Applied Clay Science 243 (2023) 107080

Contents lists available at ScienceDirect

Applied Clay Science

journal homepage: www.elsevier.com/locate/clay

Research Paper

Prioritization of low-grade kaolinite and mixed clays for performance

evaluation of Limestone Calcined Clay Cement (LC3):
Multi-criteria assessment
Nosheen Blouch a, Khuram Rashid a, *, Idrees Zafar b, Mounir Ltifi b, c, Minkwan Ju d, *
a
Department of Architectural Engineering and Design, Faculty of Civil Engineering, University of Engineering and Technology, Lahore, Pakistan
b
Department of Civil Engineering, Imam Mohammad Ibn Saud Islamic University, Riyadh, Saudi Arabia
c
Department of Civil Engineering, National Engineering School of Gabes, University of Gabes, Tunisia
d
Department of Civil and Environmental Engineering, Yonsei University, Seoul, Republic of Korea

A R T I C L E I N F O A B S T R A C T

Keywords: Limestone calcined clay cement (LC3) is an emerging ternary binder, while typically only kaolinitic clays have
Limestone calcined clay cement (LC3) been used or investigated for its production. There is a dire need to scientifically investigate the potential uti­
Low-grade clays lization of low-grade kaolinitic and mixed clays. In this study, nineteen different natural clays were examined
Multi-criteria decision-making
from various locations in Pakistan, with a focus on clay quarries near cement plants. From the dataset, seven
Characterization
Performance
significant criteria were identified: mineral type, and the contents of kaolinite, SiO2, Al2O3, Fe2O3, CaO, and the
Al2O3/SiO2 mass ratio. Nineteen clays were characterized using X-ray diffraction, X-ray fluorescence, and
thermogravimetric analysis. These clays were then prioritized against the identified criteria using multi-criteria
decision-making analysis. After characterizing them, the top ten clays were calcined to evaluate the performance
of LC3. A correlation analysis was conducted between the characterization of calcined clays and the compressive
strength of LC3. It was revealed that all criteria were positively correlated with compressive strength, except for
Fe2O3 and CaO, which showed a negative correlation. The correlation coefficients for Fe2O3 and CaO were −
0.44 and − 0.47, respectively. According to the OPC Type I, incorporating low-grade kaolinite with a compo­
sition of 40–55% leads to greater compressive strength gain compared to the standard 42.5 MPa compressive
strength of OPC Type I. Even clays with lower grades, ranging from 38 to 41 MPa, can achieve up to 96.5% of the
compressive strength of OPC Type I. Therefore, low-grade clays could be incorporated into a ternary binder
system to produce carbon-neutral cement.

1. Introduction materials (SCMs). This approach is widely regarded as the most practical
and robust alternative. However, most conventional supplementary
The global demand for cement is expected to increase in line with the cementitious materials (SCMs) being studied, such as fly ash, slag, and
growth of the construction industry. However, this increase in demand is silica fume, are in limited supply (Scrivener et al., 2018). Alternatively,
paradoxical because the use of cement should actually be reduced due to calcined clay and limestone are emerging as promising SCMs and are
the significantly higher levels of carbon dioxide (CO2) emissions it being used to create a ternary blend known as “LC3” (Limestone
produces, which contribute to climate change (IPCC, 2022). The pri­ Calcined Clay Cement) (Avet and Scrivener, 2018). The latest research
mary cause of the significant CO2 emissions in the production of ordi­ has recognized the significant contribution of clinker replacement in
nary Portland cement (OPC) is the clinker manufacturing process, achieving carbon-neutral development of LC3 (Scrivener et al., 2019;
during which CaCO3 is subjected to calcination at temperatures Dhandapani et al., 2021). To ensure the binder's high integrity, the
exceeding 1450 ◦ C. Therefore, reducing the amount of clinker used in calcined clay must exhibit abundant pozzolanic performance. However,
cement production is crucial for achieving carbon neutrality. One because clays can vary in their mineralogical and thermal properties, it
strategy to reduce the economic and environmental impact of modern is important to investigate the priority of each clay in order to develop
cement is to replace the clinker with supplementary cementitious LC3 with optimal performance.

* Corresponding authors.
E-mail addresses: khuram_ae@uet.edu.pk (K. Rashid), j_dean21@yonsei.ac.kr (M. Ju).

https://doi.org/10.1016/j.clay.2023.107080
Received 11 March 2023; Received in revised form 17 June 2023; Accepted 17 July 2023
Available online 26 July 2023
0169-1317/© 2023 Elsevier B.V. All rights reserved.
N. Blouch et al. Applied Clay Science 243 (2023) 107080

LC3, a ternary blend of cement, has been successfully tested in both compressive strength (Ghorbel and Samet, 2013). High CaO concen­
laboratory and industrial trials (Scrivener et al., 2019; Zhou et al., tration reduces metakaolin formation during calcination (Alujas et al.,
2022). In LC3, the reactive alumina-silica phases from calcined clay react 2015). A high Al2O3/SiO2 ratio indicates higher pozzolanic activity
with hydration product of cement i.e., CH (calcium hydroxide) to form (Dondi et al., 2014). Other impurities like SO3, Na2O, K2O, and MgO
C-S-H (calcium silicate hydrate) and C-A-S-H (calcium alumino-silicate have minor direct impact but can affect porosity, setting time, and heat
hydrate). On the other hand, limestone reacts with reactive alumina of hydration of LC3 (Horkoss et al., 2011; Ma and Qian, 2018). However,
(Al2O3), whether from calcined clay or cement-aluminates, to produce their concentration in clays is typically low, resulting in negligible
low-density and microstructure-densifying carbo-aluminate phases impact on LC3 performance.
(Sharma et al., 2021). These synergistic reactions in LC3 have been The LC3 system uses clay as a major source of reactive Al2O3 and
proven to produce 28 d compressive strength equivalent to that of the SiO2, and therefore SiO2/Al2O3 ratio decides the composition and bal­
control OPC paste, even with a 50% replacement of clinker by using ance between hydration products, especially the relative formation of C-
high-grade kaolinitic clays (Nguyen et al., 2020a; Dhandapani et al., S-H and C-A-S-H. It has been observed that C-S-H formed in calcined
2021). Therefore, the use of clay as a major replacement for clinker is of clay-based binder is different from original OPC pastes with lesser Ca/Si
utmost importance, as its pozzolanic reactivity significantly influences and higher Si/Al ratios, and longer chain length (Rakhimova and
the performance of LC3 (Rakhimova and Rakhimov, 2020). Under­ Rakhimov, 2020). A study investigated the chemical composition of
standing the impacts of clay minerals, oxides, and thermal performance various types of clay for each clay mineral and stated the average Al2O3/
is essential for evaluating the performance of LC3. SiO2 ratios for kaolinite, illite, and montmorillonite to be 0.78, 0.57 and
Clay is commonly classified based on its mineralogical properties, 0.35, respectively (Sivrikaya et al., 2017). Kaolinites provide almost
such as kaolinite, montmorillonite, and illite. It has been found that equal amounts of SiO2 and Al2O3, whereas 2:1 clay is abundant in SiO2.
among all calcined clays, kaolinite is the most pozzolanic material It has been reported that bentonite (montmorillonite) favor silicate
(Fernandez et al., 2011; Msinjili et al., 2019; Schulze and Rickert, 2019). phase formation (C-S-H), while kaolinite triggers the formation of alu­
The calcination of kaolinite at an ideal temperature results in the for­ minohydrates and stratlingite (C2ASH8) (Tironi et al., 2013; Rakhimova
mation of metastable metakaolinite, which is a highly pozzolanic ma­ and Rakhimov, 2020). It has also been concluded that the reactive SiO2
terial. Halloysite is a promising mineral for making LC3. However, the contribution from pozzolans/SCMs should be at least 25% (ASTM C595,
performance of the resulting binder is sensitive to its replacement ratio 2010).
(Yu et al., 2023). The properties of LC3 produced using metakaolinite are This study investigates the potential of various low-grade and mixed
directly dependent on the kaolinite content of the raw clay. Higher clays and their suitability for developing LC3 cement. Initially, seven
kaolinite content results in better properties. Since high-grade or pure criteria are identified by establishing a database of the latest literature
kaolinite clay has value for other industries such as paper, ceramics, and on LC3. Nineteen clay samples are collected from nearby reserves to
pigment industries, it comes at a high cost (Alujas Diaz et al., 2022a). It existing cement plants in Pakistan and characterized using XRF, XRD
has been proven that clay with as little as 40% kaolinite content can and TGA-DTA analyses. These clays are classified based on their
produce cement pastes with equal compressive strength to pure OPC mineralogy, and a multi-criteria analysis is used to determine the pri­
paste (Avet and Scrivener, 2018). Therefore, there has been consistent ority of each clay for making LC3. The top 10 ranked clays are then
demand to use low or moderate quality clay for preparing the LC3. processed for calcination. Finally, LC3 is produced using ten types of
Therefore, there has been a consistent demand to use low or moderate clay, and its compressive strength is evaluated at 7 and 28 d. The
quality clay for preparing LC3. The global reserves of this type of clay are compressive strength is then compared and correlated with the char­
largely unexplored and uninvestigated, but they represent an important acterization of the calcined clays. Furthermore, multicriteria analysis is
opportunity with the potential to produce carbon-neutral cement. performed using the parameters of compressive strength and charac­
Clay reserves commonly contain impurities such as calcite, quartz, terization of calcined clay in the LC3 mixes. Consequently, the perfor­
feldspars, anatase, and albite (Rakhimova and Rakhimov, 2020; Alujas mance appraisal of LC3 using different types of clay will be discussed.
Diaz et al., 2022b; Alujas Diaz et al., 2022a). These impurities affect the The criteria selected for prioritization are verified using the heat map.
chemical and mineralogical composition of the clay and its pozzolanic
properties (Ghorbel and Samet, 2013). Inclusion of Fe2O3 up to 2.7% 2. Materials and methods
enhances clay's pozzolanic characteristics, but higher amounts deterio­
rate its performance. Fluxing oxides like Fe2O3, Na2O, and K2O can 2.1. Exploring raw clay
induce sintering, reducing the specific surface area and reactivity of the
clay (Schulze and Rickert, 2019; Rehman et al., 2020). Illite clay with To investigate the feasibility of using local clay in Pakistan for LC3
low hematite content exhibits similar pozzolanic reactivity to low or production, nineteen different types of clay (CL) were examined. These
moderate grade kaolinite clay (Msinjili et al., 2019). Clay reserves with clays were sourced from 11 different districts across Pakistan. Most of
calcite or dolomite should be used cautiously, as they reduce pozzolanic the samples were collected from clay quarries located near cement
reactivity when exposed to high temperatures (beyond 800 ◦ C). Calcite plants to ensure the practical feasibility of producing LC3 on an indus­
or dolomite impurities lead to the formation of vitreous forms that are trial scale in the future. Details regarding the sources of clay are pre­
rich in calcium, silicon, and aluminum, but have low specific surfaces sented in Fig. 1. The colors of all clays are shown in Fig. 2. The OPC and
(Ghorbel and Samet, 2013).The CaO content in clay can also reduce its limestone powder were obtained from Maple Leaf Cement Factory and
pozzolanic activity, but a range of 0.02 to 2.90 has no harmful impact Fatima Fertilizer Company, respectively. The analysis of these clay
(Díaz et al., 2018). When selecting clay for LC3 development, consid­ samples, along with cement and limestone, using X-ray fluorescence
eration must be given to the content of impurities like quartz or calcite- (XRF), is presented in Table 2.
rich soils, which often contribute a significant amount of CaO. Most LC3 The collected clay samples were of different particle sizes. Before
materials are based on kaolinite clay, with limited studies using mont­ calcination, the samples underwent a series of crushing and milling
morillonite, illite, or halloysite. Low-to-moderate-grade kaolinite and processes. First, a Denver black jaw crusher was used for primary
mixed clay are sustainable alternatives for the cement industry and are crushing. Samples with sizes <2 mm were processed through a disk mill
widely available (Msinjili et al., 2019; Alujas Diaz et al., 2022b). The and then a ball mill for 30 min. Samples with a particle size >2 mm were
kaolinite content, along with Al2O3, SiO2, Fe2O3, and CaO, as well as the first passed through a rod mill before undergoing ball milling. The raw
Al2O3/SiO2 ratio, affect LC3 properties. Al2O3 and SiO2 enhance early clays were then characterized. Details of the characterization process are
compressive strength and durability (Krishnan et al., 2019). Higher presented in Section 3.1.
Fe2O3 concentration in LC3 prolongs setting time and reduces early

2
N. Blouch et al. Applied Clay Science 243 (2023) 107080

durability of the LC3 by forming a calcium silicate hydrate (C-S-H) gel,

providing strength and stability to the cement paste and therefore its
higher content is preferable. Whereas Fe2O3 influences the mechanical
properties and color of the cement. However, its higher concentration
increases the setting time and decreases the early compressive strength
of LC3 (Ghorbel and Samet, 2013), and therefore its less amount is
preferable in LC3. Yet, a high CaO concentration within the clay matrix
results in the formation of calcium silicates and calcium aluminates,
which reduces the conversion of calcium carbonate to calcium oxide,
leading to a reduction in metakaolin formation during the calcination
process (Alujas et al., 2015), and its less amount is required in making
durable LC3. The Al2O3/SiO2 ratio is also a crucial factor in determining
the pozzolanic activity of the clay (Dondi et al., 2014). A higher value
indicates a higher pozzolanic activity due to an increased reaction with
calcium hydroxide Ca(OH)2 to form additional cementitious com­
pounds, enhancing the compressive strength and durability of the
resulting cement by forming C-A-S-H and C-S-H gels. Table 1 summa­
rizes the compressive strength of both types of LC3 blends, LC3–30 (OPC
replacement of 30 wt%) and LC3–50 (OPC replacement of 50 wt%),
along with the compressive strength of OPC. A comprehensive summary
Fig. 1. The locations of the source of clay explored in this study. was prepared in Table 1 and criteria that may influence the compressive
strength of LC3 are plotted in Fig. 4. The most common and important
2.2. Key process parameter is the type of mineral, and kaolinitic clay is used in the
ternary blend in almost all studies, and only in very few studies, mont­
The first step carried out in this work was the scientific literature morillonite and quartz, kaolinite with nacrite and halloysite have been
review and extraction of important parameters (Table 1) that influence used (Table 1). The second parameter is obviously the amount of
the performance of LC3. Few criteria were further extracted, and nine­ kaolinite content, and it varies from 0–100% as shown in Fig. 4. The
teen clays were characterized by conducting x-ray diffraction (XRD), x- variation in the chemical contents of different types of clay can be easily
ray fluorescence (XRF), and thermogravimetric analysis (TGA) for the visualized in Fig. 4. Whereas Fig. 5 shows the box plot and describes the
quantification of each criterion. Two types of multi criteria decision variation of maximum and minimum values of each parameter. Large
making (MCDM) techniques were applied with the intention of priori­ variations in the contents of kaolinite, Al2O3, SiO2, Fe2O3 and CaO can
tizing clay to be used for LC3. The top ten clays were selected and be observed. Therefore, these values must be part of the selection
calcined, and comparison of raw and calcined clays were made with criteria for selecting low-grade kaolinite or impure clay. Along with that,
respect to diffractograms and thermograms. The activation of clays was the ratio of Al2O3 to SiO2 (Al2O3/SiO2) has a significant impact on the
verified and used to fabricate a ternary blend. Then, LC3 were made of performance of LC3 and needs to be adopted for the selection of clay. The
top ranked 10 clays and compressive strength of each LC3 was measure early compressive strength (up to 7 d) may be affected by other oxides
and compared with each other and with OPC paste. The ranking of the present in clay, for instance, alkali oxides (Na2O and K2O) and SO3
compressive strength was also compared with the analytical priority of content, but these factors have little impact on the compressive strength
calcined clay, which was obtained by employing both MCDM tech­ at 28 d. Along with that, the variation in their contents is also less
niques. The key methodology of this work is illustrated in Fig. 3. compared to other parameters and therefore not used as the selection
criterion for adopting low-grade kaolinite or impure clay.
2.3. Criteria of clay for decision making analysis Based on the histogram analysis (Fig. 4), the selected criteria have
also been highlighted in Fig. 5. A study based on experimental and
The requirement for clay to be used as a binding ingredient is its high analytical methodology proposed that kaolinitic clay resources should
pozzolanic reactivity. Several researchers have analyzed the pozzolanic meet few criteria to be suitably used as SCMs: Al2O3/SiO2 > 0.3, Al2O3
reactivity of the three commonly available clay minerals: kaolinite, > 18.0%, and CaO < 3.0% (Díaz et al., 2018). Based on that, positive and
illite/micas, and smectite/montmorillonite (Fernandez et al., 2011; negative criteria have been investigated. Among them, contents of
Msinjili et al., 2019; Schulze and Rickert, 2019; Khalifa et al., 2020), and kaolinite, Al2O3, SiO2, and Fe2O3 have large variations and are posi­
concluded that these clays have exhibited pozzolanic reactivity at tively related to the compressive strength of LC3 due to the formation of
different calcination temperatures. The illitic and illitic-smectite clays C-A-S-H and C-S-H gels as well as mono- and hemicarboaluminates,
have been identified as potential SCMs in recent studies (Irassar et al., except Fe2O3, therefore, its least value is preferable and for other criteria
2019; Msinjili et al., 2019; Rossetti et al., 2021; Sposito et al., 2022). maximum value is preferable. Similarly, Al2O3/ SiO2 and CaO must be
And therefore, illitic and illitic-smectite clays may be used in the maximum and minimum, respectively, to develop LC3.
development of LC3. But undoubtedly, the kaolinite based LC3 have been
investigated extensively (Table 1). The variation in the kaolinite content 2.4. MCDM
and the varied chemical and mineralogical characteristics may in­
fluences the performance of LC3. The chemical composition of the clay Clays were selected based on their mineralogical composition and
plays a significant role in determining the reactivity and compressive the quantity of oxides present, as defined in Section 2.1. There are
strength of the LC3. The oxides of Al2O3, SiO2, Fe2O3, and CaO, as well as several factors that influence the performance of clay, and all of these
the Al2O3/SiO2 ratio, are critical elements that directly affect the criteria must be taken into account when making a decision on the
properties of LC3 (Díaz et al., 2018). For example, the presence of Al2O3 specific clay. Therefore, multi-criteria problems can be better solved
in the clay enhances the early compressive strength development of the using MCDM techniques. Two techniques have been adopted in this
cement by promoting the formation of calcium aluminate phases, more work and are explained in the following subsections.
precisely calcium aluminate silicate hydrate (C-A-S-H) (Krishnan et al.,
2019). SiO2, a primary component of cementitious materials, affects the 2.4.1. TOPSIS
setting and hardening of the blended cement and contributes to the The one of MCDM techniques employed in this work is TOPSIS

3
N. Blouch et al. Applied Clay Science 243 (2023) 107080

Fig. 2. The colors of various clay specimens used in this study.

(Technique for Order Preference by Similarity to Ideal Solution), which

A1 ⎡
is widely used in construction systems for material selection (Rashid c11 c12 ⋯ c1n
⎤
et al., 2017, 2020; Abed et al., 2022). TOPSIS is easy to apply and only [ ] A2 ⎢ c c22 ⋯ c2n ⎥
requires five steps presented in Eq. (1) through Eq. (5). In the first step, Decison Matrix [D] = cij m×n = ⎢
⎣ ⋮
21 ⎥ (1)
⋮ ⋮ ⋮ ⎦
the decision matrix “D” was established by quantifying all the criteria ⋮
cm1 cm2 ⋯ cmn
“Cmn” for each alternative “A” (Eq. (1)). The values from Eq. (1) were An
normalized in the second step (Eq. (2)). And in Eq. (2) “wj” is the weight
function and its value estimation has been explained in Section 2.4.3. Cij
Pij = wj ∑m (1 ≤ i ≤ m, 1 ≤ j ≤ n) (2)
Then, using Eqs. (3) and (4), a set of BEST and WORST ideal solutions
2
i=1 Cij

were made, respectively. In step four, a separation matrix was made for {〈 ( ) 〉 〈 ( ) 〉}
the BEST and WORST solutions by using Eq. (5). The relative closeness B+ = max Pij |i = 1,2, …, m |j ∈ Nb , min Pij |i = 1,2, …, m |j ∈ Nc
[ ]
coefficient (RCC), which was quantified in step five (Eq. (6)), shows the B+ = a+ + +
1 , a2 , …, an
priority of an alternative based on its highest value. The descending (3)
order of the RCC is the ranking or priority vector of all alternatives.

4
N. Blouch et al. Applied Clay Science 243 (2023) 107080

Table 1
Summary of kaolinite content, Al2O3/SiO2 ratio and the compressive strength of LC3 and OPC from published literatures.
Sr. No Major mineral Kaolinite (%) Al2O3/SiO2 CaO (%) Comp. Strength at 28 d (MPa) Refs.
3 3
LC –30 LC –50 OPC

1. Kaolinite 40.0 0.86 0.11 – 62.0 55.0 (Bernal et al., 2021)

2. 50.9 0.32 0.49 45.6 48.6
3. Kaolinite 78.8 0.86 0.12 – 49.5 50.5 (Nguyen et al., 2020a)
4. 95.0 0.84 0.02 – 50.0
5. Kaolinite 50.9 0.32 0.49 56.6 – 64.3 (Nguyen et al., 2020b)
6. 50.0 0.8. 2.20 60.7 –
7. Kaolinite 45.0 0.57 3.60 55.7 – 60.5 (Hanpongpun, 2019)
8. 50.0 0.27 2.07 63.5 50.5
9. Kaolinite 50.9 0.32 0.49 35.5 36.3 52.3 (Khan et al., 2018)
10. Kaolinite 60.0 0.62 0.20 58 59.0 63.0 (Wang et al., 2020)
11. Kaolinite 60.0 0.84 0.53 – 47.0 50.5 (Agarwal et al., 2020)
12. Kaolinite 75.0 0.57 0.40 58 – 62.0 (Kandagaddala et al., 2020)
13. Kaolinite 75.0 0.68 0.40 – 18.7 43.6 (Mishra et al., 2020)
14. Kaolinite 60.0 0.83 0.40 – 36.0 49.0 (Gunjal and Kondraivendhan, 2020)
15. 60.0 0.7 0.04 – 36.2 33.0
Kaolinite (Huang et al., 2020)
16. 60.0 0.7 0.04 – 36.9 53.9
17. Kaolinite 50.9 0.32 0.49 34 33.0 34.0 (Nguyen et al., 2018)
18. Kaolinite 56.5 0.72 0.06 41 42.0 – (Krishnan and Bishnoi, 2020)
19. 60.3 0.73 0.81 – 50.0 50.0
Kaolinite (Zunino and Scrivener, 2021)
20. 95.0 0.84 0.03 – 78.0 70.0
21. Kaolinite 60.0 0.93 0.02 56 54.0 58.0 (Antoni et al., 2012)
22. Kaolinite 70.0 0.84 0.46 57 54.0 45.0 (Yu et al., 2021)
23. Kaolinite 46.0 0.84 0.20 – 50.0 45.0 (Avet et al., 2019)
24. Kaolinite 57.0 0.61 0.40 – 48.1
49.0 (Huang et al., 2020)
25. Quartz 0.0 – 0.10 – 29.0
26. 65.3 0.31 0.10 – 34.0
Kaolinite 36.0 (Vizcaino Andres et al., 2015)
27. 80.6 0.7 – – 37.0
28. Kaolinite 70.0 0.72 0.06 – 41.0 40.5 (Akindahunsi et al., 2020)
29. Kaolinite 38.9 0.49 0.30 54.5 49.4
50.2 (Avet et al., 2016)
30. 95.0 0.84 – 77.2 67.3
31. Kaolinite 65.0 0.75 0.46 – 40.0 43.0 (Medepalli et al., 2016)
32. Kaolinite 50.9 0.32 0.49 – 49.3 52.3 (Nguyen and Castel, 2020)
33. Montmorillonite 0.0 0.27 1.55 70 54.0
69.0 (Ferreiro et al., 2017)
34. Kaolinite 92.0 0.75 0.05 76 62.0
35. Kaolinite 30.0 0.27 1.66 – 40.0 47.0 (Marangu, 2020)
36. 19.5 0.25 0.04 43 –
37. 20.8 0.25 1.71 41.5 –
Kaolinite 52.0 (Zolfagharnasab et al., 2021)
38. 28.7 0.32 2.70 47 –
39. 66.0 3.42 2.10 48 –
40. Kaolinite 20.0 0.31 0.07 – 41.0
41. Kaolinite, nacrite 57.8 0.5 0.28 – 39.0
48.0 (Almenares Reyes et al., 2018)
42. Kaolinite, nacrite, halloysite 46.1 0.43 2.94 – 49.0
43. Kaolinite, halloysite 81.1 0.8 0.05 – 51.0
44. 0.0 0 0.00 – 29.5
45. 95.0 0.84 0.00 – 68.5
46. 79.4 0.82 0.10 – 64.5
Kaolinite 50.0 (Maraghechi et al., 2018)
47. 50.3 0.72 1.30 – 56.5
48. 41.9 0.65 0.10 – 56.0
49. 17.0 0.26 0.60 – 38.0

{〈 ( ) 〉 〈 ( ) 〉}
W− = min Pij |i = 1,2, …, m |j ∈ Nb , max Pij |i = 1,2, …, m |j ∈ Nc which presents the priority vector of all alternatives.
[ ] ( )
W − = a−1 , a−2 , …, a−n
∑m a*j − cij
(4) Si = wj ( * ) (7)
j=1 aj − c−j
⎡√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ ⎤ ⎡√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ ⎤
√∑ √∑
√ n ( )2 √ n ( )2 ⎡ ( )⎤
S = ⎣√
+
Pij − aj ⎦ ; S = ⎣√
+ −
Pij − a−j ⎦ (5) a*j − cij
j=1 j=1
Ri = max⎣wj ( )⎦ (8)
j a*j − c−j
Si−
RCC = (6)
Si+ + Si− (Si − S* ) (Ri − R* )
Qi = v + (1 − v) − where v = 0.5 (9)
(S− − S* ) (R − R* )
2.4.2. VIKOR
VIKOR is another MCDM technique, which has been used by re­ 2.4.3. Entropy weight
searchers for the selection of sustainable material (Abed et al., 2022; The entropy method has been used to calculate the weight function
Zafar et al., 2022). The first step is again the establishment of the de­ used in MCDM techniques. In both techniques, the weight function “wj”
cision matrix (Eq. (1)), then its normalization (Eq. (2)), and making a set has been used (Eq. (2) and Eq. (7)). The calculation of weight functions
of BEST and WORST solutions (Eqs. (3) and (4)). Afterwards, Si, Ri, and depends on the experimental data available for each alternative. Step 1
Qi are quantified from Eqs. (7) to (9), respectively. The best alternative is is the development of the decision matrix (Eq. (1)). Its normalization
shown by having the lowest value or ascending order of the Qi value, expression is different (Eq. (10)) as compared to the expression used in

5
N. Blouch et al. Applied Clay Science 243 (2023) 107080

Fig. 3. The methodology to develop LC3 in this work.

Table 2
XRF analysis of ternary binder system: raw clay-OPC-limestone.
Sample ID Oxides (wt.%) Cl (%)

SiO2 Al2O3 Fe2O3 CaO MgO K2O Na2O SO3

CL-01 49.74 18 11.93 2.18 1.7 1.86 0.412 0.412 0.04

CL-02 40.6 7.59 3.53 22.75 1.83 1.35 0.71 0.02 0.02
CL-03 42.36 25.99 4.06 13.7 1.76 0.19 2.1 0.03 0.01
CL-04 49.3 13.58 5.79 8.15 2.71 0.01 0.94 0.03 –
CL-05 49.64 16.93 7.44 5.3 3.77 0.01 1.15 0.07 0.03
CL-06 41.45 29.23 2.32 9.6 0.68 2.55 0.65 0.03 –
CL-07 40.13 26.55 5.2 9.53 2.28 3.27 0.75 0.06 –
CL-08 43.27 39.63 1.06 0.11 0.26 – 0.18 0.05 0.02
CL-09 41.84 38.74 0.97 0.11 0.26 – 0.19 0.03 0.01
CL-10 41.23 37.55 1.08 0.12 0.26 – 0.91 0.03 0.01
CL-11 63.24 3.75 1.27 15.21 0.51 1.35 0.195 0.03 –
CL-12 44.67 12.59 2.93 19.75 1.83 1.35 0.71 0.02 0.04
CL-13 44.67 36.82 1.3 0.61 0.12 0.32 0.03 0.13 0.21
CL-14 65.09 17.59 3.43 0.40 1.14 2.32 0.56 0.15 0.36
CL-15 71.53 10.59 4.76 0.34 0.84 2.02 0.70 0.06 0.47
CL-16 43.79 15.12 13.08 4.21 3.18 1.16 1.94 0.09 2.40
CL-17 68.16 14.33 4.81 0.32 1.03 2.07 0.48 0.00 0.28
CL-18 4.74 1.16 0.77 49.6 1.12 0.73 2.28 1.41 1.01
CL-19 1.30 0.52 0.43 53.04 1.29 0.37 0.23 0.06 1.01
OPC 20.16 4.86 3.01 64.11 1.86 64.11 0.28 2.87 0.01
Limestone 0.10 0.03 0.023 54.61 0.21 – – – –

both techniques (Eq. (2)). Then the weight function is calculated from ⃒ ⃒
⃒1 − ej ⃒
Eq. (11) and the unknown values are calculated from Eq. (12). The sum wj = ∑ (11)
m ⃒ ⃒ (1 ≤ j ≤ n)
of all weight functions must be equal to 1.0. ⃒1 − ej ⃒
i=1
xij
Pij = ∑
m (1 ≤ i ≤ m, 1 ≤ j ≤ n) (10)
xij ∑
m
( ) 1
i=1 ej = − k Pij lnPij (1 ≤ j ≤ n), k = (12)
i=1
lnm

6
N. Blouch et al. Applied Clay Science 243 (2023) 107080

Table 3
Decision matrix for the shortlisting of raw clay.
Clay Major mineral Kaolinite (%) Al2O3 (%) SiO2 (%) Fe2O3 (%) Al2O3/SiO2 (%) CaO (%)

CL-01 8.5 32.0 18.0 49.7 11.9 0.4 2.2

CL-02 8.5 15.7 7.6 40.6 3.5 0.2 22.8
CL-03 7.0 10.8 26.0 42.4 4.1 0.6 13.7
CL-04 7.0 10.7 13.6 49.3 5.8 0.3 8.2
CL-05 7.0 20.0 16.9 49.6 7.4 0.3 5.3
CL-06 8.5 21.0 29.2 41.5 2.3 0.7 9.6
CL-07 7.0 11.5 26.6 40.1 5.2 0.7 9.5
CL-08 10.0 53.0 39.6 43.3 1.1 0.9 0.1
CL-09 10.0 40.0 38.7 41.8 1.0 0.9 0.1
CL-10 10.0 35.0 37.6 41.2 1.1 0.9 0.1
CL-11 5.5 0.0 3.8 63.2 1.3 0.1 15.2
CL-12 5.5 0.0 12.6 44.7 2.9 0.3 19.8
CL-13 10.0 55.1 36.8 44.7 1.3 0.8 0.6
CL-14 7.0 21.0 17.6 65.1 3.4 0.3 0.4
CL-15 7.0 9.8 10.6 71.5 4.8 0.2 0.3
CL-16 7.0 20.0 15.1 43.8 13.1 0.4 4.2
CL-17 8.5 6.5 14.3 68.2 4.8 0.2 0.3
CL-18 4.0 0.0 1.2 4.7 0.8 0.2 49.6
CL-19 4.0 0.0 0.5 1.3 0.4 0.4 53.0

Table 4
Entropy and Equal weight values for each criterion.
Weight Criteria Sum

Mineral Kaolinite SiO2 Al2O3 Fe2O3 Al2O3/SiO2 CaO

Equal 0.14 0.14 0.14 0.14 0.14 0.14 0.14 1.00

Entropy 0.02 0.20 0.05 0.11 0.16 0.09 0.37 1.00

3. Results and discussion

Table 5
Priority of nineteen raw clay by both MCDM techniques.
The chemical composition and mineralogy of raw clay are presented
MCDM TOPSIS VIKOR Average priority initially. Based on the characterized material properties, MCDM analysis
Weight Equal Entropy Equal Entropy was conducted to determine the priority of the raw clay. Then, 10 sig­
CL-01 12 9 12 8 10 nificant clay samples were selected for calcination to be used in the
CL-02 15 17 13 15 15 implementation of LC3. All details and discussions will be presented in
CL-03 8 14 9 11 11 the following subsections.
CL-04 13 12 10 13 12
CL-05 11 8 6 7 8
CL-06 5 6 5 5 5 3.1. Performance evaluation of raw clay for LC3
CL-07 7 11 8 9 9
CL-08 1 1 1 1 1 3.1.1. Characterization of raw clay
CL-09 3 3 3 3 3 Raw clay samples were analyzed for chemical composition, miner­
CL-10 4 4 4 4 4
CL-11 14 15 16 16 15
alogy, and thermal properties. PANalytical Smart Zetium XRF spec­
CL-12 17 16 17 17 17 trometer was used to assess the oxide composition of clay, limestone
CL-13 2 2 2 2 2 powder, and OPC. Mineralogical characterization of clay was performed
CL-14 6 5 7 6 6 out through XRD using a BTX III Benchtop XRD Analyzer. The XRD
CL-15 10 7 14 10 10
analysis was performed with Cu-kα radiation using 30 mA current and
CL-16 16 13 15 14 15
CL-17 9 10 11 12 11 30 kV voltage, respectively. XRD patterns were collected in the 5–45o 2θ
CL-18 19 18 19 18 19 range with a step time of 1.58 s and step size of 0.025o 2θ. Thermog­
CL-19 18 19 18 19 19 ravimetric analysis was performed to determine the dehydroxylation
behavior and kaolinite content of raw clay using a LINSEIS STA PT 1600
(TGA-DTA) thermogravimetric analyzer. Powder samples of 25 mg were
Table 6 heated at a rate of 10 ◦ C/min from 30–900 ◦ C under nitrogen environ­
Mix proportions for OPC and LC3 cement paste. ment. The weight loss during kaolinite dehydroxylation (wt%kaol-OH)
Sample ID Portland Calcined Limestone Water/binder1
was calculated using Eq. (13), where Mkaolinite (258.16 g/mol) and
cement clay ratio Mwater (18.02 g/mol) are the molecular weights of kaolinite and water,
respectively. The weight loss during the dehydroxylation of kaolinite at
OPC 1.0 0.0 0.0 0.4
LC3 0.7 0.2 0.1 0.4 400–600 ◦ C, expressed as wt%kaol-OH (Alujas et al., 2015; Avet and
1
Scrivener, 2018).
Binder denotes summation of ordinary Portland cement, limestone, and
calcined clay. wt.%kaol− OH × Mkaolinite
wt.%kaolinite = (13)
2Mwater

3.1.2. Mineralogical analysis

XRD analysis was used to determine the mineralogical composition
of the raw clay. The XRD patterns of clay are presented in Fig. 6.

7
N. Blouch et al. Applied Clay Science 243 (2023) 107080

Table 7
The properties of top-ranked clay for calcination and compressive strength of developed LC3.
Clay Major mineral Kaolin (%) SiO2 (%) Al2O3 (%) Fe2O3 (%) Al2O3/SiO2 CaO (%) Comp. Strength (MPa)

7d 28 d

CL-01 8.5 32.0 49.7 18.0 11.9 0.4 2.2 28.7 41.1
CL-03 7.0 10.8 42.4 26.0 4.1 0.6 13.7 25.6 39.9
CL-05 7.0 20.0 49.6 16.9 7.4 0.3 5.3 26.1 40.3
CL-06 8.5 21.0 41.5 29.2 2.3 0.7 9.6 19.9 38.8
CL-07 7.0 11.5 40.1 26.6 5.2 0.7 9.5 23.9 40.0
CL-08 10.0 53.0 43.3 39.6 1.1 0.9 0.1 34.2 46.2
CL-09 10.0 40.0 41.8 38.7 1.0 0.9 0.1 24.6 42.1
CL-10 10.0 35.0 41.2 37.6 1.1 0.9 0.1 29.8 38.5
CL-13 10.0 55.1 55.1 36.8 1.3 0.8 0.6 32.4 48.6
CL-16 7.0 21.0 65.1 17.6 3.4 0.3 0.4 21.6 37.7

Table 8
Entropy and equal weight values for each criterion of calcined clay.
Weight Criteria Sum

Mineral Kaolinite SiO2 Al2O3 Fe2O3 Al2O3/SiO2 CaO

Equal 0.14 0.14 0.14 0.14 0.14 0.14 0.14 1.00

Entropy 0.01 0.10 0.01 0.04 0.26 0.06 0.53 1.00

samples exhibit a sharp peak at around 6◦ 2θ that confirms the presence

Table 9
of montmorillonite (Fernandez et al., 2011). Kaolinite and quartz were
Comparison of priority of compressive strength with calcined clay potential for
also present along with traces of illite and hematite (Hollanders et al.,
the development of LC3.
2016; Patarachao et al., 2019). Calcite and quartz were the minerals
Clay TOPSIS VIKOR Average Comp.
reflected in CL-11, 12, 18, and 19 (Ranjan et al., 2019; Tavares et al.,
priority of strength
Equal Entropy Equal Entropy
clay priority 2020). These clays were identified as inert clay.

CL-
01
7 6 7 7 6.8 4 3.1.3. Chemical composition
CL- The chemical composition of all clay is presented in Fig. 7. It was
10 10 10 10 10.0 7
03 extracted that clay from different regions predominantly consisted of
CL-
8 7 9 6 7.5 5 Al2O3 (3%–39%) and SiO2 (4%–64%). The Fe2O3 content of clay varied
05
from 0.9%–10.2%. The CaO content ranged between 0.11% - 50% while
CL-
06
6 8 5 8 6.8 8 the SO₃ content of all clay was below 4%.
CL- Following the concept from previous studies (Sivrikaya et al., 2017;
9 9 8 9 8.8 6
07 Khalifa et al., 2020), the contents of SiO2, Al2O3, and CaO (Fig. 7(a)), as
CL- well as SiO2, Al2O3, and other oxides (Fig. 7(b)) are normalized to the
2 1 2 1 1.5 2
08
sum of these components to identify the type of clay minerals under
CL-
3 2 3 3 2.8 3 study, and presented in Fig. 7. The separation of different clay minerals
09
CL- allows their potential shortlisting for producing LC3. Ternary diagrams
4 4 4 4 4.0 9
10 based on chemical composition can help to identify inert clay from
CL-
1 3 1 2 1.8 1 kaolinitic/illitic clay. But are ineffective, to some extent, in differenti­
13
CL- ating low-grade kaolinitic or mixed clay as the chemical compositions of
5 5 6 5 5.3 10 kaolinitic and illitic clay can be similar sometime as indicated by the
16
overlapped region in Fig. 7(b). However, inert clay can be easily iden­
tified by their predominantly high SiO2 or CaO content and extremely
Different mineralogical compositions of clay have been observed from low Al2O3/SiO2 ratios. Indeed, the chemical composition is helpful for
different sources. Three different clay minerals were found, namely preliminary classification categorization.
kaolinite, illite, and montmorillonite; while quartz, calcite, and hematite It can be observed from the ternary diagram based on SiO2 - Al2O3-
were found as impurities and/or traces additionally. CaO in Fig. 7(a) four clay were falling into 1:1 clay category, while
Clays were sorted and categorized based on major clay minerals as almost four clays were falling in 2:1 clay category out of the nineteen
kaolinitic (kaolinite-rich or kaolinite as major mineral), clays. Fig. 7(b) placed 4 clays in the kaolinite zone, 3 in illite zone, and 5
montmorillonite-rich clay and illitic (illite-rich), while a fourth category in smectite (montmorillonite-rich in this study) zone, which better agree
was termed as inert-clay had mostly quartz and calcite as major minerals with the XRD results. This shows that this chemical composition based
and either no clay minerals or there were traces of them. The XRD of clay graphical description can be a good source for identifying suitable clays
samples, CL-08, 09, 10, and 13, showed strong reflections of kaolinite for SCM used especially in LC3.
minerals, which were the first group termed as kaolinitic clay (Fig. 6).
One prominent peak of these clay at 11.3◦ 2θ shows kaolinite (Hinckley, 3.1.4. Thermogravimetric analysis
1962; Meroufel et al., 2013). The XRD curves of CL-03, 04, 05, 07, 14, Ground and sieved samples of raw clay were subjected to TGA in the
15, and 16 showed that illite was the dominant mineral in these clays, temperature range of 30–900 ◦ C to examine the dehydroxylation
with some reflections of quartz, kaolinite, calcite, and hematite as im­ behavior. The derivative thermograms are shown in Fig. 8. It shows that
purities. These clays were classified as illitic clay. The XRD curves of clay below 300 ◦ C, kaolinitic-clay (CL-08, 09, 10, and 13) underwent negli­
CL-01, 02, 06, and 17 have strong peaks of montmorillonite. These gible mass loss, whereas illitic-clay (CL-03, 04, 05, 07, 14, 15, and 16)

8
N. Blouch et al. Applied Clay Science 243 (2023) 107080

Fig. 4. Histograms presenting average composition of clay from the literatures (Note: all values were extracted from Table 1).

them less stable at temperatures below 300 ◦ C (Fernandez et al., 2011),

On the other hand, montmorillonite clays have a more complex layered
structure that can undergo significant structural changes at these tem­
peratures, resulting in the observed mass loss (Fernandez et al., 2011;
Garg and Skibsted, 2016).
The second peak in the derivative weight loss curves at mid-
temperature level (300–600 ◦ C) is attributed to the removal of hy­
droxyls from the clay structure. The specific temperature at which this
transformation occurs depends on the type of clay (Bernal et al., 2017).
Clays (CL-08, 09, 10, 13) that underwent dehydroxylation between
400 ◦ C – 650 ◦ C are considered to be kaolinitic clays, according to
studies (Ortega et al., 2010; Gasparini et al., 2013). The presence of
kaolinite in these clays was also confirmed by XRD analysis (Section
3.1.2). Clay minerals that exhibit dehydroxylation at temperatures
exceeding 600 ◦ C are of the 2:1 type, such as illite or montmorillonite.
Based on their dehydroxylation behavior, Fernandez et al. (2011) clas­
Fig. 5. Selected criteria for prioritizing clay from dataset. sified CL-01, 02, 03, 04, 05, 06, 07, 14, 15, 16, and 17 as 2:1 type clay. In
this category, clay samples CL-03, 04, 05, 07, 14, 15, and 16 exhibited
showed small mass loss. The montmorillonite clay (CL-01, 02, 06, and complete dehydroxylation up to 700 ◦ C, indicating that they are likely
17) exhibited a significant mass loss in this range. The mass loss in composed of illite clay (Irassar et al., 2019; Msinjili et al., 2019). The
different clay below 300 ◦ C occurs due to the dehydration of absorbed XRD patterns (Section 3.1.2) also confirm the predominant presence of
water molecules between the layers of clay (Bernal et al., 2017). illite peaks in these clay samples. The montmorillonite clay samples (CL-
Kaolinitic clays are composed of layered structures and are relatively 01, 02, 06, 17) underwent dehydroxylation up to 800 ◦ C. Literature
stable at low temperatures (Arab et al., 2015). Illitic-clays have a suggests that dehydroxylation of montmorillonite clay primarily occurs
different mineral composition and usually contain impurities making within the temperature range of 600–800 ◦ C (Garg and Skibsted, 2016;

9
N. Blouch et al. Applied Clay Science 243 (2023) 107080

Fig. 6. X-ray diffractograms of raw clay. I: Illite, C: Calcite, K: Kaolinite, M: Montmorillonite, Q: Quartz, H: Hematite.

Fig. 7. Ternary diagram of the clays studied (a) CaO-SiO₂- Al2O3 (b) Al2O3-SiO2-other oxides.

Alujas Diaz et al., 2022b). Clay samples CL-11, CL-12, CL-18, and CL-19 nature and in consideration of TGA results. A higher calcination tem­
exhibit inert behavior, with only a minor dehydroxylation process perature may initiate recrystallization in clay minerals.
occurring until 700 ◦ C. These clays demonstrated complete mass loss at These weight loss values were then used to quantify the kaolinite
approximately 850 ◦ C. In this study, a calcination temperature of 850 ◦ C content in each sample. The overlapping dehydroxylation zones of
was selected for ten chosen clays (Section 3.2.1) due to their mixed clay kaolinite and illite can make it difficult to accurately determine the

10
N. Blouch et al. Applied Clay Science 243 (2023) 107080

Fig. 8. Thermograms of the raw clays used in this study.

kaolinite content by TGA. Clay with kaolinitic content up to 40% - 60%

are the established valuable deposit of clay for the production of LC3
(Scrivener et al., 2018). Also, illitic clay with low kaolinite content or
vice versa are also suitable for LC3 production. It is worth mentioning
that 1:1 or 2:1 type of clay minerals are considered reactive if these
minerals (i) have a relatively high content of SiO2 and/or Al2O3 and (ii)
exhibit a significant weight loss due to elimination of OH group during
calcination process (Díaz et al., 2018).

3.1.5. Prioritizing raw clay for calcination

The selection of appropriate clay for LC3 production is a complex
process that demands extensive testing. Multiple criteria, which may
sometimes conflict, must be considered when making a decision. In this
study, two MCDM techniques (TOPSIS and VIKOR), were employed to
prioritize clay. After conducting a comprehensive literature review
(Section 1) and analyzing data from previous studies (Table 1), we
investigated seven criteria: (i) mineral type, (ii) kaolinite content, (iii)
silica content (SiO2), (iv) alumina content (Al2O3), (v) Fe2O3 content,
(vi) Al2O3/SiO2 ratio, and (vii) CaO content. This is summarized in
Fig. 5. The type of minerals and the amount of kaolin content are the Fig. 9. Priority of clay for further treatment determined by TOPSIS and
primary variables that influence the properties of LC3 cement. The type VIKOR techniques.
of major mineral, which is an immeasurable criterion, is converted into a
measurable criterion by assigning a ranking of 1 to 10 to each mineral. development of LC3. It can be observed that CL-08 is the best clay as
Kaolinite received the highest score of 10 among the minerals used in depicted by both techniques for both weights, equal (EQ) or entropy
LC3, followed by montmorillonite with 8.5, illite with 7, non-reactive (EN). The trend is almost followed for both MCDM techniques and the
quartz with 5, and calcite with 4. All other criteria have already been same priority was achieved for each clay. The ten clays were selected
measured and incorporated into the decision matrix using Eq. (1). It is and therefore the limit value of priority 10 has also been plotted in
the initial step in analyzing any MCDM technique, whether it be TOPSIS Fig. 9. And all points exceeding that limit are selected for calcination and
or VIKOR, and is displayed in Table 3. finally for the development of LC3.
TOPSIS and VIKOR were applied to the decision matrix (Table 3) by
considering both equal and entropy weights for each criterion. The en­ 3.2. Performance evaluation of calcined clay for LC3
tropy weight was calculated by following the Eqs. (10)–(12) and their
corresponding weights are described in Table 4. The sum of entropy 3.2.1. Calcination of raw clay
calculated with equal weight is equal to 1.00. The detailed procedures The prioritized raw clays were calcined in a laboratory furnace. One
for these techniques are explained in Section 2.4.1 and Section 2.4.2, kilogram of each type of clay was poured into an alumina crucible and
respectively. placed in the furnace at 850 ◦ C for a retention time of 1 h. The clay was
The priority of each clay has been made and detail of ranking is rapidly cooled to room temperature after calcination by removing the
described in Table 5 and plotted in Fig. 9. The outermost intersecting crucible from the hot furnace and placing it on a metal plate to cool. The
line across the concentric circles in Fig. 9 presents the best clay for the calcination temperature is determined based on the dehydroxylation

11
N. Blouch et al. Applied Clay Science 243 (2023) 107080

behavior of the clay samples, as determined by TGA, and taking into clays (CL-08 and 13) were also ranked at the top when analyzed using
account the mixed nature of the clay minerals present. XRD and TGA MCDM analysis in Section 3.1.5. The trend in compressive strength
were also performed on the calcined samples to observe the effects of development at 28 d is consistent with the trend observed at 7 d.
thermal treatment on the structure of clay minerals and the resulting Notably, LC3 with CL-08 and CL-13 exhibit the highest compressive
mineralogical changes. The characterization of calcined clays has been strength at 28 d among all calcined clays. Moreover, the compressive
explained in detail in Section 3.2.3. strength performance of LC3 with CL-13 has been greatly enhanced and
it has even outperformed the compressive strength of OPC. The higher
3.2.2. Development and assessment of compressive strength of LC3 reactivity of these two calcined clays is strongly correlated with their
higher kaolinite content. The LC3 with low-grade kaolinite clays (CL-09
3.2.2.1. Fabrication of LC3. The ten clay samples that were short-listed and CL-10) exhibit less compressive strength compared to both OPC and
based on their characterization were used to make LC3 pastes. Addi­ LC3 with high-grade kaolinite clays, as shown in Fig. 10. However, the
tionally, a control paste using OPC was prepared for comparison pur­ compressive strength of LC3 based on this material is greater than that of
poses. The water-to-binder (w/b) ratio of the test specimens was illite and montmorillonite.
maintained at 0.4. The LC3 paste was prepared by mixing 20% by weight The compressive strength gains of illitic-based LC3 specimens at 7
of calcined clay, 10% by weight of limestone, and 70% by weight of d appear to be largely independent of the kaolinite content and/or
OPC, while maintaining a weight ratio of 2:1 between calcined clay and chemical composition. However, the compressive strength gain after 7
limestone (Table 6). This ratio is based on the stoichiometry of the re­ d is quite significant (56% - 76%) and consistent when compared to the
action between metakaolin and calcium carbonate, which forms mono kaolinite based LC3. This result suggests that clay rich in illite is rela­
carboaluminates. This ratio has been utilized by several authors (Antoni tively less reactive than clay rich in kaolinite, which is consistent with
et al., 2012; Avet and Scrivener, 2018; Hameed et al., 2020). The mixing previous literature (He et al., 1995; Scrivener et al., 2019). The calcined
process was as follows: (a) Dry mixing clay, limestone powder, and OPC montmorillonite-based LC3 specimens (CL-01, C-06) exhibit compres­
for 2 min using a mechanical mixer; (b) Water was added; (c) The sive strength comparable to that of illite-based LC3. The LC3 with CL-01
mixture was stirred for an additional 5 min to create a homogeneous specimen has higher compressive strength than LC3 with CL-06,
paste. Specimens of 50-mm cubes were cast in steel molds and then although both are montmorillonite clays in LC3. The possible reason
wrapped with polyethylene sheets to prevent moisture evaporation. for this variation is the difference in the amount of kaolinite content. The
After 24 h, the fabricated cubes were removed from the mold and cured CL-01 has almost 34% kaolinite content, whereas CL-06 has only 23%.
in water. Afterwards, the compressive strength test was conducted at 7 This observation again verifies that a higher content of kaolinite
and 28 d by following ASTM C109 (ASTM C618, 2023). promotes early compressive strength development, while other clay
minerals are relatively slow to react and promote late compressive
3.2.2.2. Discussion on compressive strength of LC3. The compressive strength development (Schulze and Rickert, 2019). According to the
strength of blended cement paste (LC3) is measured at 7 and 28 d and OPC Type I (BS EN 197–1:2011, 2011), low-grade kaolinite with a
compared to both, each other and with the compressive strength of OPC composition of 40% - 55% results in higher compressive strength gain
paste. The compressive strength of ten LC3 mixes was categorized ac­ than the standard 42.5 MPa compressive strength of OPC Type I. Even
cording to the major clay mineral present (kaolinite, illite, and mont­ clays with lower grades, at 38–41 MPa, can achieve up to 96.5% of the
morillonite), and the results are presented in Fig. 10 for comparison. In compressive strength of OPC Type I.
each bar of Fig. 10, the difference in compressive strength between OPC It can be observed from Fig. 10 that, in all LC3 mixes, the 7
and LC3 has been indicated. It was observed that LC3 with calcined d compressive strength is lower than that of OPC compressive strength.
kaolinitic clays (CL-08, 09, 10, and 13) exhibited greater compressive However, at 28 d, the compressive strength of LC3 with kaolinitic clays
strength development. Among them, LC3 with CL-08 and CL-13 exhibit (CL-08 and CL-13) is almost similar to that of the OPC specimens. The
the highest compressive strength among all calcined clays. These two compressive strength difference between LC3 and OPC at 7 d can be

50 60
3%
-2%
50
40 -10%
-13%
-15% -14% -15%
Compressiev Strength (MPa)

-18% -18%
-20%
Kaolinite Content (%)

-11%
40
30 -16%
-22%
-25%
-32% 30
-33%
-36% -38%
20 -44%
-48%
20
LC³-7 d
LC³-28 d
10 OPC-7d (38.4 MPa) 10
OPC-28d (47.0 MPa)
Kaolinite Content (%)
0 0
CL-08 CL-09 CL-10 CL-13 CL-03 CL-05 CL-07 CL-16 CL-01 CL-06
Kaolinite Clay Illite Clay Montmorillonite
Clay
Type of Clay for LC3

Fig. 10. Compressive strength of all types of LC3 specimens and OPC.

12
N. Blouch et al. Applied Clay Science 243 (2023) 107080

attributed to two factors. Firstly, the pozzolanic and synergistic re­ signifies the amorphous state (He et al., 1995; Elimbi et al., 2011). Peaks
actions of clay and limestone blends depend on the reaction products of of quartz were observed in the calcined specimens showing the unal­
OPC, such as portlandite and ettringite. Secondly, the composition of tered behavior of this mineral in this temperature range. The XRD pat­
mixed clay, including the relative proportions of clay minerals and other terns of selected calcined illitic clay (CL-03, 05, 07, and 16) showed
impurities, can also influence the compressive strength development small peaks of illite and muscovite. While calcined montmorillonite clay
during hydration reactions. Consequently, this indicates that the amount (CL-01 and 06) showed no peaks of montmorillonite with minor peaks of
of kaolinite present plays a significant role in the development of mullite and albite. It has been reported in the literature that 2:1 clay can
compressive strength. In order to gain a better understanding of the retain part of their crystalline structure even after complete dihydrox­
relationship between the chemical and mineralogical composition of ylation (Alujas Diaz et al., 2022b). These clay show strong quartz peaks,
clay and the compressive strength development of blended cements, a indicating that the minerals have not been transformed in this temper­
correlation matrix was developed and analyzed in Section 3.2.4, using ature range (Fernández-González et al., 2013; Maier et al., 2021).
various input parameters.
3.2.3.2. Thermogravimetric analysis of calcined clay. The DTG curves for
3.2.3. Characterization of selected calcined clay raw (R) and calcined (C) clay are displayed in Fig. 12. The effectiveness
Ten top-ranked clay samples, which may be kaolinitic, illitic, or of the treatment strategy was evaluated by comparing the difference in
montmorillonitic (or a combination of these), were selected based on mass loss between the DTG curves of the raw and calcined clay. The DTG
their Al2O3 content of over 15%, SO3 content of approximately 3%, curves of the calcined clay were almost straight, indicating that the
varying CaO content, and an Al2O3/SiO2 ratio >0.3 (see Table 7). selected clay samples underwent almost complete dehydroxylation
Moreover, clays with <10% kaolinite content were not selected, even if during the studied calcination cycle. However, the materials studied
they ranked in the top ten according to TOPSIS and VIKOR. After were impure because they contained varying amounts of associated
excluding such clays, the next two closely ranked clays were selected, minerals, such as quartz and other clay structures, as indicated by the
calcined, and their characterization is explained in the following XRD results (Fig. 11). The weight loss curve may be influenced by im­
subsections. purities, therefore, the DTG curves of these clay samples are not exactly
straight lines. The DTG results agree with the XRD results.
3.2.3.1. Mineralogical analysis of calcined clay. The calcined clay sam­
ples were also subjected to XRD analysis to evaluate the efficacy of the 3.2.3.3. Validation of correlation between selected criteria and compressive
calcination process. Fig. 11 displays the XRD patterns of the raw clay strength. To establish a stronger correlation between the compressive
and calcined clay (C) samples that were selected. In the case of kaolinitic strength performance and the mineralogical and chemical composition
clay samples (CL-08, 09, 10, and 13), a significant decrease in crystal­ of clay, a correlation matrix was developed using six input parameters
linity was observed after calcination at a temperature of 850 ◦ C. The and two output parameters, namely the 7 d and 28 d compressive
peaks corresponding to kaolinite disappeared due to dehydroxylation, strength. Fig. 13 shows the correlation matrix, indicating that the 7
indicating that the kaolinite had been completely transformed into d and 28 d compressive strengths are highly dependent on the kaolinite
amorphous metakaolin. The broad hump between 18◦ and 32◦ 2θ

Fig. 11. X-ray diffractograms of raw and calcined clay; I: Illite, C: Calcite, K: Kaolinite, M: Montmorillonite, Q: Quartz, H: Hematite.

13
N. Blouch et al. Applied Clay Science 243 (2023) 107080

Fig. 12. Derivative thermograms of selected raw and calcined clay.

Fig. 13. Heat map of the correlation matrix along with the criteria.

14
N. Blouch et al. Applied Clay Science 243 (2023) 107080

content, with correlation coefficients of 0.77 and 0.82, respectively. correspondence was observed between the compressive strength and
These coefficients are the highest among all the coefficients. The amount priority of clay. More specifically, CL-08 and CL-13 had the highest
of CaO is a significant factor that influences compressive strength, as priority and observed the highest compressive strength among all
indicated by its negative correlation coefficients with maximum values specimens. A detailed comparison with respect to both MCDM tech­
of − 0.61 and − 0.47 for 7 and 28 d, respectively. The presence of CaO niques and by opting equal and entropy weight function has been made
delays reactivity, as explained in Section 3.2.2.2. Furthermore, it has in Table 9. The average of four values is considered as the priority of that
been observed that the correlation coefficient is higher for 7 clay and compared with the rank of the compressive strength. The
d compressive strength (− 0.61) compared to 28 d compressive strength agreement and disagreement can be easily visualized through the
(− 0.43). Al2O3 and Al2O3/SiO2 are other factors that positively influ­ overlap and contrast. It can be concluded that the proposed framework
ence the compressive strength after the kaolinite content. Although SiO2 of analytical techniques closely correlates clay properties with their
has a positive effect on compressive strength, its influence is marginal. performance in LC3 blends, across a wide range of clay deposits. This
Furthermore, its coefficient of correlation is lower compared to other framework can serve as a guideline for selecting clay that is suitable as
coefficients for both 7 d and 28 d compressive strength outputs. The SCMs for producing low-emission and environmentally friendly
correlation matrix was used to determine the influence of compressive cements.
strength between 7 and 28 d. The coefficient value obtained was 0.76,
indicating that the compressive strength at 28 d will be significantly 4. Conclusions
higher than that at 7 d. It is important to note that in the correlation
matrix (Fig. 13), CaO and Fe2O3 have negative correlation coefficients, This study identified the criteria and investigated the feasibility of
while the remaining four parameters have positive correlation co­ utilizing locally available low-grade kaolinitic and mixed clay to
efficients. The results of the correlation matrix confirm the BEST and formulate calcined clay limestone cement (LC3). After characterizing the
WORST solutions that were adopted for prioritizing clay. Therefore, the raw clays and using MCDM techniques, the top ten ranked clays were
priority of each clay is determined and compared with its compressive calcined to configure LC3, and their performance was appraised. The
strength value, as explained in Section 3.2.4. following conclusions were drawn from extensive experimentation and
analytical work.
3.2.4. Decision making for LC3 from the relationship between
characterization and compressive strength 1. Nineteen raw clays were analyzed and based on physico-chemical
The characterization of the calcined clay is presented in Section and thermal analyses, they were classified into: (i) kaolinite clay,
3.2.3, while the corresponding compressive strength of LC3 is explained (ii) illite clay, (iii) montmorillonite clay and (iv) quartz. It was found
in Section 3.2.2.2. The compressive strength and correlation matrix that impure clay consisting mainly of kaolinite and montmorillonite
verify that the top ranked clay must have highest compressive strength. was completely dehydroxylated after being heated at 800 ◦ C for an
For each clay the compressive strength values are described in Table 7, hour. On the other hand, the illite-containing clay required a higher
and it can be easily ranked with respect to the compressive strength. But temperature of 850 ◦ C to achieve complete dehydroxylation.
in the case of clay, many factors (SiO2, Al2O3, Fe2O3, Al2O3/SiO2 and 2. The dehydroxylation, amorphous, and crystalline phases of
CaO) influence the performance as discussed in detail in Section 2.3. kaolinite, illite, and montmorillonite clays were confirmed through
Therefore, both types of MCDM techniques were applied and a decision XRD and TGA-DTA analyses. Mineral type, kaolinite content, SiO2,
matrix is presented in Table 7. However, the compressive strength Al2O3 and Al2O3/SiO2 Fe2O3 and CaO were identified as criteria for
values were eliminated, and the remaining seven columns are the prioritizing clays using multicriteria analysis.
criteria for ten alternatives (ten clays selected for calcination). TOPSIS 3. It was found that kaolinite content, SiO2, Al2O3 and Al2O3/SiO2 have
and VIKOR techniques were applied by considering equal weight and positive value of coefficient of correlation with the compressive
entropy weights for each criterion (Table 8) and the ranking of clay has strength and high value of aforementioned parameters is preferable
been plotted in Fig. 14. The priority of LC3 based on compressive for developing LC3, whereas Fe2O3 and CaO have negative value of
strength was also plotted for the comparison purpose, close

Fig. 14. Priority comparison of compressive strength with the potential of the calcined clay for the LC3.

15
N. Blouch et al. Applied Clay Science 243 (2023) 107080

coefficient of correlation and therefore their less value is preferable Alujas, A., Fernández, R., Quintana, R., Scrivener, K.L., Martirena, F., 2015. Pozzolanic
reactivity of low grade kaolinitic clays: influence of calcination temperature and
in LC3 by using mixed mineral clays.
impact of calcination products on OPC hydration. Appl. Clay Sci. 108, 94–101.
4. According to the OPC Type I, low-grade kaolinite with a composition Antoni, M., Rossen, J., Martirena, F., Scrivener, K., 2012. Cement substitution by a
of 40–55% results in higher compressive strength gain than the combination of metakaolin and limestone. Cem. Concr. Res. 42, 1579–1589.
standard 42.5 MPa compressive strength of OPC Type I. Even clays Arab, P.B., Araújo, T.P., Pejon, O.J., 2015. Identification of clay minerals in mixtures
subjected to differential thermal and thermogravimetry analyses and methylene blue
with lower grades, at 38–41 MPa, can achieve up to 96.5% of the adsorption tests. Appl. Clay Sci. 114, 133–140.
compressive strength of OPC Type I. ASTM C595, 2010. Standard Specification for Blended Hydraulic Cements. West
5. Correlation analysis revealed that the characterization criteria and Conshohocken, PA.
ASTM C618, 2023. Standard Specification for Coal Ash and Raw or Calcined Natural
strength are robustly correlated. Input weight assumptions, such as Pozzolan for Use in Concrete. West Conshohocken, PA.
the use of various types of low-grade kaolinite clay, can provide Avet, F., Scrivener, K., 2018. Investigation of the calcined kaolinite content on the
significant information for the performance appraisal of LC3 in hydration of Limestone Calcined Clay Cement (LC3). Cem. Concr. Res. 107,
124–135.
MCDM analysis. Avet, F., Snellings, R., Alujas Diaz, A., Ben Haha, M., Scrivener, K., 2016. Development of
a new rapid, relevant and reliable (R3) test method to evaluate the pozzolanic
According to the mineralogy of clays, the reactivity of each clay can reactivity of calcined kaolinitic clays. Cem. Concr. Res. 85, 1–11.
Avet, F., Boehm-Courjault, E., Scrivener, K., 2019. Investigation of C-A-S-H composition,
be assessed using the latest method, known as the R3 test. Afterward, the morphology and density in Limestone Calcined Clay Cement (LC3). Cem. Concr. Res.
optimal reactive condition can be utilized to treat the clay, and subse­ 115, 70–79.
quently, LC3 can be developed to achieve optimal performance. There­ Bernal, S.A., Juenger, M.C.G., Ke, X., Matthes, W., Lothenbach, B., De Belie, N., Provis, J.
L., 2017. Characterization of supplementary cementitious materials by thermal
fore, it is recommended to use the R3 test for each type of clay, and the
analysis. Mater. Struct. 50 (26), 1–13.
performance of LC3 may be further improved. Bernal, I.M.R., Shirani, S., Cuesta, A., Santacruz, I., Aranda, M.A.G., 2021. Phase and
microstructure evolutions in LC3 binders by multi-technique approach including
synchrotron microtomography. Constr. Build. Mater. 300, 124054.
CRediT authorship contribution statement
Dhandapani, Y., Santhanam, M., Kaladharan, G., Ramanathan, S., 2021. Towards ternary
binders involving limestone additions — a review. Cem. Concr. Res. 143, 106396.
Nosheen Blouch: Investigation, Data curation, Formal analysis, Díaz, A.A., Almenares Reyes, R.S., Carratalá, F.A., Martirena Hernández, J.F., 2018.
Proposal of a methodology for the preliminary assessment of kaolinitic clay deposits
Writing – original draft, Writing – review & editing. Khuram Rashid:
as a source of SCMs. In: Martirena, F., Favier, A., Scrivener, K. (Eds.), Calcined Clays
Supervision, Conceptualization, Validation, Formal analysis, Writing – for Sustainable Concrete. Springer, Netherlands, Dordrecht, pp. 29–34.
review & editing. Idrees Zafar: Writing – review & editing, Project Dondi, M., Raimondo, M., Zanelli, C., 2014. Clays and bodies for ceramic tiles:
administration. Mounir Ltifi: Writing – review & editing, Project Reappraisal and technological classification. Appl. Clay Sci. 96, 91–109.
Elimbi, A., Tchakoute, H.K., Njopwouo, D., 2011. Effects of calcination temperature of
administration, Funding acquisition. Minkwan Ju: Supervision, Vali­ kaolinite clays on the properties of geopolymer cements. Constr. Build. Mater. 25,
dation, Writing – original draft, Writing – review & editing. 2805–2812.
Fernandez, R., Martirena, F., Scrivener, K.L., 2011. The origin of the pozzolanic activity
of calcined clay minerals: a comparison between kaolinite, illite and
Declaration of Competing Interest montmorillonite. Cem. Concr. Res. 41, 113–122.
Fernández-González, M.V., Martín-García, J.M., Delgado, G., Párraga, J., Delgado, R.,
2013. A study of the chemical, mineralogical and physicochemical properties of
The authors declare that they have no known competing financial peloids prepared with two medicinal mineral waters from Lanjarón Spa (Granada,
interests or personal relationships that could have appeared to influence Spain). Appl. Clay Sci. 80, 107–116.
the work reported in this paper. Ferreiro, S., Herfort, D., Damtoft, J.S., 2017. Effect of raw clay type, fineness, water-to-
cement ratio and fly ash addition on workability and strength performance of
calcined clay–limestone Portland cements. Cem. Concr. Res. 101, 1–12.
Data availability Garg, N., Skibsted, J., 2016. Pozzolanic reactivity of a calcined interstratified illite/
smectite (70/30) clay. Cem. Concr. Res. 79, 101–111.
Gasparini, E., Tarantino, S.C., Ghigna, P., Riccardi, M.P., Cedillo-González, E.I.,
Data will be made available on request. Siligardi, C., Zema, M., 2013. Thermal dehydroxylation of kaolinite under isothermal
conditions. Appl. Clay Sci. 80, 417–425.
Acknowledgment Ghorbel, H., Samet, B., 2013. Effect of iron on pozzolanic activity of kaolin. Constr.
Build. Mater. 44, 185–191.
Gunjal, S.M., Kondraivendhan, B., 2020. Study on fresh and harden properties of
The authors extend their appreciation to the Deanship of Scientific limestone calcined clay cement (LC 3) production by marble stone powder. In:
Research at Imam Mohammad Ibn Saud Islamic University for funding Calcined Clays for Sustainable Concrete. Springer, pp. 535–543.
Hameed, R., Seo, J., Park, S., Amr, I.T., Lee, H.K., 2020. CO2 uptake and physicochemical
this work through Research Group no. RG-21-12-06.
properties of carbonation-cured ternary blend Portland
cement–metakaolin–limestone pastes. Materials (Basel). 13, 4656.
References Hanpongpun, W., 2019. Investigation of the use of Limestone Calcined Clay Cement
(LC3) Applied to Thailand. EPFL.
He, C., Osbaeck, B., Makovicky, E., 1995. Pozzolanic reactions of six principal clay
Abed, M., Fořt, J., Rashid, K., 2022. Multicriterial life cycle assessment of eco-efficient
minerals: activation, reactivity assessments and technological effects. Cem. Concr.
self-compacting concrete modified by waste perlite powder and/or recycled concrete
Res. 25, 1691–1702.
aggregate. Constr. Build. Mater. 348, 128696.
Hinckley, D.N., 1962. Variability in “crystallinity” values among the kaolin deposits of
Agarwal, S.K., Palla, S., Chaturvedi, S.K., Mohapatra, B.N., Bishnoi, S., Maity, S., 2020.
the coastal plain of Georgia and South Carolina. Clay Clay Miner. 11, 229–235.
Investigation on limestone calcined clay cement system. In: Bishnoi, S. (Ed.),
Hollanders, S., Adriaens, R., Skibsted, J., Cizer, Ö., Elsen, J., 2016. Pozzolanic reactivity
Calcined Clays for Sustainable Concrete. Springer Singapore, Singapore,
of pure calcined clays. Appl. Clay Sci. 132, 552–560.
pp. 443–454.
Horkoss, S., Lteif, R., Rizk, T., 2011. Influence of the clinker SO3 on the cement
Akindahunsi, A.A., Avet, F., Scrivener, K., 2020. The Influence of some calcined clays
characteristics. Cem. Concr. Res. 41, 913–919.
from Nigeria as clinker substitute in cementitious systems. Case Stud. Constr. Mater.
Huang, Zhenyu, Huang, Y., Liao, W., Han, N., Zhou, Y., Xing, F., Sui, T., Wang, B.,
13, e00443.
Ma, H., 2020. Development of limestone calcined clay cement concrete in South
Almenares Reyes, R.S., Díaz, A.A., Rodríguez, S.B., Rodríguez, C.A.L., Hernández, J.F.M.,
China and its bond behavior with steel reinforcement. J. Zhejiang Univ. A 21,
2018. Assessment of Cuban Kaolinitic clays as source of supplementary cementitious
892–907.
materials to production of cement based on Clinker–Calcined Clay–Limestone. In:
IPCC, 2022. AR6: Sixth Assessment Report, Climate Change 2022: Mitigation of Climate
Calcined Clays for Sustainable Concrete. Springer, pp. 21–28.
Change.
Alujas Diaz, A., Almenares Reyes, R.S., Hanein, T., Irassar, E.F., Juenger, M.,
Irassar, E.F., Bonavetti, V.L., Castellano, C.C., Trezza, M.A., Rahhal, V.F., Cordoba, G.,
Kanavaris, F., Maier, M., Marsh, A.T., Sui, T., Thienel, K.-C., 2022a. Properties and
Lemma, R., 2019. Calcined illite-chlorite shale as supplementary cementing
occurrence of clay resources for use as supplementary cementitious materials: a
material: thermal treatment, grinding, color and pozzolanic activity. Appl. Clay Sci.
paper of RILEM TC 282-CCL. Mater. Struct. 55, 1–22.
179, 105143.
Alujas Diaz, A., Almenares Reyes, R.S., Hanein, T., Irassar, E.F., Juenger, M.,
Kandagaddala, R.K., Manare, V., Nanthagopalan, P., 2020. Limestone calcined clay as
Kanavaris, F., Maier, M., Marsh, A.T., Sui, T., Thienel, K.-C., Valentini, L., Wang, B.,
potential supplementary cementitious material—an experimental study. In: Calcined
Zunino, F., Snellings, R., 2022b. Properties and occurrence of clay resources for use
Clays for Sustainable Concrete. Springer, pp. 491–501.
as supplementary cementitious materials: a paper of RILEM TC 282-CCL. Mater.
Struct. 55.

16
N. Blouch et al. Applied Clay Science 243 (2023) 107080

Khalifa, A.Z., Cizer, Ö., Pontikes, Y., Heath, A., Patureau, P., Bernal, S.A., Marsh, A.T.M., Rehman, M.U., Ahmad, M., Rashid, K., 2020. Influence of fluxing oxides from waste on
2020. Advances in alkali-activation of clay minerals. Cem. Concr. Res. 132, 106050. the production and physico-mechanical properties of fired clay brick: a review.
Khan, M.S.H., Nguyen, Q.D., Castel, A., 2018. Carbonation of limestone calcined clay J. Build. Eng. 27, 100965.
cement concrete. In: Calcined Clays for Sustainable Concrete. Springer, pp. 238–243. Rossetti, A., Ikumi, T., Segura, I., Irassar, E.F., 2021. Sulfate performance of blended
Krishnan, S., Bishnoi, S., 2020. A numerical approach for designing composite cements cements (limestone and illite calcined clay) exposed to aggressive environment after
with calcined clay and limestone. Cem. Concr. Res. 138, 106232. casting. Cem. Concr. Res. 147, 106495.
Krishnan, S., Emmanuel, A.C., Bishnoi, S., 2019. Hydration and phase assemblage of Schulze, S.E., Rickert, J., 2019. Suitability of natural calcined clays as supplementary
ternary cements with calcined clay and limestone. Constr. Build. Mater. 222, 64–72. cementitious material. Cem. Concr. Compos. 95, 92–97.
Ma, Y., Qian, J., 2018. Influence of alkali sulfates in clinker on the hydration and Scrivener, K., Martirena, F., Bishnoi, S., Maity, S., 2018. Calcined clay limestone cements
hardening of Portland cement. Constr. Build. Mater. 180, 351–363. (LC3). Cem. Concr. Res. 114, 49–56.
Maier, M., Beuntner, N., Thienel, K.-C., 2021. Mineralogical characterization and Scrivener, K., Avet, F., Maraghechi, H., Zunino, F., Ston, J., Hanpongpun, W., Favier, A.,
reactivity test of common clays suitable as supplementary cementitious material. 2019. Impacting factors and properties of limestone calcined clay cements (LC3).
Appl. Clay Sci. 202, 105990. Green Mater. 7, 3–14.
Maraghechi, H., Avet, F., Wong, H., Kamyab, H., Scrivener, K., 2018. Performance of Sharma, M., Bishnoi, S., Martirena, F., Scrivener, K., 2021. Limestone calcined clay
Limestone Calcined Clay Cement (LC3) with various kaolinite contents with respect cement and concrete: a state-of-the-art review. Cem. Concr. Res. 149, 106564.
to chloride transport. Mater. Struct. 51 (125), 1–17. Sivrikaya, O., Uzal, B., Ozturk, Y.E., 2017. Practical charts to identify the predominant
Marangu, J.M., 2020. Physico-chemical properties of Kenyan made calcined Clay clay mineral based on oxide composition of clayey soils. Appl. Clay Sci. 135,
-Limestone cement (LC3). Case Stud. Constr. Mater. 12, e00333. 532–537.
Medepalli, S., Shah, V., Bishnoi, S., 2016. Production of lab scale limestone calcined clay Sposito, R., Maier, M., Beuntner, N., Thienel, K.-C., 2022. Physical and mineralogical
cements using low grade limestone. In: 7th Int. Conf. Sustain. Built Environ. properties of calcined common clays as SCM and their impact on flow resistance and
Meroufel, B., Benali, O., Benyahia, M., Benmoussa, Y., Zenasni, M.A., 2013. Adsorptive demand for superplasticizer. Cem. Concr. Res. 154, 106743.
removal of anionic dye from aqueous solutions by Algerian kaolin: Characteristics, Tavares, L.R.C., Junior, J.F.T., Costa, L.M., da Silva Bezerra, A.C., Cetlin, P.R.,
isotherm, kinetic and thermodynamic studies. J. Mater. Environ. Sci 4, 482–491. Aguilar, M.T.P., 2020. Influence of quartz powder and silica fume on the
Mishra, D.K., Yu, J., Leung, C.K.Y., 2020. Comparison of strength activity of limestone- performance of Portland cement. Sci. Rep. 10, 1–15.
calcined clay and class F fly ash. In: Calcined Clays for Sustainable Concrete. Tironi, A., Trezza, M.A., Scian, A.N., Irassar, E.F., 2013. Assessment of pozzolanic
Springer, pp. 481–490. activity of different calcined clays. Cem. Concr. Compos. 37, 319–327.
Msinjili, N.S., Gluth, G.J.G., Sturm, P., Vogler, N., Kühne, H.-C., 2019. Comparison of Vizcaino Andres, L.M., Antoni, M.G., Alujas Diaz, A., Martirena Hernandez, J.F.,
calcined illitic clays (brick clays) and low-grade kaolinitic clays as supplementary Scrivener, K.L., 2015. Effect of fineness in clinker-calcined clays-limestone cements.
cementitious materials. Mater. Struct. 52 (94), 1–14. Adv. Cem. Res. 27, 546–556.
Nguyen, Q.D., Castel, A., 2020. Reinforcement corrosion in limestone flash calcined clay Wang, B., Sui, T., Sui, H., 2020. Identification and activation of coal gangue and
cement-based concrete. Cem. Concr. Res. 132, 106051. performance of limestone calcined gangue cement. In: Bishnoi, S. (Ed.), Calcined
Nguyen, Q.D., Khan, M.S.H., Castel, A., 2018. Engineering Properties of Limestone Clays for Sustainable Concrete. Springer Singapore, Singapore, pp. 381–389.
Calcined Clay Concrete. J. Adv. Concr. Technol. 16, 343–357. Yu, J., Wu, H.-L., Mishra, D.K., Li, G., Leung, C.K.Y., 2021. Compressive strength and
Nguyen, Q.D., Afroz, S., Castel, A., 2020a. Influence of calcined clay reactivity on the environmental impact of sustainable blended cement with high-dosage Limestone
mechanical properties and chloride diffusion resistance of limestone Calcined Clay and Calcined Clay (LC2). J. Clean. Prod. 278, 123616.
Cement (LC3) concrete. J. Mar. Sci. Eng. 8, 1–14. Yu, T., Zhang, B., Guo, H., Wang, Q., Liu, D., Chen, J., Yuan, P., 2023. Calcined nanosized
Nguyen, Q.D., Kim, T., Castel, A., 2020b. Mitigation of alkali-silica reaction by limestone tubular halloysite for the preparation of limestone calcined clay cement (LC3). Appl.
calcined clay cement (LC3). Cem. Concr. Res. 137, 106176. Clay Sci. 232, 106795.
Ortega, A., Macías, M., Gotor, F.J., 2010. The multistep nature of the kaolinite Zafar, I., Rashid, K., Tariq, S., Ali, A., Ju, M., 2022. Integrating technical-environmental-
dehydroxylation: kinetics and mechanism. J. Am. Ceram. Soc. 93, 197–203. economical perspectives for optimizing rubber content in concrete by multi-criteria
Patarachao, B., Tyo, D.D., Davis, Z.K., Kung, J., Ng, S., Mercier, P.H.J., 2019. XRD analysis. Constr. Build. Mater. 319, 125820.
Analysis of Illite-smectite Interstratification in Clays from Oil Sands ores. Adv. X-ray Zhou, Y., Gong, G., Xi, B., Guo, M., Xing, F., Chen, C., 2022. Sustainable lightweight
Anal. 62, 22–31. engineered cementitious composites using limestone calcined clay cement (LC3).
Rakhimova, N., Rakhimov, R., 2020. Advances in development of calcined clays as Compos. Part B Eng. 243, 110183.
supplementary cementitious materials. IOP Conf. Ser. Mater. Sci. Eng. 890, 012085. Zolfagharnasab, A., Ramezanianpour, A.A., Bahman-Zadeh, F., 2021. Investigating the
Ranjan, R., Narnaware, S.D., Patil, N.V., 2019. A novel technique for synthesis of calcium potential of low-grade calcined clays to produce durable LC3 binders against
carbonate nanoparticles. Natl. Acad. Sci. Lett. 41, 403–406. chloride ions attack. Constr. Build. Mater. 303, 124541.
Rashid, K., Razzaq, A., Ahmad, M., Rashid, T., Tariq, S., 2017. Experimental and Zunino, F., Scrivener, K., 2021. The reaction between metakaolin and limestone and its
analytical selection of sustainable recycled concrete with ceramic waste aggregate. effect in porosity refinement and mechanical properties. Cem. Concr. Res. 140,
Constr. Build. Mater. 154, 829–840. 106307.
Rashid, K., Rehman, M.U., de Brito, J., Ghafoor, H., 2020. Multi-criteria optimization of
recycled aggregate concrete mixes. J. Clean. Prod. 276, 124316.

17