*3.6. Nitrogen Adsorption Results (Surface Area and Pore Structure of Cement Pastes)*

Figure 8 shows the comparison of cumulative nitrogen intrusion volume with respect to pore width for different paste samples after curing for 91 days. The results demonstrate lesser pore volume for ternary mixes having WSA together with SF as compared to control as well as binary mixes containing WSA. The relatively lesser pore volume of ternary mixes indicates formation of dense pore structure due to accelerated pozzolanic reactivity caused by the highly reactive SF. The least amount of nitrogen intrusion volume (0.043 cm3/g) was observed for a ternary mix with 40% cement substitution (WSA33SF7) followed by WSA25SF5 (0.045 cm3/g), WSA40SF10 (0.046 cm3/g), WSA10 (0.048 cm3/g), WSA20 (0.049 cm3/g), control (0.050 cm3/g), and WSA30 (0.051 cm3/g). The binary mixes with lower percent of WSA (WSA10 and WSA20) exhibited lesser nitrogen intrusion volume as compared to control, however, the maximum nitrogen intrusion volume was observed for WSA30. This is most obviously due to reduced pozzolanic reactivity when a high percent of cement replaced with WSA, which consequently had led to increased porosity and vascularity in the paste matrix.

**Figure 8.** Comparison of the cumulative nitrogen intrusion volumes versus pore widths for control, binary (C/WSA), and ternary (C/WSA/SF) paste samples after 91 days of curing.

A comparison of Brunauer–Emmett–Teller (BET) surface areas between control and other binary and ternary mixes is presented in Figure 9. All ternary mixes exhibited higher BET surface areas as compared to those of binary and CC. Similar to the least amount of nitrogen intruded volume, WSA33SF7 demonstrated a larger BET surface area (17.6 m2/g) as compared to all other mixes. An increased BET surface area of the ternary mix indicates its improved and denser microstructure of C-S-H gels [68,69]. As described in the preceding section, SEM analysis with EDS demonstrated a dense cementitious matrix with few pores. In general, BET surface area increased with increasing percent of WSA and SF in ternary mixes, except for WSA40SF10 (15.7 m2/g), which showed a slight reduction that could be due to a high amount of cement substitution leading to a large amount of unreacted WSA in the mixture [70]. Similar to ternary mixes, binary mixes, especially those containing relatively low percentage of WSA such as WSA10 (12.8 m2/g) or WSA20 (13.0 m2/g), also showed slightly larger BET surface areas than that of control (12.7 m2/g), which ultimately

had led to their improved microstructure of C-S-H gel. However, contrary to this, BET surface area decreased with further increase in percent of WSA (more than 20%) as was observed for WSA30 (12.0 m2/g). This clearly indicates the presence of unreacted WSA that had caused the hydration products to jam the pores and form a porous paste matrix. These results suggest a restricted use of WSA in binary mixes by not more than 20% when used as a sole substitute of cement.

**Figure 9.** Comparison of BET surface area between control, binary (C/WSA) and ternary (C/WSA/SF) paste samples after 91 days of curing.

#### *3.7. Thermogravimetric Analysis (TGA) of Cement Pastes*

The thermal decomposition of 91-day cured cement paste samples of control, binary, and ternary mixtures was performed by TGA to evaluate the effect of WSA and SF on the amount of C-H (%), which occurred due to the weight loss between 400 and 500 ◦C [71,72]. Based on the TGA results, a comparison of weight loss (%) with respect to temperature is presented between control, binary, and ternary mixes (Figure 10). Consequently, using the TGA results, the amount of C-H for different mixes and their normalized C-H (%) values with respect to OPC was calculated, as listed in Table 7. The normalized C-H values for each mix were calculated by dividing their C-H values with their respective OPC (%) content.

From Table 7, it can be seen that the maximum C-H content exists in control sample and a gradual decrease in C-H phase occurred in binary mixes with increasing percent substitution of cement with WSA. The gradual decrease in C-H content in binary mixes is obviously due to their decreasing cement content and partly because of the pozzolanic reactivity caused by WSA due to the formation of hydration products as a result of C-H consumption. However, as compared to binary mixes, a significant decrease in C-H phase occurred in ternary mixes, which was due to the simultaneous effects of high reactivity of SF along with pozzolanic reaction caused by WSA. This would ultimately lead to significantly high densification and compactness of C-S-H gel microstructure for ternary mixes as compared to binary and control mixes.

Similar to C-H, it can be further seen from Table 7 that the normalized C-H values of all the binary and ternary mixes are lower as compared to control. However, in comparison to ternary, all the binary mixes exhibited relatively higher normalized C-H values, which demonstrated their lesser pozzolanic reactivity. In fact, the same was also evidenced earlier

through XRD, SEM–EDS, and FTIR analyses. Moreover, the reason for relatively lower normalized C-H in ternary mixes, despite their large volume substitution of cement, is mainly due to the high reactivity of SF jointly with pozzolanic WSA, which ultimately causes a seeding effect to produce more C-S-H gel. These important findings with clear scientific proofs may justify the effectiveness of using both WSA and SF jointly as a high-volume replacement of cement for the production of a strong, durable, and sustainable concrete.

**Figure 10.** Comparison of thermogravimetric analysis of 91-day cured paste samples among control, binary (C/WSA), and ternary (C/WSA/SF) paste samples after 91 days of curing.

**Table 7.** Actual amount of calcium hydroxide (C-H) and normalized with respect to the percent content of cement in each mix (C-H/OPC) after 91 days of cement hydration.


#### *3.8. Global Warming Potential*

3.8.1. Comparison of CO2-eq for Unit Concrete Production among Control, Binary, and Ternary Concrete Mixes

Figure 11 shows the comparison of estimated GWP (kg CO2-eq per unit volume concrete) between control and other binary as well as ternary concrete mixtures. In this figure, the distribution of CO2-eq for each concrete mixture is illustrated by its ingredients and major production processes. The value of CO2-eq for all concrete mixtures was calculated by using the green LCA tool according to the data listed in Tables 4 and 5. Subsequently, the total emission for each concrete mixture was calculated by adding its direct and supply chain emissions associated with the quarrying, production, and transportation processes that occur within a systems' boundary.

**Figure 11.** Comparison of GWP distribution by concrete ingredient and phase between control, binary (C/WSA), and ternary (C/WSA/SF) concrete mixtures.

It can be seen in Figure 11 that among all the different ingredients used in unit concrete production, cement is responsible for maximum CO2 emission for all the mixtures. Following cement production, transportation of the raw materials and their products represents the second-highest source of CO2 emissions, which varies between 8.5% and 19.5%. Most importantly, the CO2 emissions associated with noncementitious materials are very low and remained almost same for all concrete mixes. According to the calculations, the SCMs (WSA and SF) constituted only 0.3% to 1.4% of total calculated CO2 emissions. Similarly, concrete batching and mixing account for very small percentages, between 0.4% and 0.5%.

According to the comparison, the CC exhibited the highest GWP at 533 kg CO2-eq/m<sup>3</sup> among all mixes (Figure 11). On the other hand, the ternary concrete mix containing WSA together with SF (WSA40SF10) as 50% cement substitution possessed lowest GWP at 313 kg CO2-eq/m3. Generally, by decreasing the amount of Portland cement in concrete mixes and increasing the amount of SCMs, the carbon footprint for cement production decreased from 89% for CC to 44% for WSA40SF10.

3.8.2. Comparison of Normalized CO2-eq for Unit Concrete Production per Unit Strength (MPa) among Control, Binary, and Ternary Concrete Mixes

Figure 12 shows the comparison of normalized values of GWPs as CO2-eq/m3/MPa among all the mixes, including control, binary, and ternary (Mix #1 to 7). As illustrated in Figure 12, the normalized values of CO2-eq/m3/MPa for all the mixes were drawn with respect to their compressive strength and aging (7, 28, and 91 days). The CO2-eq intensity of different mixes is used as a measure to evaluate their important impact on compressive strength of concrete and associated GWP per unit concrete volume and strength [73].

**Figure 12.** Comparison of normalized GWPs as kg CO2-eq with respect to strength among control, binary (C/WSA), and ternary (C/WSA/SF) concrete mixtures.

It can be seen in Figure 12 that each individual mix, in general, showed a gradual decreasing CO2-eq/m3/MPa trend due to increasing compressive strength with aging. Moreover, it can be further noted that the amount of CO2-eq/m3/MPa decreases with decreasing quantity of cement in all the binary and ternary mixtures. The only exception is the binary mix WSA30, where it slightly increased as compared to WSA20. This is due to drop of strength in WSA30 because of the high amount of cement substitution (30%) in this mix as compared to WSA20. A sharp decrease in CO2-eq/m3/MPa in a binary (WSA20) and ternary mix (WSA33SF7) is due to their high compressive strengths as compared to other mixes.

The current results dictate the possibility of lowering CO2-eq intensities by replacing cement with only one type of SCM (WSA) or blends of SCMs (WSA with SF). According to the current findings, the intensities of CO2-eq improved significantly in ternary blends as compared to binary, without compromising any strength potentials. For instance, despite of almost similar compressive strength of binary WSA20 (43.6 MPa) and ternary WSA33SF7 (44.6) mixes at 91 days, CO2-eq intensity decreased to 7.9 kg/m3/MPa in ternary mix (mix #6 having 60% cement) from 10 kg/m3/MPa in binary mix (mix #3 having 80% cement). Furthermore, the addition of higher amounts of WSA (40%) and SF (10%) in ternary mix (mix #7 having 50% cement) causes a further reduction in CO2-eq intensity as 7.8 kg/m3/MPa as compared to 12.5 kg/m3/MPa in binary mix (mix #2 having 90% cement). This might be due to slightly higher 91-day strength of ternary mix (40.2 MPa) as compared to binary mix (38.3 MPa). These findings are equally applicable for these mixes at other ages as well, such as 7 and 28 days. Consequently, these results suggest that the intensities of CO2-eq can be reduced without compromising the strength of concrete by optimizing concrete mixes with appropriate amount of cement replacement and selection of a suitable type of SCMs with their correctly chosen or tested blend percentages.

## **4. Conclusions**

This study investigated the use of high-volume WSA and its blend with SF as a partial substitute of cement for the development of high-performance and sustainable concrete. Besides control (100% cement), several other concrete mixtures were prepared by partially

substituting cement with only WSA as binary system (10%, 20%, and 30%) and WSA together with SF as ternary system (25%/5%, 33%/7%, and 40%/10%). Subsequently, the influence of adding WSA and WSA with SF on the hardened mechanical properties (compressive and tensile strengths, WA, and AP) was assessed and compared to those of CC. Finally, paste samples were prepared for all mixes to examine their microstructures and pore structures using SEM–EDS, FTIR, TGA, and N2 adsorption techniques to scientifically understand the impact of adding WSA and SF on the resulting paste matrix. At the end, GWP as kg CO2-eq per unit volume of concrete and kg CO2-eq per unit volume of concrete/MPa were also calculated using the LCA tool and compared among different mixes used in this study.

The main findings of this study are summarized as follows:

The current findings demonstrated a decrease in strength of binary concrete corresponding to their relatively low (WSA10) and high (WSA30) cement substitution rates, while a significant increase in strength was observed for moderate substitution rate of cement (20%) for binary concrete WSA20. In a very similar manner, the ternary mixes also showed a decreasing trend of strengths both at low (WSA25SF5) and high (WSA40SF10) blends of cement substitution, and the significant increase in strength was obtained for the moderate ternary blend of 33% WSA and 7% SF (WSA33SF7). Moreover, these increased strengths of WSA20 and WSA33SF7 were validated by their relatively lower apparent porosity and water absorption values among all mixes.

A correlation between experimental compressive and tensile strengths showed close agreement to models proposed by Noguchi–Tomosawa and De Larrard and Malier. Based on the results of current findings, it is recommended to use these models to properly estimate the tensile strength of tested concretes containing WSA alone or WSA jointly with SF, except the CC and WSA33SF7, as the JSCE model demonstrated a close agreement for these two concrete mixes. In addition, the JSCE model safely predicts the tensile strength of other binary and ternary concrete mixes with their slightly underestimated values.

Analysis of SEM–EDS data reveals that the incorporation of WSA as 20% replacement of cement (WSA20) leads to the densification of the paste matrix by decreasing the Ca/Si ratio in both the C-S-H and C-H phases. Furthermore, adding a 7% SF jointly with 33% WSA in the ternary mix (WSA33SF7) resulted in the lowest Ca/Si ratio among all the concrete mixtures tested. These findings suggest better refinement of microstructure for these mixes, which ultimately would lead to improvement of their engineering performance.

A visible shift in Si-O band was observed through FTIR analysis for almost all the binary and ternary mixes. Comparatively, the shift was more pronounced in all ternary mixes containing WSA together with SF, which clearly demonstrates the presence of high levels of C-S-H gels. Moreover, the portlandite peaks (3641–3644 cm−1) were also significantly smaller in all ternary mixes as compared to binary mixes. This, consequently, suggests an improved pozzolanic reactivity and the formation of more C-S-H gels that ultimately leads to a more refined and denser microstructure.

Similar to FTIR analysis, a densification of the paste matrix and refinement of the pore structure also suggested by the results of N2 adsorption tests was due to the decrease in intruded pore volume and an increase in BET surface area, especially for mixes WSA20, WSA25SF5, and WSA33SF7. However, other mixes such as those containing a large amount of WSA (WSA30 and WSA40SF10) showed a smaller surface area and more intruded pore volume. This could be possibly due to the presence of unreacted WSA in these mixes that might have caused high porosity and vascularity in their paste matrices.

As a matter of further validation of these findings, TGA results also showed a reduction in the portlandite phase of binary and ternary mixes. This occurred especially in those binary mixes that contained high doses of WSA (WSA20 and WSA30), partly because of a lesser amount of cement in these mixes and due to the pozzolanic reactivity of amorphous silica present in WSA. A further reduction in the proportion of portlandite phase occurred in ternary mixes due to very fine and amorphous silica present in both WSA and SF, which would ultimately consume the C-H phase to generate additional C-S-H phases and lead to densification of the paste matrix.

The GWP per unit volume of concrete (CO2-eq/m3) mixes decreased with decreasing amount of Portland cement simultaneously with an increase in amounts of SCMs. The highest GWP of 533 was generated by the control mixture (100% OPC) while the least GWP, only 313, was produced by the ternary concrete mixture (WSA40SF10) having 40% WSA jointly with 10% SF and only 50% OPC.

Regardless of aging, all the binary and ternary concrete mixes containing SCMs (WSA or WSA with SF) exhibited lower CO2-eq intensities as compared to CC, with the only exception of a binary mix having 10% WSA that showed almost identical CO2-eq intensities to that of control at later ages of 28 and 91 days. Furthermore, as compared to binary, ternary mixes having WSA together with SF showed good potential for further reducing the normalized CO2-eq intensities/MPa at all ages (7, 28, and 91 days). The ternary mixes containing highest percentages of SCMs, at 40% (WSA33SF7) and 50% (WSA40SF10), resulted in lower CO2-eq intensity/MPa as compared to that of CC, regardless of aging. Consequently, it may be safe concluding that the efforts for using larger amounts of regionally available SCMs can have an important positive effect on producing green concretes together with reduced GWP without compromising any strength potentials.

**Author Contributions:** Conceptualization, K.K., M.I., M.N.A. and K.S.; Methodology, K.K., M.I. and N.W. Validation, M.N.A. and K.K.; Formal analysis, M.N.A. and K.K.; Investigation, M.N.A.; Resources, M.N.A. and K.S.; Data curation, M.I.; Writing—original draft preparation, M.N.A., K.K., M.I. and N.W.; Writing—review and editing, M.N.A.; Visualization, M.I.F. and K.S.; Supervision, M.N.A., K.S. and K.K.; Project administration, M.N.A.; Funding acquisition, M.N.A., M.I.F. and K.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the Al Bilad Bank Scholarly Chair for Food Security in Saudi Arabia, the Deanship of Scientific Research, Vice Presidency for Graduate Studies and Scientific Research, King Faisal University, Saudi Arabia [Grant No. CHAIR21]. The APC was funded by the same "Grant No. CHAIR21".

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** All the data utilized in current research are available upon a reasonable request from the corresponding author.

**Acknowledgments:** The authors acknowledge the Al Bilad Bank Scholarly Chair for Food Security in Saudi Arabia, the Deanship of Scientific Research, Vice Presidency for Graduate Studies and Scientific Research, King Faisal University, Saudi Arabia, for the financial support (Grant No. CHAIR21).

**Conflicts of Interest:** The authors declare no conflict of interest. The founding sponsors had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; nor in the decision to publish the results.

### **References**

