*3.1. Material Characterisation*

#### 3.1.1. Physical Properties

The physical characteristics of the sand analysed through the sieve particle size distribution (PSD) were fineness modulus (FM) = 2.87; coefficient of uniformity (Cu) = 2.39; coefficient of curvature (Cc) = 0.94; dust content = 0.45%, specific gravity (SG) value of 2.65 and water absorption of 1.2% was recorded in the physical properties' tests and presented in Figure 1 and Table 2, respectively. The granite analysed for the study had a specific gravity of 2.7, water absorption of 1.05%, aggregate crushing and impact values of 28% and 11%, respectively.

**Figure 1.** Particle size distribution of aggregates.


**Table 2.** Physical and mechanical properties of aggregates.

The PSD plot for the MCC and CEM II is presented in Figure 2, which shows that 90% (D90) of CEM II and MCC particles are smaller than 4000 and 950 nm, respectively. The median particle size, D50 of CEM II and MCC are 48.8 and 450 nm, respectively. Furthermore, the particle size below 10% (D10) falls within 4.88 and 275 nm, respectively, for the same samples. From Table 3, the SSA measured via the SinglePoint and MultiPoint BET model showed 5.590 × 102 and 3.026 × 102 <sup>m</sup>2/g and 8.182 × <sup>10</sup><sup>2</sup> and 4.649 × <sup>10</sup><sup>2</sup> m2/g for CEM II and MCC, respectively. The pore diameter of the binders analysed through the DA BET mode indicated corresponding values of 2.92 and 2.88 nm for CEM II and MCC samples. Based on these values, CEM II is finer than MCC. The differences in particle size seen in the binders may be attributed to the different production methods used. Comparing the two binders DA BET analysis, the MCC sample has a lower pore size diameter, conforming to macro-mesoporous material [37].

**Figure 2.** Particle size distribution of MCC and CEM II.



## 3.1.2. Chemical and Microstructure Analyses of Binders

Figure 3 depicts the morphology of the dark brown MCC powder at 200 μm XRF magnifications, revealing that MCC powder particles are broad, solid masses of the wider surface area and irregularly shaped. Table 4 further reveals that MCC mainly comprises SiO2, Al2O3 and Fe203 while CEM II oxide components are CaO, SiO2 and Al2O3.

**Figure 3.** SEM Image of MCC.


The XRD phase spectra presented in Figure 4 shows that illite and kalicinite minerals dominate with 13% mineral contents of the MCC, followed by quartz with 8% mineral representation. Calcite and garnet recorded 6% mineral contents, respectively. Rutile recorded the least mineral content of 5% mineral.

The FTIR/ATR absorption spectra for MCC are shown in Figure 5. The spectra depict the intensities of the OH stretching, Si–O stretching and bending and Al–OH bending bands, indicating the development of kaolinite and illite minerals sensitive to cation exchange [38]. The OH stretching of the inner surface and the outer hydroxyl groups are observed at 3683, 3623 and 3534, 3414 cm−<sup>1</sup> for sandwiched octahedral sheets between two layers for kaolinites tetrahedral-octahedral-tetrahedral structure of illite. The high region with a band at 3623 cm−<sup>1</sup> resulted from the low frequency of kaolinites' inner surface hydroxyls, indicating kaolinite and illite minerals sensitive to cation exchange [15]. The bands at 1651 cm−<sup>1</sup> are due to the deformation of water molecules. Functional Al–OH bending

bands at 1033–913 cm−<sup>1</sup> typical for all smectite mineral clay groups are observed, thus validating the previous work on kaolin's inner hydroxyls [39]. The vibrational modes of the octahedral aluminium ions of kaolinite and the Si–O stretching of quartz are recorded for the MCC sample at 693 cm−<sup>1</sup> validating the XRD result.

**Figure 4.** XRD patterns of MCC.

**Figure 5.** FTIR/ATR patterns of MCC.

Characterising the MCC for TGA (as shown in Figure 6) was borne out of knowing the dehydroxylation trend of the samples. The studied samples' carbonation effects were controlled by maintaining the nitrogen environment with a 50 mL/min flow rate within the heating chamber. As shown in Figure 6, after subjecting the MCC sample to about 270 ◦C, the absorbed water located in the clay sample's interlayer space got dehydrated. Further measurement to about 350 ◦C to 450 ◦C, brought about dehydroxylation of possible kaolinite. At 500 ◦C, it shows complete dehydroxylation of the remaining kaolinite. At about 650 ◦C to 888 ◦C, the TGA curve remains flattened, showing the maximum calcination temperature to which MCC manufacturers subjected the raw clay material. Zhou [39] averred that at 800 ◦C, a compete dehydroxylation of illite and montmorillonite clay minerals appears. Garg and Skibsted [40] pointed out that illite/smectite mineral-based clay attains dehydration, dehydroxylation, amorphisation and recrystallisation at the temperature ranges of 25 to 200 ◦C; 600 to 800 ◦C; 800 to 900 ◦C; 950 ◦C and above, respectively. Thus, the MCC showed an amorphous phase considering the flattering of the TGA curve from 650 ◦C to 888 ◦C and supports the postulations of Garg and Skibsted [40].

**Figure 6.** TGA Curve of MCC.

#### *3.2. Fresh Properties*

#### 3.2.1. Slump Flow Test

Before casting into moulds, fresh HPC mixtures were examined for workability by slump flow test. The slump flow test results on the HPC mixtures containing MCC is presented in Figure 7.

**Figure 7.** Slump flow of HPC made with different contents of MCC.

The result demonstrates an average flow value of 520 mm for the control mixture. In contrast, the mixtures containing MCC show an improvement inflow up to 20% MCC content. Specifically, there is an improved flow of 15.39%, 12.50%, 8.65%, and 5.77% for MCCC-5 to MCCC-20 compared with the control. MCCC-25 recorded the same flow value as the control, while MCCC-30 had a 5.77% decrease in flow value. These findings suggest the MCC's ability to absorb the Masterglenium Sky 504—a PCE superplasticiser used in the study, causing the dispersion and water reduction tendencies affecting MCC-based HPC [41–43]. The lowest flow value observed by the control mixture may be linked to the more angular shape of the CEM II particle size as compared with the broad irregular shape of clay particle size, as corroborated by SEM characterisation. However, the data obtained from slump flow tests is consistent with HPC mixtures in literature, with a slump flow range of 450–600 mm [23].

#### 3.2.2. Initial and Final Setting Times

Figure 8 portrays the setting times test results for the various binary blend of HPC mortar mixtures using a concrete penetrometer under ASTM C403 [33]. From Figure 8, MCC-based HPC mixtures show that MCC's addition resulted in a gradual increase in the HPC's initial and final setting times. The initial setting time for control, MCCC-5, MCCC-10, MCCC-15, MCCC-20, MCCC-25 and MCCC-30, is 300, 420, 490, 540, 600, and 660 min, respectively and observed to be higher than the control. From the same Figure, the final setting times for control, MCCC-5, MCCC-10, MCCC-15, MCCC-20, MCCC-25 and MCCC-30 are 840, 840, 900, 960, 1020, and 1080 min, respectively. As observed from Figure 9, a relatively lower final setting time occurred in MCC modified HPCs than the initial setting time. The higher initial and final setting values seen for the HPC mixtures are linked to MCC's gradual addition into the HPCs. Cia et al. [44] inferred that the retardation of setting conditions for Portland cement pastes containing mineral clay is connected with the concentration influence of mineral clay and the partial performance of clay mineral's pozzolanic reactivity at an early age.

**Figure 8.** Initial and finals setting times of MCC-based HPC mortar.

**Figure 9.** Density of HPC made with different contents of MCC.

#### *3.3. Mechanical Properties*

#### 3.3.1. Density of HPC

The average dry densities of the control and various MCC-based HPCs at 7, 28, 56 and 90 hydration days are presented in Figure 9. The figure shows that the average density of the activated HPC samples containing MCC and the control varies from 2390 kg/m<sup>3</sup> to 2387 kg/m3 for all ages. It also showed that the densities slightly increased by 0.04% as the curing age increases. Comparing the individual HPCs containing MCC and the CEM II, MCCC-5, MCCC-10, and MCCC-15 showed an average increase in dry densities of 2.92%, 1.80% and 0.46% over the control, respectively.

MCCC-20, MCCC-25, and MCCC-30 showed decreased dry densities corresponding to 1.35%, 1.51%, and 1.62% compared to the control over the ninety-day observation. The initial increases in density with lower MCC contents (MCCC-5 to MCCC-15) may be attributed to the binary combination of the cementitious materials filling the voids between the fine aggregates, thereby achieving a denser assembly. The fact that the dry densities of MCCC-5, MCCC-10 and MCCC-15 were improved, which was not the case for MCCC-20, MCCC-25 and MCCC-30, could be the reason that MCC was added as CEM II, replacement by weight. When incorporated to substitute cement to a higher mass, MCC has a lower density, the total volume of powder (CEM II+ MCC) was increased. The addition of MCC's higher weight with PSD closer to CEM II may have reduced the HPC density. This result agrees well with Zhou's [39] reported lower density with the higher replacement of London calcined clay in concrete.

#### 3.3.2. Compressive Strength of HPC with MCC

HPC specimens' compressive strength data having MCC tested at 7, 28, 56 and 90 days are presented in Figure 10. HPC specimen with 5% cement replacement (MCCC-5) had the highest compressive strength at 7 days curing age, followed by MCCC-15, control, MCCC-20 and MCCC-25. There are comparable strength values between the control and the MCC-based HPCs. The control specimen showed a superior strength value at 28 days among other HPCs of the same age. On the other hand, the MCCC-10 HPC specimen had the highest compressive strength at 56- and 90-days curing ages, followed by MCCC-5 and MCCC-15 mixture types. MCCC-25 and MCCC-30 had a reduced strength at 56 and 90 days compared to control. The control and specimen with a higher MCC content had lower strength at later age following dilution effects and high content of calcined clay with low pozzolanic reactivity [6].

**Figure 10.** Compressive strength development of MCC-based HPC.

The HPC samples with higher MCC contents produce lower C3S, b-C2S phases, and high-water demand for cement hydration, giving room for decreased compressive strength [16]. The 56 and 90 days MCC-based HPC specimens gained strength at a greater rate than the control. Strength development continues beyond 28 days, and this result

demonstrated later age minor pozzolanic reactions, filler effect, MCC fineness and cement hydration accomplishment [45]. It can be inferred that the insignificant content of decomposed kaolinite found in the MCC also had a marginal compressive strength impact on the HPCs design mix.

As NBBRI Pozzolan (MCC) was never used for HPC development before, the comparison with other researchers' results had to be done with similar compositions. Vejmelková et al. [46] prepared a ternary HPC mix containing 10–60% CEM I, 52.5 R cement class replaced with an industry prepared calcined Czech claystone calcined at a maximum temperature of 700 ◦C and silica fume with a target design strength of 120 MPa. The XRD analysis revealed that the mineralogical compositions of the clay were mostly kaolinite and illite. The authors achieved over 120 MPa compressive strength with 10, 20 and 30% replacements at 28, 90, 180- and 365-days age curing. The best result of Vejmelková et al. [46] was observed at 30% of the clay with compressive strength of ~20% higher than the control. Thus, the attainment of over 120 MPa may be linked to the introduction of silica fume and 1:1 clay structure in the HPC matrix, leading to their samples' improved microstructure. Trümer et al. [13] obtained a C40/50 concrete class with a 30/70 ratio with a CEM I, 42.5 R and calcined montmorillonite-based clay. The raw bentonite's calcination up to 900 ◦C brings about the total decomposition of montmorillonite mineral culminating in the clay's amorphousness and attainment of the designed strength. Schulze and Rickert [45] reached a strength class of 42.5 N using calcined clay, irrespective of the clay minerals (kaolinite, montmorillonite and muscovite/illite) content investigated. Laidani et al. [16] reported compressive strength of 74 and 70 MPa as their most successful strength, with 5% to 30% of calcined illite and quartz mineral-based clay in the cement-based blend. They found that their best mix showed an improved compressive strength of 20% and 15% over the control.

Strong evidence emanating from the entire result showed that blended CEM II with MCC could not produce the target designed strength of Class 1 HPC (50–75 MPa) at 28 days, while later age curing showed a promising result. Garg and Skibsted [40] pointed out that when illite/smectite clays are blended with Portland cement in mortar and concrete; there appears to be a little form of clay reaction at an early age while there is usually a considerable amount of reaction in a clay-based mortar or concrete mixtures at a later age. A further shortfall in the rate of strength development at an early age may be linked to the MCC's lower calcination temperature below 700 ◦C.

## 3.3.3. Splitting Tensile Strength of HPCs with MCC

Splitting tensile strength results of HPCs with varied MCC contents is presented in Figure 11.

**Figure 11.** Splitting tensile strength results of HPCs at different treatments with MCC.

From Figure 11, the 28 days of hydration period of MCCC-5, MCCC-10, MCCC-20 and MCCC-25 enhances good splitting tensile strength value of the HPCs than the control. Only the splitting strength value of MCCC-10 was higher by 20.64% than the control specimen at 56 days. At the same age, MCCC-5, MCCC-15, MCCC-20, MCCC-25 and MCCC-30 decreased in strength with ~7%, 9%, 13%, 14% and 17%, respectively compared with control mix. Comparing 90 days splitting tensile strength values of MCCCs with control (4.61 MPa), MCCC-10 performed best with 6.15 MPa, followed by MCCC-5 with a slight margin of 4.67 MPa. Other MCCCs (MCCC-15, MCCC-20 and MCCC-25) recorded close splitting strengths values of 4.14, 4.37 and 4.59 MPa with control. Only MCCC-30 had a decreased value of about 16% compared with the control. The addition of MCC has a moderately positive effect on the splitting tensile strength of the HPC, especially at later hydration ages. These results indicate that 10% cement replacement with MCC (MCCC-10) is sufficient to accelerate 28, 56 and 90 days splitting tensile strength of the HPCs. These results were consistent with the splitting tensile strength result found in the literature [40,46].

#### 3.3.4. Flexural Strength of HPCs with MCC

Flexural strength (modulus of rupture) results of HPCs as influenced by the MCC content is presented in Figure 12.

**Figure 12.** Flexural strength of HPCs at different treatment conditions with RHA.

At 28 days, for Figure 12, the HPC mixes with MCC had higher flexural strength than the reference mix, which can be attributed to the pozzolanic reactivity of MCC compared to the control mix at 28 days. The MCCC-15 with 15% cement replacement with MCC had the highest flexural strength (1.40 MPa). As the hydration progressed for 56 and 90 days, the strength increased faster with MCC mixes than the reference, especially for mix MCCC-15 and MCCC-10, indicating that MCC's pore structure refinement was more efficient in these mixes than others. Possible explanations for the strength improvement are contained in the studies of Garg and Skibsted [40], Zhou et al. [6], Vejmelková et al. [46] and Trümer et al. [13].

## *3.4. Non-destructive Tests on Hardened HPC Mixtures*

SEM/EDX was conducted on selected HPC samples (Control, 10% and 20%) in furtherance to quantitatively assess the hydration products, molecular structure and the bond between the cement paste and aggregate at the interfacial transition zone (ITZ). The SEM images taken at 100 and 200 μm and oxides atomic concentrations for the control, MCCC-10 and MCCC-20 specimen at 90 days, are shown in Figure 13, Figure 14, Figure 15 and Table 5, respectively.

**Figure 13.** (**a**,**b**) SEM images of control HPC at 90 days.

**Figure 14.** (**a**,**b**) SEM images of MCCC-10 HPC at 90 days.

**Figure 15.** (**a**,**b**) SEM images of MCCC-20 HPC at 90 days.


**Table 5.** Oxides Atomic Concentration of the HPCs from EDX.

#### 3.4.1. SEM/EDX Analysis

Figure 13a,b depicts the SEM images of the reference HPC. The images revealed the general morphology and crystalline structure of the internal surface of the HPCs. The greyscale generally assists in identifying and analysing the specific elements for a welldefined chemical composition's evaluation. As shown in the SEM images taken at 100 μm, the part labelled 1 represents the sand of angular shape, darker in colour surrounding the coarse aggregate. The item labelled 2 represents the coarse aggregate, dark grey in colour, large, and irregular in shape. The cement paste matrix contains the light grey background portion, labelled 3 being spread all over the surface, which is assigned hydration products (C3S, C2S, C3A, C4AF, CH, C-S-H, AFt and AFm). The 200 μm image also revealed the presence of a consolidated bond at the ITZ and unified surface. There appears to be a good bond between the aggregate and the cement paste leading to a dense interface with

less porosity. Mindess, Young and Darwin [47] linked this phenomenon to factors such as chemical interaction between the aggregate and cement paste, surface roughness of aggregate, and micro filler inclusion. A good bonding system between the paste and aggregate can come from the chemical interaction between the concrete and the cement paste for the control mixture. The coarse aggregate's angularity is another important factor in forming the denser bond at the ITZ zone. Furthermore, the compositions of the oxide (Table 5) revealed specific atomic concentrations (%) by each constituent, giving silicon (34.35%), calcium (18.78%), oxygen (22.24%) and aluminium (10.20%) prominence in the phases. On the other hand, potassium traces (2.94%) and sodium (2.59%) were found in the constituents' phases.

Figures 14a,b and 15a,b show the SEM micrograph of MCCC-10 and MCCC-20 HPC specimen cured for 90 days. The SEM images indicated bright grey colour irregular shaped coarse aggregate surrounded by darker angular shaped sand. The cement paste matrix highlights the light grey background fragment spread all over the surface, which is assigned hydration products (C3S, C2S, CH, C-S-H, C-A-H, C-A-S-H, AFt and AFm). As can be seen from the images, there is evidence of a more dense and uniform transition zone between the cement paste and aggregates than the control. The addition of 10% and 20% of MCC resulted in a more compact and cohesive paste than the control sample; this can be correlated with these mixtures' high strength. This fact can be related to MCC's pozzolanic activity, which generated a close grid of hydration products with fewer portlandite residues. As shown in Figure 15a,b, the MCCC-20 mixture produced a few spot dark pores that interfered with the HPC's strength compared with the MCCC-10 mixture. EDX result (Table 5) of MCCC-10 and MCCC-20 mixtures showed the dominance of calcium (38.99%; 70.90%), silicon (28.78%; 5.08%) and oxygen (22.22%; 15.37%) in the elemental atomic weight % compositions, respectively. This result points to the formation of C-S-H and portlandite in the HPC mixtures. The elevated calcium formed in the tested MCC blended HPCs at 90 days of curing indicates portlandite formation compared to the reference mix.

#### 3.4.2. XRD Analysis

The XRD diffraction pattern was used to determine the crystalline phase in the hardened HPC pastes, and the crystalline phase's amount at 90 days of hydration. XRD analysis from Figures 16–18 present the XRD patterns of control and MCC blended HPCs.

From Figure 16 quartz (SiO2), calcite (CaCO3), portlandite (Ca(OH)2, phlogopite (KMg3AlSi3O10(F,OH)2), biotite (K(Mg,Fe) 3(AlSi3O10)(F,OH) 2), andradite (Ca3Fe2Si3O12) are the major mineral phases of the hardened pastes. Calcite recorded the highest mineral phase (16%), possibly due to the constituent materials' carbonation during the production and samples preparation process. Quartz, phlogopite and biotite were 8%, 7%, and 5% of mineral phase contents. Portlandite and andradite recorded the lowest crystalline phase contents of 3% and 2%, respectively. Thus, these mineral compositions detected via XRD analysis are typical for a calcined clay blended cement [6,13,27,46]. XRD patterns of MCCC-10 and MCCC-20 mixtures at 90 days of hydration are demonstrated in Figures 17 and 18, respectively. As can be seen from the Figures, the percentages intensities of portlandite peaks reduced to 3% for both MCCC-10 and MCCC-20 mixtures, indicating depletion of portlandite in the MCC blended cement with the control sample. This phenomenon is consistent with the improved mechanical properties results of this study.

**Figure 16.** XRD pattern of the hardened control sample at 90 days.

**Figure 17.** XRD pattern of the hardened MCCC-10 sample at 90 days.

**Figure 18.** XRD pattern of the hardened MCCC-20 sample at 90 days.

#### **4. Conclusions**

In this paper, the potential use of MCC as an SCM in a binary blended cement for HPC production was investigated for the underlying mechanisms. The binder consists of MCC and PC, and the W/B was 0.3. SAP was introduced at 0.3 bwob as an internal curing agent to increase internal moisture availability and prevent autogenous shrinkage. MCC was substituted at 5%, 10%, 15%, 20%, 25% and 30% of PC. The fresh properties were determined by slump flow test and setting times techniques, and the findings were compared with the recommendations of Neville [23] and ASTM C403 [33] for HPC mixtures. The hardened HPC samples were compared by compressive, splitting tensile and flexural strengths. The microstructural and mineralogical phases of the selected hardened HPC samples were analysed via SEM/EDX and XRD advanced techniques. The following inferences are drawn from the study.


EDX result of MCCC-10 and MCCC-20 mixtures showed the dominance of calcium, silicon and oxygen in the elemental atomic weight% compositions, respectively;

• The hardened HPC control sample at 90 days revealed quartz, calcite, portlandite, phlogopite, biotite, andradite as the major mineral phases of the hardened paste. The percentages intensities of portlandite peaks reduced to 3% for both MCCC-10 and MCCC-20 mixtures, indicating depletion of portlandite in the MCC blended cement with the control sample.

**Author Contributions:** Conceptualization, D.O.N. and B.J.O.; methodology, D.O.N., B.J.O. and B.G.F.; investigation, D.O.N., B.J.O. and O.I.F.; writing—original draft preparation, D.O.N.; writing—review and editing, B.J.O., O.I.F. and B.G.F. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding, and the APC was funded by Covenant University Center for Research, Innovation and Discovery (CUCRID).

**Acknowledgments:** The authors extend their appreciation to the suppliers of the superplasticiser-Masterglenium Sky 504—BASF Limited, West Africa; the Superabsorbent Polymers (SAP)—SNF Floerger-ZAC de Milieux, France and 100 mm cube metal moulds—the Nigerian Building and Road Research Institute (NBRRI), Ota.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Abbreviations**

