The effect of marble powder on physico‐mechanical and microstructural properties of kaolin‐based geopolymer pastes

As an alternative to ordinary Portland cement (OPC), alkali‐activated materials (AAMs) have more recently been studied and found to have certain suitability in reducing ecological footprint of OPC binding systems. Nonetheless, due to recent concerns over the availability of certain precursors used AAMs (such as the reduction of coal fly ash availability), this study utilized marble powder in various quantities to replace the naturally available kaolin as precursor. This composite use of natural precursors has been entertained to assimilate the production of carbon zero AAMs. To evaluate the physico‐mechanical and microstructural characteristics of the materials, 48 mixes with different sodium (Na) concentrations, curing temperatures, and marble powder content have been used. The results show that rising curing temperature is more effective than other variables, such as the Na and marble powder content on the physico‐mechanical performance of tested geopolymers. In this regard, it is found that a very high Na content can have adverse effect on the properties, potentially due to altered Na/Si, Na/Al ratio in the mixes. Furthermore, the inclusion of marble powder is found to be effective in decreasing the overall porosity up to ~31% and enhancing the physico‐mechanical properties of the specimens cured at 20 and 80°C. Nonetheless, results show that when specimens containing marble powder are exposed to higher curing temperatures (above 80°C) the presence of marble powder adversely affects physico‐mechanical properties. It is concluded that this phenomenon is caused by the dehydration of chemically bound water in higher temperatures when marble powder is used. This result is further confirmed by microstructural tests.


K E Y W O R D S
alkali activation, alkali-activated materials (AAMs), kaolin, marble powder, sodium hydroxide

| INTRODUCTION
The need for ordinary Portland cement (OPC) has been growing with a continuous rise in the global population and economic expansion in recent decades.It is reported that the OPC industry is one of the most energy-consuming sectors and a leading cause for CO 2 emissions due to the highly energy-intensive thermal processes performed in cement manufacturing. 1,2It this regard, it is estimated that the global cement market will experience a 200% increase by the year 2050, compared to the 2010 levels. 3herefore, global efforts to find eco-friendly materials to replace cement have recently emerged as an alternative way to eliminate the detrimental environmental impacts caused by OPC production. 4,5Alkali-activated materials (AAMs), in general (also geopolymers), have been studied for about a century and are considered as a sustainable binder that can potentially replace or address the need for OPC due to their reduced embodied energy and minimized CO 2 emissions. 6,7][10] In this way, geopolymers can achieve up to 80% less CO 2 emission compared to the OPC. 8,11ccordingly, substituting OPC with geopolymers can surmount some crucial concerns, such as high energy utilization, human-caused greenhouse gas emissions, and depletion of natural resources. 12he determination process of aluminosilicates to employ in geopolymers is quite significant, since it remarkably affect the geopolymers' final characteristics.4][15][16] Nevertheless, it is believed that the global access of these popular aluminosilicates will be limited in the future to meet the need for the potential large-scale use of geopolymer in the concrete industry. 5,17Therefore, choosing suitable precursors for geopolymers has also broadened enormously in the last decades.Palm oil fuel ash, 18 red mud, 19 as well as some other byproducts, [20][21][22] and minerals [23][24][25] have been investigated and validated for utilization in geopolymer binders.In addition, clay-based minerals (i.e., kaolinite, smectite, illite, quartz and feldspar) are a group of the most abundant natural aluminosilicates on the earth that account for nearly 90% of Earth's crust. 26In general, clays show higher pozzolanic activity after a thermal treatment process (between 500 C to 900 C), reducing crystallinity for potential use as precursors for geopolymer formulations. 27The calcination of clay between temperatures of 500-900 C increases the reactivity, creates an amorphous phase, and significantly improve its geopolymerization rate. 28][31][32] Nevertheless, thermal treatment of clays inevitably raises energy utilization and production costs, limiting the large-scale applications of geopolymers in building industry.Nonetheless, there has been a recent research effort to employ the non-calcined clays, namely kaolin, as a precursor for the preparation of geopolymers.Kaolin, also known as China clay, refers to a group of kaolinitic clays, and it has been widely used in the production of OPC, geopolymers, and as supplementary cementitious materials (also known as SCMs). 33Recently, Vincent et al. 1 developed an eco-friendly geopolymer mortar including FA, GGBS, and non-calcined kaolin activated by solid alkaline and examined the effect of kaolin on fresh and mechanical properties of geopolymer.The authors reported the marked potential of non-calcined kaolin for geopolymer concretes' development in terms of improved mechanical properties and lowered embodied CO 2 ($134%).Nnaemeka and Singh 34 examined the durability properties of FA-based geopolymers including kaolin with the presence of sodium hydroxide/silicate.The study showed that substituting FA with kaolin up to 40 wt% (weight percent) improved the resistance of geopolymers to acid and sulfate attacks.The effect of curing temperature (40, 60, 80, and 100 C) and duration (1-3 days) on characteristics of kaolin-based geopolymer was conducted by Heah et al. 35 The authors found that the mechanical performance of produced pastes was significantly influenced by curing conditions and curing at 60 C for up to 3 days optimally contributed samples' mechanical and microstructural performance.Although the viability of kaolin as a precursor in geopolymers has been proven, an in-depth understanding is required to increase the reactivity and geopolymerization rate of kaolin, especailly when used in composite form to other natural precursors. 36To address this, several studies suggested some strategies, such as increasing the surface area of precursor, 37 thermal curing, 38 as well as the blended use of alternative activators and additives. 39Very recently, Kaya et al. 40 investigated the effect of natural salts (sodium and magnesium sulfate) on physicomechanical performance of kaolin and ceramic powderbased geopolymer mortars.The outcomes of this study emphasized the suitability of using kaolin as precursor for a cleaner production of geopolymer concrete.
2][43] In this regard, Yip et al. 43 investigated the effect of alkaline earth carbonate minerals, including calcite and dolomite, on the microstructural and mechanical properties of metakaolin-based geopolymers.The authors found that the incorporation of up to 20 wt% of calcite or dolomite has a supportive effect on mechanical performance of metakaolin-based geopolymeric binder.Further to this, Qian et al. examined the effect of limestone powder on fresh and mechanical performance of metakaolin-based geopolymer binder.It was reported that a 10 wt% addition of limestone powder improved the 7-day compressive and flexural strength of mortars by almost 17% and 14%, respectively.Moreover, Xie et al. 16,44 examined the combined utilization of GGBS and FA or metakaolin in different studies to show the coupling effects of Ca and Al, Si sources on the physicochemical performance of geopolymer concrete.The authors confirmed that using high source of Ca materials (GGBS) provided an excellent synergetic effect on physicochemical and mechanical performance of FA and metakaolin-based geopolymers.It can be summarized as the addition of an appropriate amount of calcium-containing material to the kaolin or FA-based geopolymers has found to have a synergetic effect, leading to the formation of sodium (alumino) silicate hydrate (N-A-S-(H)) and calcium (alumino) silicate hydrate (C-A-S-H) gel, favorably affecting the mechanical properties. 7One material for such potential use, is marble powder that has been utilized as a common construction material for several purposes for decades.Quarrying, sawing, shaping, and polishing processes generate a waste known as marble powder which can exceed millions of tons annually, thus creating an irreversible environmental and health problem and reducing the fertility of lands. 45,46Although the disposal of this waste material is one of the major environmental issues, it has been successfully utilized in the construction industry for several decades.2][53] Moreover, inclusion of marble powder as an aluminosilicate source to synthesize geopolymers has gained so far limited interest from researchers.It is observed that marble powder shows a low reactivity due to having lower Si and Al content. 54However, combining marble powder with other aluminosilicates with high concentrations of Si and Al may improve the reactivity and geopolymerization.Very recently, Komnitsas et al. 55 activated marble powder in conjunction with metakaolin, NaOH, and Na 2 SiO 3 , forming geopolymer.It is reported that incorporation of marble powder into the metakaolin-based geopolymer enables the formation of geopolymer with better physicomechanical properties, as well as contributing to the utilization of the waste and reducing the detrimental environmental effects of marble waste.Du et al. 45 examined the fluidity, compressive strength, and microstructure of Na 2 CO 3 activated superfine slag-based geopolymer with presence of marble powder.The results showed that a 30 wt% incorporation of marble powder increased the fluidity of mortar by almost 22% but yielded an almost 20% decrease in compressive strength.][56][57] Aside from the selection of aluminosilicates for the preparation of geopolymers, the type of alkaline activators and their concentration are critical in understanding the dissolution of minerals.In general, sodium NaOH and KOH are usually utilized as activators for clay-based geopolymers. 58Although the geopolymerization rate of KOH is faster than NaOH's, a higher rate of dissolution is typically observed in NaOH-activated systems. 26Moreover, researchers believe that NaOH have tendency to form zeolite rather than geopolymers. 26However, the wide availability, low viscosity, and cost efficiency of NaOH affect the selection of alkaline activator for synthesizing geopolymers. 58Another key parameter for the determination of activator concentration is the higher activator ratio, yielding faster dissolution and geopoloymerization of clay minerals, which influence the density, strength, and microstructure characteristics. 59,60The curing temperature of clay-based geopolymers significantly influences geopolymerization rate.Although the room temperature can be adequate for geopolymerization of calcined clay-based (metakaolin) geopolymers, higher curing temperatures (60-100 C) may be needed to achieve proper reaction of non-calcined clay minerals. 61,62Therefore, it is advised to identify optimum curing temperatures to both accelerate the formation of geopolymers and avoid to drying shrinkage cracks, thus improved mechanical and durability characteristics can be obtained. 63Overall, several parameters including activator concentrations and curing temperatures linked with materials design on physico-mechanical and microstructural performance of final products needs more fundamental assessment.
From the limited number of previous studies, kaolin and marble powder have potential to react and generate aluminosilicate formations under particular scenarios but combining these materials for synthesizing geopolymers has not yet been investigated.This study answers the question of how the abundant resources of kaolin and marble powder can be best valorized for environmentalfriendly and cost-effective geopolymer systems with improved mechanical and microstructural performance.Although considerable progress has been made in adopting NaOH in geopolymers, it is necessary to further examine the NaOH concentration in the chemistry of geopolymer systems that incorporates both kaolin and marble powder as a precursor.To do so, a set of experimental tests, including compressive and flexural strengths, dry unit weight, ultrasonic pulse velocity (UPV), porosity, and water absorption, were investigated in the present study to enable a path to adopt these materials in geopolymer.Beyond these common experimental tests, geopolymers' thermal decomposition and chemical crystallographic structure of raw materials were assessed by thermogravimetric analysis (TGA) and X-ray diffraction (XRD) analysis, respectively.Finally, the microstructure of the produced samples was examined via scanning electron microscopy (SEM) and energy dispersive X-ray spectroscopy (EDS).The effect of different concentrations of marble powder (0, 10, 20, and 30 wt%), NaOH contents (8, 10, and 12 wt% Na), and diverse curing temperatures (20, 80, 100, and 120 C) on characteristics of geopolymer system were carefully explored.This paper presents a comprehensive experimental study that unlocks opportunities for a better understanding of kaolin-based geopolymers including marble powder that offer improved physico-mechanical and microstructural properties and reduced carbon footprint.

| Materials
Kaolin and marble dust (also noted as marble powder) were obtained from Utelka and Arısu Inc. in Turkey, respectively.The chemical composition of kaolin and marble powder were examined through X-ray fluorescence (XRF) analysis and the results are listed in Table 1.Particle size distribution of kaolin and marble powder are exhibited in Figure 1.

| Mix proportions, specimen preparation, and testing procedures
In this study, the mixture was prepared by replacing 10, 20, and 30 wt% marble powder with kaolin.NaOH F I G U R E 1 PSD of kaolin and marble powder.PSD, particle size distribution was utilized as activatior with a constant liquid to binder ratio of 0.35.The mixtures were designed with three different Na/binder ratios of 8%, 10%, and 12% Na.The raw components, including kaolin and marble powder, were mixed in a dry form in a mechanical mixer for 60 s.Then, the NaOH + water solution was introduced to the dry mixture, and mixing was performed for 3 more minutes.The prepared pastes were placed in 40 Â 40 Â 160 mm prism molds, and molds were placed in the oven and kept at the curing temperatures of 80, 100, and 120 C for 24 h.The reason for using different curing temperatures was to evaluate the impact of various curing temperatures and quicker strength gain for samples since the used precursors were not very highly reactive cementitious materials. 59,60As a result, a group of samples produced were kept at room temperature (20± 2 C) without being placed in an oven to observe the effect of ambient curing activation temperature on the properties of samples.In this regard, a total of 48 mixtures were prepared and the demolded samples were kept at room temperature for 28 days.The mechanical performance of produced pastes, including compressive and flexural strength, were tested based on ASTM C67. 64The unit weight, porosity, and water absorption tests of the geopolymers were carried out per ASTM C20. 65 Further from Figure 3, it can be seen that the variation in Na/binder percent content values has had a slight impact on the developed strengths.At either of curing temperatures, for instance, the mean compressive strength values of samples with Na of 8%, 10%, and 12% is found to be 8.2, 7.4, and 6.7 MPa, respectively.This shows about 18.5% reduction of strength values when a Na% of 12 is used, as opposed to Na% of 8. Reporting from Ref. 69, an unconditionally high  Nonetheless, as can be seen in the figures, the inclusion of marble powder favorably affect strength values, as long as the curing temperature does not exceed 80 C.This is found to be the case in either of the Na% values of 8, 10, or 12.The reason for this can be due to increased internal porosity, as concluded based on the results of Refs.70, 71 and that subjecting geopolymer mortar to higher temperatures can significantly increase internal strain values, due to shrinkage, reducing the physico-mechanical properties of the produced samples.Further from Figure 4, the incorporation of marble powder had a generally positive impact on specimens produced with any of the Na% of 8, 10, and 12, as long as the curing temperature did not exceed 80 C. In this regard, in the Na% of 8, 10, and 12 for specimens cured at 80 C, the flexural strength values are found to increase by 54%, 68%, and 84% when marble powder at the rate of 30% is used.However, at 120 C curing temperature, the mentioned values change to a reduction of $14%, 55%, and 51%, for samples produced with a Na% of 8, 10, and 12, respectively.This indicates that the inclusion of 30% marble powder has significantly reduced the flexural strength values when exposed to higher curing temperatures.This phenomenon can be due to increased internal porosity of mixes containing marble powder that when exposed to higher temperatures results in dehydration of samples, as also reported by Refs.47, 74.

| Unit weight
Figure 5 represents the result of unit weight test conducted on various mixes.The unit weight values range from 1.49 to 1.82 g/cm 3 for N12M0-20 and N8M0-120, respectively.Further, it can be seen that an increase in Na% from 8 to 10 and 12 has reduced the mean unit weight values of specimens by 3.2% and 2.8%, respectively.This reduction can be related to variation in geopolymerization as the Si/Na lowers at higher activator concentration. 75,76urther from Figure 5, the rise in the curing temperature also favorably affects the unit weight values.At Na% of 8, 10, and 12, for instance, the increase in temperature from 20 C to 120 C has increased the unit weight values by $13.6%, 10.9%, and 13.8%, respectively.The increased values of unit weights can be associated with increased polycondensation and formation of geopolymer paste. 77,78onetheless, as the marble powder is added to the mixes, its impact is found to be relatively positive in unit weight values, until exposed to 100 and 120 C. As can be seen in Figure 5, the inclusion of marble powder at 30% with Na% of 8 and cured at 20, 80, 100, and 120 C has changed the unit weight values by +4%, 1%, À4%, and À3%, respectively.Similar trend can be seen when a Na% of 10 and 12 are also used and point to lower unit weight values when 30% marble powder is used and cured at higher temperatures.

| Porosity
The porosity of 28-day cured samples is presented in Figure 6.The range of porosity values is found to be F I G U R E 6 (a-c) Porosity of various mixes and (d) relationship between porosity and compressive strength from 4.45% to 18% for N8M0-120 and N12M0-20, respectively.Based on the results, the increasing curing temperature from 20 C to 120 C lowered the mean porosity values by 55%, 48%, and 55% for mixes produced with Na% of 8, 10, and 12, respectively.This shows that the mean impact of thermal curing for all geopolymer mixes is found to be positive.Similar to the effect of temperature, as the Na% is increased from 8 to 10 and 12, the porosity values are found to increase from 10.5% to 11.2% and 11.9%, respectively.This change in porosity value is believed to have been caused by higher ratios of Na/Si and Na/Al that in very high values can adversely affect the geopolymerization process. 69In this regard, as reported by Ref. 79, a higher molarity value can cause higher shrinkage.Given this fact, and since porosity in this study has been tested based on ASTM C20 that evaluates porosity based on surface permeability, higher surface cracking caused by autogenous shrinkage can have contributed to higher porosity values of samples with higher molarity.
Despite the impact of curing temperature and Na% concentrations, the inclusion of marble powder is found to be more dependent on the curing temperature.As illustrated in Figure 6, the inclusion of 30% marble powder is found to reduce the porosity values when cured at temperatures of 20 and 80 C while, in contrast, it increased the pore content when cured at 100 and 120 C. For samples that were cured at 100 and 120 C, for instance, the inclusion of 30% marble powder has increased porosity by $21% and 120% when activator had a Na% of 10 as well as 37.5% and 67% when a Na% of 12 has been used, respectively.This shows that the increase in curing temperature above 80 C results in higher porosity values, irrespective of the Na% of mixture and points to dehydration of chemically bound water in higher temperatures when marble powder is used.Similar results are reported by Ref. 80 that this phenomenon takes place at a temperature of around 110 C and is intensified when Na/Si ratio is higher.Figure 6d presents the relationship between porosity and compressive strength values reported in Section 3.1.Based on the figure, the variables have a linear relationship that further confirms the findings of this study.

| Water absorption
Figure 7a-c presents the result of water absorption of samples tested after 28 days of curing.It can be seen that water absorption values ranged from 2.2% to 12% for N8M0-120 and N12M10-20, respectively.Further from Figure 7, mean water absorption values for specimens Water absorption of various mixes cured at temperatures of 20, 80, 100, and 120 C is 8.8%, 6.85%, 7.11%, and 3.79%, respectively, when Na% of 8 is used.Similar trend can be seen when Na% of 10 and 12 are also used which point to the average impact of thermal curing in reducing the absorption values.This is believed to have been caused by enhanced geopolymerization at higher curing temperatures, 66,69 which is aligned with the results of Section 3.6.Nonetheless, unlike the impact of curing temperature, the mean absorption values for specimens produced with a Na% of 8, 10, and 12 is found to be 6.5%, 7%, and 7.5%, respectively.As can be seen, the higher intensity of NaOH concentration has slightly increased the absorption values.Yet, regardless of the Na% concentration, the mean absorption of specimens cured at temperatures of 20, 80, 100, and 120 C is found to be 10, 7.8, 6.4, and 4.6, respectively.This shows that on average, the higher curing temperature has significantly reduced the absorption values which can point to enhanced microstructural development of geopolymer samples.Similar results are reported by Ref. 81 that used NaOH molarity of 2-16 and cured specimen at temperatures of 27-60 C.
In addition, as can be seen in the figure, increasing the marble powder loading is found to have a varying impact on the absorption values which is similar to the results reported in Section 3.4.In this sense, it can be seen that when 30% marble powder is used with various Na%, it changes the mean water absorption values from 5.7% to 6.7%.This shows $17% increase in the mean water absorption when 30% marble powder is used which mostly resulted from specimens cured at 100 and 120 C.This result is aligned with other outcomes of this study and point to the slightly adverse impact of marble powder in absorption values when specimens are exposed higher temperatures.Ref. 82 used marble powder in the production of structural brick and reported that the inclusion of marble content is proportional to the increase in water absorption values and achieved an absorption rate of 35%, when 30% marble powder is used.

| UPV
Figure 8a-c presents the result of UPV conducted on 28-day cured samples.Based on the figure, the mean UPV of samples produced with Na% of 8, 10, and 12 is found to be $1550, 1489, and 1452 m/s, respectively.This shows that the increased concentration of Na in the mixture, has slightly reduced the UPV values which points to higher content of pores.Ref. 83 associated this variation to the extent of geopolymerization.Nonetheless, this is in line with test results in Sections 3.4 and 3.5 that higher ratio of Na/Si has increased the internal porosity.In contrast, it can be seen from Figure 8 that the mean UPV values for samples cured at 20, 80, 100, and 120 C is found to be $1095, 1310, 1589, and 1993 m/s, respectively.This shows that the increase in curing temperature has resulted in higher UPV values.This is the opposite to the result of porosity outlined in Section 3.4 and can be credited to the increased content of unconnected internal porosity when higher curing temperature is exercised. 66Similar to the result of porosity outlined in Section 3.4, when marble powder at the rate of 30% is used, it tends to slightly increase the UPV, but reverses to comparatively lower the UPV in higher temperatures.This result further confirms that thermal curing has increased the content of internal porosity that is intensified as the marble powder is included in the mixes.Ref. 84 also showed that when marble powder is used up to 30%, it reported a slightly lowered UPV, associating it with higher porosity of the resulting specimens.
Further, Figure 8d illustrates the UPV with porosity values discussed in Section 3.4 in a scatter plot.As can be seen in the figure, a semi-linear relationship between the two parameters can be made which confirm their impact on each other.

| XRD
XRD diffractograms of geopolymers for different content of marble powder replacing kaolin are shown in Figure 9.As can be seen from the crystallographic structure, that N8M0 and N12M0 specimens comprised quartz (Q), kaolinite (K), and alunite (A) as major mineral phases.XRD patterns for the geopolymers with 30 wt% of marble powder contain calcite (C) peaks in addition to the quartz, kaolinite, and alunite mineral phases.The presence of quartz, kaolinite, and alunite peaks come from unreacted or semi-reacted kaolin utilized as precursor, while calcite peaks are accounted for high content of CaO coming from the marble powder.A remarkable difference in the intensity of mineral phases between the N8M0 and N12M0 has not been observed.However, unlike the trend observed in specimens with 0 wt% marble powder, increased Na% from 8 to 12 weakened the intense peaks of calcite and quartz for the specimens with 30 wt% of marble powder.It is likely that Ca ions either react with silicate to form N(C)-A-S-H or form crystalline precipitates, including Ca(OH) 2 or CaCO 3 . 43It is, however, challenging to understand from only XRD data how exactly Ca ions participate in geopolymerization.Although there is still a significant amount of undissolved calcite existing in both N8M30 and N12M30, the increasing Na concentration yielded a reduction in calcite peaks which can be confirmed with mechanical test results.The 20% compressive strength gain of specimens (N8M30 and N12M30) with enhanced Na% from 8 to 12 also affirms the participation of the intense peaks in the gelation process.Moreover, the major quartz peak around 25 -37 2θ is more pronounced in the N8M30 specimen compared to the N8M0.Furthermore, it can be seen that an almost 33% loss in the compressive strength of N8M0 with the

Intensity (a.u)
F I G U R E 9 XRD analysis of geopolymers cured at 100 C. Q: quartz; K: kaolinite; A: alunite; C: calcite; M: muscovite.XRD, X-ray diffraction substitution of 30 wt% marble powder can be attached to the presence of this prominent quartz peak.Overall, a number of intense kaolinite, quartz, and calcite peaks were witnessed in the diffractogram of tested specimens due to the semi-reacted kaolin and marble powder content. 85However, differences between the mechanical performances and microstructural observations are highly dependent on the activation temperature, making it challenging to draw a specific trend based on a single curing temperature.

| TGA and DTA
The TGA of the kaolin-based geopolymers with different marble powder content and alkaline activator concentration were examined and are presented in Figure 10.Based on the figure, the mass loss can be divided into several steps based on the marble powder content and Na concentration of activators.The first and significant mass loss was detected in TGA and derivative thermogravimetr (DTG) curves for all tested specimens up to $150 C. The mass loss at this temperature range was identified as 5.2%, 7.1%, 6.5%, and 7% for N8M0, N8M30, N12M0, and N12M30 specimens, respectively.This mass loss can be attributed to the free water from liquid activator and in the pore network of gel structure. 86,87The evaporation of physicochemical bound water from sodium/(sodium-calcium) aluminosilicates was determined up to temperatures of $400 C. The mass loss caused due to the loss of physico-chemical bound water was determined as 4.6%, 3.3%, 5.1%, and 5.0% for N8M0, N8M30, N12M0, and N12M30 specimens, respectively.The minor peak observed from 450 C to 600 C is associated with dehydroxylation of OH groups and polycondensation process of geopolymer gels. 88,89Mass loss due to the dehydroxylation of OH groups and polycondensation process for N8M0 and N12M0 specimens accounted for 3.0% and 3.6%, respectively.This is also in line with mechanical test results, which indicates higher strength values for N12M0 specimens.The onset of calcite decomposition can also be witnessed between 600 C and 700 C in activated geopolymers with 70 wt% kaolin and 30 wt% marble powder binders.The decomposition of calcite for N8M30 and N12M30 specimens caused almost 5.7% and 7.9% mass loss, respectively.This is in good agreement with the calcite mineral peaks found in the XRD test results of N8M30 and N12M30.The higher calcite decomposition for N12M30 confirms higher amount of reaction products, which is also consistent with compressive strength test results.witnessed in all the examined surfaces, but the density of pores altered with the extent of geopolymerization and gel formation.It is worth noting that the density decreases with increasing marble powder content for specimens cured at 100 C, independent of the mixture's Na% concentration.This microstructural observation can be associated with increased porosity values of geopolymers via dehydration of chemically bound water at higher curing temperatures exceeding 80 C with the presence of marble powder.The unfavorable influence of high curing temperatures on the microstructure of geopolymers with marble powder is also confirmed by mechanical test results, as well as some previous studies. 74,80It is found that a more porous microstructure has formed in specimens with a Na % of 12 as compared to 8 regardless of marble powder content.This variation in the microstructure is considered to have been resulted from higher ratios of Na/Si and Na/Al that detrimentally influence the geopolymerization process which is also documented by earlier studies. 69,72he EDS analysis was conducted to identify substances formed through geopolymerization, and the analyzed areas are numbered and highlighted in Figure 11.The elemental compositions (by wt%) of analyzed spectrum areas can be seen in Table 3. From elemental analysis, three different gel formations of N-(C)-A-S-H, N-A-S-H, and C-A-S-H, are determined as the hydration products of tested geopolymers.For the N8M0 specimen, sodium aluminosilicate is the primary hydration gel, while calcium and sodium-calcium aluminosilicates are found as gel formations for N8-M30 and N12M30.This is attributed to the presence of calcite in marble powder, contributing to the formation of calcium aluminosilicates.The Si/Al ratio lies in the range of 4.5-6.6,4.8-6.4,and 1.7-4.2 in N8M0, N8M30, and N12M30 specimens' detected areas, respectively.Moreover, the Ca/Si ratio varies between 0.33-1.99 and 0.35-0.46 in N8M30 and N12M30 specimens, respectively.It is noteworthy that it is challenging to report a specific relationship between mechanical properties and Si/Al and Ca/Al ratios due to the limited number of analyzed points with EDS. Figure 12 demonstrates the ternary composition plot of Al-Si-Ca of different areas in produced geopolymers.The graph presents the gel formation variation with available Al, Si, and Ca content in analyzed areas.

| CONCLUSIONS
In this study, the effect of replacing marble powder with kaolin at different curing temperature and variety of activator ratio has been examined.Based on the test results, the following conclusions can be drawn: • The compressive and flexural strength results showed that the curing temperature is most effective in developing strength and a higher activator ratio is found to reduce strength values.This is believed to have been caused by a high Na/Si ratio in the mixture.Furthermore, the inclusion of marble powder is found to slightly increase but then decrease strength values, when the samples are exposed to 20 C versus 120 C, respectively.The initial increase can be due to micro pore filling effect of marble powder which can result in further incompatibility when exposed to high temperatures.The temperature tipping point for this is found to be 80 C, after which, the impact of marble powder becomes negative.• Similarly, the result of unit weight, porosity, water absorption and UPV tests point to the reduction of pores and a slight increase in unit weight values when a higher curing temperature is used.Nonetheless, when activator ratio is increased, a higher pore content and a lower unit weight is achieved.Furthermore, the inclusion of marble powder is found to slightly increase unit weight values and reduce porosity which reverses after 80 C. It is believed that this is caused by dehydration of chemically bound water that takes place in higher temperatures when marble powder is used.• The XRD diffractogram of the activated samples at 100 C shows that the increased Na% of the activator from 8 to 12 did not significantly affect the intensity of peaks for N8M0 and N12M0 while it reduced the intensity of calcite and quartz peaks in samples with marble powder.This is attributed to the participation of Ca from marble powder contributing to gel formation at higher Na%.• TGA and DTG curves reveal dehydroxylation of OH groups and polycondensation process of geopolymer gels between 450 C and 600 C and decomposition of calcite can be witnessed between 600 C and 700 C in activated geopolymers with 70 wt% kaolin and 30 wt% marble powder binders.• SEM micrographs indicate that a more porous microstructure has observed in specimens with Na of 12% as compared to the 8% regardless of marble powder concentration.This alteration in the microstructure is considered to have been resulted from higher fractions of Na/Si and Na/Al that detrimentally influence the geopolymerization process.From elemental analysis, three different gel formations, N-(C)-A-S-H, N-A-S-H, and C-A-S-H, are determined as the main hydration products of tested geopolymers.
Results of this study have shown that the utilization of marble powder as a substitution for kaolin as a precursor is promising and improves the physico-mechanical properties of kaolin-based geopolymers under different environmental conditions.Nonetheless, future studies are required to draw further detailed information on the geopolymerization process when marble powder is used in conjunction with various cementitious materials to optimize the mix design without the requirement of thermal curing.
analysis of (a) kaolin and (b) marble powder.XRD, X-ray diffraction T A B L E 2 Mixture proportions and activation temperatures activator (very high molarity values) can remarkably change the physico-mechanical and porosity of the produced samples, resulting in higher porosity, efflorescence, and a reduction of mechano-durability properties.

Figure
Figure 4a-c illustrates flexural strength test results of 28-day cured specimens.As can be seen in the figure, the strength values range from 0.33 to 4.46 MPa for N12M0-20 and N8M0-120, respectively.Similar to the results discussed in Section 3.1, the increase in the curing temperature is found to increase the flexural strength values from a mean of 0.97 MPa when cured at 20 C, to 4.04 MPa when cured at 120 C, for samples produced with Na 8%.Similar trend is observed for specimens produced with a Na% of 10 and 12. Nonetheless, it can be seen that the increase in Na% has not had a positive impact on the mean strength values, similar to the results discussed for compressive strength.For instance, for specimens cured at 120 C,

F
I G U R E 8 (a-c) UPV result of various mixes and (d) the relationship between compressive strength and porosity values.UPV, ultrasonic pulse velocity

F
I G U R E 1 0 TGA and DTG curves of geopolymers cured at 100 C. TGA, thermogravimetric analysis

Figure
Figure 11a-d represents the micrograph of 28-day cured (100 C) kaolin-based geopolymers with different contents of marble powder (0 and 30 wt%) and Na% of alkaline activators (8% and 10%).Overall, the pores and voids are

2
Ternary graph of elemental composition (Si-Al-Ca) T A B L E 1 XRF analysis of kaolin and marble powder Abbreviation: XRF, X-ray fluorescence.