for each binder category were found using similar error-fitting curves. Table 3 also shows the Dapp statistics for each binder group. As the penetration depth increased, the total chloride was observed to be reduced across all binder types (Fig. 15). The alkali-activated blends, blended with more calcined clay brick waste, compared to coconut shell ash, showed an increased reduction in chloride penetration. The chloride concentration in all binders drops drastically at a depth of less than 2 cm. The exposed surface's proximity to the aggressive medium may explain this. Every form of cement binds a certain amount of chloride. Chloride binding slows the chloride's diffusion into the binder's bulk. As observed by Ngui et al. [65] this can be explained by the cement mortar's chloride ion binding capability and diffusivity penetrating the cement mix. According to the profiles, PPC had less chloride infiltration and a lower Dapp than OPC (Fig. 15 and Table 3). Moderately substituted pozzolana cements have been found by other researchers to have a lower diffusion coefficient [66]. The pore solution of mixed cement contains less CH than that of OPC, which could explain this. According to Wang et al. [67] blended cement has a low CH content because of a pozzolanic reaction source, the source primary OH and -source; a reduction in OH-concentration reduces the pore water's ability to exchange OH for chlorides [68]. Blended cement is particularly effective in lowering chloride-induced corrosion due to higher aluminate contents than plain Portland cement. Al2O3 is known to bind Cl- ions, forming Friedel's salt. This enhances the chloride binding capacity, decreasing the free chlorides resulting in lower chloride ingress. Tang et al. [69] noted that the C-S-H gels, particularly in blended cement with a low calcium-to-silicon ratio, react with chloride salt, producing Friedel's salt. This salt can seal the interior pores, improve the distribution of pore sizes, and make the material more resistant to corrosion caused by chloride ions. The results (Fig. 15 and Table 3) showed that increasing CBW from 1 to 4 decreased the Dapp at all mixing liquids. The converse was true of Dapp on increasing CSA. Thus, the blended cement's Dapp was 5-1-4<5-4-1 at all mixing liquids.

5-1-4 (2 M) 5-1-4 (0.5 M) 5-1-4 (H) 5-4-1 (2 M) 5-4-1 (0.5 M) 5-4-1 (H) OPC PPC

OPC: Ordinary portland cement; PPC: Pozzolanic portland cement.

This supports the argument in 3.3 of CBW's superior mineralogical properties compared to CSA Figures 7 and 8. This contributed positively to the mortar microstructural development [24, 70]. Alkali-activated blend cement offered greater resistance to chloride ingress than OPC at any given alkali concentration. There was an increasing trend of Dapp with the increase of alkali-activator concentration as well as the increase of CSA from 1-4 as [5-4-1(H) < 5-1-4(H) < PPC] < OPC < [5-4-1(0.5 M) < 5-4-1(2 M) < 5-1-4 (0.5 M) <5-1-4 (2 M)]. The introduction of an alkali activator increases the rate of alumina-silicate dissolution. As a result, the concrete becomes denser, attains higher resistivity, and reduces chloride diffusivity due to the greater cementitious material produced by the more pozzolanic reaction. The cementitious material can be produced via alkali-activated cement with increased time [71]. According to Abdulkareem et al. [72], this causes the final mortar or concrete to have less permeability. Chemical activation at an 80% BFS replacement level with a high dose of 10.17 kg/m3 KOH was able to speed up hydration and portlandite consumption, according to [72].

3.8. Thermal Resistance

After exposure to various thermal loads, quantitative evaluation of the hardened paste behavior was determined by measuring their residual strengths following heat exposure. Heat strength profiles of 5-1-4 and OPC pastes binders after being heated to 100, 300, 500, 700, and 800 °C are displayed in Figure 17. On exposure of cement paste to temperatures ranging from 25 to 100 oC, OPC's compressive strength dropped sharply while 5-1-4 blends increased marginally (Fig. 4). According to Zivica et al. [73], crystalline products can be formed in the structural pores of pozzolana blend and alkali-activated paste when the reaction rate of the unreacted materials is increased by raising the temperature. A possible explanation for OPC's weakening at 0 to 100 oC could be pore water evaporation. At temperatures between 100 and 300 oC, the compressive strength of 5-1-4 blends rapidly decreased. The loss of water content in the specimens is likely to cause this. Free and chemically bonded water within specimens swiftly migrates to the surface and evaporates as the external temperature rises from ambient to 300 oC. This leads to a quick loss of strength in the 100–300 oC range and harms the specimens' internal microstructure [74]. Once the temperature reaches 300 oC, the strength of 5-1-4 blends gradually de-

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 374–390, December 2024

Figure 17. Results of the effect of heat on 5-1-4 blends compared to OPC. creases in the order of 5-1-4 (2M) < 5-1-4 (0.5M) < 5-1-4(H), indicating a small impact of temperature on strength deterioration in that order. Deterioration in strength also slows down about 300 oC because water evaporation decreases. OPC at temperatures beyond 300 oC continued to show a relatively steady decrease in compressive strength to total failure at 800 oC. This could be associated with the deterioration of the CSH and Ca(OH), the principal binding phases of the hydrated OPC mortar. OPC has previously indicated an irreversible loss of strength at a temperature above 200 oC [75]. Jeon et al. [76] reported that the main cause of OPC strength reduction on exposure to high temperature was mainly due to the expansion of lime after chilling and that the breakdown of Ca(OH)2 had minimal impact. The OPC system was found to degrade at elevated temperatures up to 600 °C, while Rashad [77] found that Na2SO4-activated slag exhibited superior resistance. Their research showed that a higher concentration of Na2SO4 and finer slag resulted in superior heat resistance. Figure 17 shows that at room temperature and after exposure to high temperatures, the compressive strength of CBW-CSA-blend pozzolana and alkali-activated-based paste is greater than that of OPC mortar. Because of this, structural components made with CBW-CSA blend alkali-activated-based mortar can offer the same or even greater fire resistance, and there are many other environmental benefits, including reduced energy consumption and carbon dioxide emissions.