Heat of hydration, water sorption and microstructural characteristics of paste and mortar mixtures produced with powder waste glass

Abstract

This paper investigates the effects of powder waste glass (PWG) at 10%, 15%, and 20 wt.% replacement of cement on the flow characteristic, setting time, compressive strength, water sorption, rate of the heat of hydration, cumulative heat of hydration; alkali-silica reaction (ASR), and microstructure characteristics of the resulting paste and mortar mixtures. A total of eight cementitious mixtures including two control mixtures have been investigated in this experimental program to evaluate the effect of PWG on the fresh and hardened state characteristics of paste and mortar mixtures. Test results show that the incorporation of PWG in the cementitious mixtures results in significant enhancement of their microstructure, increase in compressive strength, improvement in moisture barrier characteristics, and considerable reduction in the rate of the heat of hydration and cumulative heat of hydration. Up to 20 wt.% replacement of cement with fine PWG is seen to be innocuous from the standpoint of ASR. Furthermore, the use of PWG in paste results in delaying the initial and final setting times and increases the flow characteristic of the mortar mixture.

Keywords:

• Alkali silica reaction
• compressive strength
• heat of hydration
• powder waste glass
• water sorption

Reviewing Editor:

• Sanjay Kumar Shukla
• Edith Cowan University
• Australia

Subjects:

• Materials Science
• Clean Tech
• Civil, Environmental and Geotechnical Engineering

1. Introduction

The manufacturing of cement is an unsustainably resource consuming and environmentally hazardous process. It is estimated that production of each tonne of cement is accompanied by emission of approximately an equal amount of carbon dioxide into the atmosphere (Environmental Protection Agency & U, 2010). Cement industry alone is responsible for 5 to 8% of the total worldwide carbon emissions (Miller et al., 2018; Osman et al., 2020) and it is the largest global energy consuming industry (Amrina & Vilsi, 2015; Usón et al., 2013). Partial replacement of cement by supplementary cementitious materials (SCMs) not only reduces the impact of above stated problems but also imparts durability and strength in concrete and cementitious mixtures. Most commonly used SCMs are fly ash and ground granulated blast furnace slag. These conventional SCMs cannot be relied on in future perspective as strict environmental protection policies will force the coal burning plants to convert into more eco-friendly technologies in future making fly ash a scarce material (Bellantuono, 2019; Le Billon & Kristoffersen, 2020). Blast furnace slag is a byproduct of steel production and not commonly available in some parts of the world. So, there is a need to investigate the possibility of alternate materials which can be utilized as SCMs. PWG is a potential candidate as a substitute to these conventionally used SCMs.

Recycling of waste glass to produce new glass products is a complex process. This complication is associated with the variability of chemical composition of glass due to its color. For quality control, glass needs to be sorted out based on its color and cleaned before recycling (Guo et al., 2020). Due to this difficulty in recycling, major part of the glass produced in the world is dumped in landfills leading to serious environmental concerns (Lu & Poon, 2019; Mohajerani et al., 2017; Hamada et al., 2012). Considering these factors, the best possible use of PWG is in cementitious mixtures where different color of waste glass provides a suitable chemical composition for use as a SCM.

Application of PWG in concrete started as a replacement of coarse and fine aggregates in concrete considering the environmental benefits of the practice (Rahman Sobuz et al., 2023). Although, the results of various research indicated an increase in compressive strength of concrete by using waste glass as aggregate but it was identified that high amount of silica in glass particles increases the chances of alkali-silica reaction in concrete (Khan et al., 2021; Yang et al., 2018; Rajabipour et al., 2010). Recently it has been reported that, by reducing the particle size of glass to a sub-micron level, the kinetics of ASR can be changed to pozzolanic reaction utilizing the alkalis in the formation of calcium silicate hydrate (C-S-H) gel before ASR gel can be formed (Fanijo et al., 2021; Zheng, 2016; Ke et al., 2018; Hendi et al., 2019; Bueno et al., 2020).

PWG is reported to have several beneficial effects on concrete and cementitious mixtures when used in partial replacement of cement. Incorporation of glass milled to particle size of 13 micron in recycled aggregate concrete resulted in a higher compressive and flexural strength at 56 days, lower voids ratio, better freeze-thaw resistance and lower ASR expansion compared to that made without glass powder (Soroushian, 2012; Nassar & Soroushian, 2011). Concrete mixtures produced with PWG have been observed to show better physicomechnical properties than normal concrete after exposure to elevated temperature of up to 900 °C (Al Saffar et al., 2012). Earlier studies have also reported beneficial effects of waste glass on the sulfate resistance of mortar mixtures produced with 10% replacement of cement with PWG (Tayeh et al., 2020). Densification of microstructure due to pozzolanic activity is the major contributing factor in enhancement of concrete behavior (Vaitkevičius et al., 2014; Schwarz & Neithalath, 2008). Microstructural analysis provides strong evidence in favor of later age pozzolanic reactivity of glass powder. An energy-dispersive X-ray spectroscopy (EDX) of cement paste with 20% milled waste glass indicates significantly higher amounts of silica present in the microstructure as compared to that of pure cement paste. A dense microstructure in a scanning electron microscope (SEM) analysis and decrease in calcium/silica ratio of glass incorporated cement pastes were observed indicating the occurrence of pozzolanic reaction (Nassar & Soroushian, 2012). Thermogravimetric analysis of cement pastes modified with vitreous calcium aluminosilicate from recycled fiber glass showed a decrease in calcium hydroxide (CH) amount after 7 days which is an indication of pozzolanic reaction (Neithalath et al., 2009). Flame emission spectroscopy revealed that glass powder releases very small amount of alkalis in pore solution of concrete thus preventing the early hydration of cement pastes and promoting the long term strength gain (Schwarz & Neithalath, 2008). Dissolution of PWG in mixing water before its use in concrete has significant positive effect on early age strength development. SiO₂, CaO and Na₂O are released during hydrolysis of glass powder and participate in hydration of cement forming more C-S-H. Using this technique, concrete cubes containing 10% and 30% glass powder showed higher compressive strength than control samples at 7 days (Elaqra et al., 2019). Shi et al. (2005) reported that pozzolanic reactivity increases with the decrease in particle size of the glass powder. Kamali and Ghahremaninezhad (2016). investigated the effect of fly ash, waste fiber glass powder and post-consumer PWG on hydration of cement. A decrease in rate of hydration was observed for samples modified with fly ash and glass powder as compared to those made with pure cement. On the contrary, another study witnessed a slight increased heat of hydration due to PWG incorporation. This increase was attributed partially to higher effective water content available for hydration and partially to pozzolanic reactivity (Mirzahosseini & Riding, 2014). Significant gains in mechanical properties and enhancement of microstructure of ultra-high performance geopolymer concrete (UHPGPC) mixtures have been reported by the researchers in react studies (Tahwia et al., 2012a, 2012b). Based on the outcome of the test results, authors have reported a 10% replacement of cement with PWG to be the optimum resulting into the highest compressive and flexural strengths of the resulting mixture. In another study Tahwia et al. (2012c) observed improvement in thermal stability of UHPGPC mixtures incorporating waste glass as partial replacement of sand with a slight reduction in compressive strength when 22.5% of sand was replaced with waste glass. In another study the researchers found that the use waste glass as partial replacement of sand in the production of geopolymer concrete for use in deep beams resulted into the reduction of compressive and flexural strengths of the resulting concrete mixtures. A 30% reduction in the load carrying capacity of the deep beams was recorded when sand was replaced with waste glass in the range of 5 - 15% (Taher et al., 2011).

In several research studies, milled waste glass was found to be effective in suppressing the expansion due to ASR. It is suggested that finely ground glass powder prevents the dissolution of silica present in the reactive aggregates by increasing the concentrations of aluminum in the pore solution thus preventing the deleterious ASR related expansion (Zheng, 2016). Cai, Y., Xuan, D. and Poon, C.S. observed that Na/Si and Ca/Si ratios of hydration products had a direct effect on mitigating ASR expansion. PWG increases these ratios which results in a more stiffer ASR gel and reduced osmotic pressure, ultimately mitigating the deleterious ASR expansion (Cai et al., 2019). Chemical composition of glass plays an important role in reducing the ASR expansion. Higher amount of glass stabilizer and formers delay the dissolution of silica in the pore solution thus reducing the ASR effects (Bignozzi et al., 2015). In mixtures where coarse glass is used as fine aggregate in concrete/mortar production, the probability of ASR occurrence is high. However, researchers have reported success in mitigating the ASR expansion through the use of finely ground industrial by products as partial replacement of cement in the relevant concrete/mortar mixtures. Duan et al. (2020) found that 20 wt.% replacement of cement with calcined and finely ground drinking water treatment sludge significantly enhanced the resulting mortar’s ASR resistance although the mixture was produced with 100% glass as fine aggregate. In a more recent study (Duan et al., 2012) the authors have reported promising ASR related performance of a ternary blended mortar mixture produced with 30 wt.% replacement of cement with a mixture of drinking water treatment sludge and ground granulated blast furnace slag.

2. Research significance

Although the use of PWG in cementitious mixtures has been investigated by earlier researchers. However, most of the studies provide limited collective information about the effects of PWG on rheology, setting time, heat of hydration, microstructure, and composition of the hydrated phases of the resulting cementitious mixtures. Furthermore, there is considerable variability in the findings reported in various studies. Hence, there is a need to further investigate the effect of PWG on such attributes of cementitious composites. Consequently, this paper aims to analyze the physical characteristics, reaction kinetics, morphology of the hydration products, ASR reactivity and microstructure of the PWG modified cement paste and mortar mixtures in a single study.

3. Materials and methodology

3.1. Materials

PWG having median particle size of 15 micron was obtained by ball milling of mixed color waste glass obtained from a local recycling facility. Type-I Ordinary Portland cement (OPC) fulfilling the requirements of ASTM C-150 was obtained from local market. Figure 1 shows the particle size distribution of PWG and OPC while Table 1 shows the chemical composition of the two materials. Some of the physical characteristics of PWG and OPC are presented in Table 2. An SEM image and EDX plot of the PWG are shown in Figure 2. Two types of sands; one normal (non-reactive) and the other with known reactive characteristics of causing ASR in cementitious mixtures were used for the production of mortar mixtures.

Figure 1. Particle size distribution curve of glass powder.

Figure 2. Morphology of PWG used in the experimental program: (a) SEM image; (b) EDX plot at the location shown.

Table 1. Chemical composition of cement and glass powder.

Compound	Cement (%)	Glass powder (%)
CaO	62.32	12
SiO2	20.98	69
Al2O3	5.06	7
Fe2O3	3.58	<1
Na2O3	–	10
MgO	2	1
SO3	2.31	–
Other oxides	–	<1
IR (insoluble residue)	0.73	–
Loss on ignition	2.91	–

Table 2. Physical characteristics of cement and glass powder.

Item	Cement	Glass powder
% passing # 325 Sieve	90.5%	97%
Specific gravity	3.12 g/cm^3	2.46 g/cm^3
Median particle size	14 µm	15 µm
Moisture content	0.05%	0.1%
Brightness	–	80%
Specific surface area	9500 cm^2/g	13100 cm^2/g

3.2. Methods

Cement pastes and mortar specimens incorporating 0, 10, 15 and 20 wt.% of PWG as partial replacement of cement were prepared according to the guidelines of ASTM C109. Table 3 provides mixture proportions of four cement paste and four mortar mixtures, one each produced with pure cement as a binder to serve as control mixtures. Fresh cementitious paste specimens with a base diameter of 70 mm, top diameter of 60 mm and height of 40 mm were prepared to investigate the effect of PWG on the setting time of cement paste according to ASTM C191-21. The effect of PWG on mortar flow was investigated following the methodology of ASTM C1437-20. A tamped mortar layer of 25 mm thickness was given 25 drops within 15 seconds on a flow table. The increase in the base diameter of mortar layer was measured as the flow measurement of the cementitious mixture. Cubic mortar specimens of 50 mm size (Figure 3a) prepared according to the protocol of ASTM C109 were tested for compressive strength (Figure 3b) and water sorption as per the practice of earlier researchers (Islam et al., 2017). Similar samples with cement paste were also prepared to compare the compressive strength of pure cement paste with that of PWG-cement mixtures. All paste mixtures were produced with water to cementitious (w/cm) ratio of 0.45, while for mortar mixtures w/cm ratio was kept at 0.48. All specimens were cured in lime saturated water at 23 ± 1 °C until the test age.

Figure 3. View of mortar cubes (a) Compression testing of cube specimen (b).

Table 3. Mix proportions of cement paste and mortar.

Mix designation	Cement (%)	Glass powder (%)	cm: sand	w/cm
Cement/glass paste mixtures
C100G0	100	0	–	0.45
C90G10	90	10	–	0.45
C85G15	85	15	–	0.45
C80G20	80	20	–	0.45
Cement/glass mortar mixtures
M100G0	100	0	1: 2.75	0.48
M90G10	90	10	1: 2.75	0.48
M85G15	85	15	1: 2.75	0.48
M80G20	80	20	1: 2.75	0.48

Notes: C = cement, G = glass powder, M = mortar, cm = cementitious materials.

According the guidelines of ASTM C1679, the evolution of heat of hydration of PWG and control mixtures were measured using a TAM air calorimeter device following the method of earlier studies (Alzeer et al., 2012). About 15 g of paste of each type of mixture was prepared in a calorimetric cup and immediately transferred to a calorimeter. The rate of hydration heat and cumulative hydration heats measurements were recorded over a period of 100 hrs. after mixing at a temperature of 23 °C.

Sorption which refers to the absorption of water by capillary pores and its subsequent transport inside cementitious matrix is an important characteristic of cementitious mixtures which significantly affects their durability. Various deleterious ions, while using water as a vehicle penetrate these mixtures and cause durability problems. In this work, sorption of mortar mixtures was measured as per the guidelines of ASTM C1585. Cubic mortar specimens upon completion of 28 days curing period were first oven dried at 45 °C for 15 days to achieve constant mass. Oven-dried specimens were then sealed on sides and top with epoxy and subjected to sorption test continuously for 8 days according to the method outlines in ASTM C1585. As per this method, sorption is obtained in the units of length (mm) by dividing the change in mass by the product of cross-sectional area of specimen and density of water.

SEM analysis of 28 days old unaged mortar specimens with and without PWG were carried out to compare their microstructures. Similarly, to assess the ASR reactivity of glass in the mixture, 28 days old cubic mortar specimens of M80G20 mixture prepared with reactive sand obtained from a known quarry were first aged in NaOH solution at 80 °C. Specimens of M80G20 mixture prepared with normal (nonreactive) sand were also subjected to similar aging condition for comparison. Subsequently, SEM analysis of the fractured specimens were carried out after 7- and 20-days exposure to NaOH solution to assess the potential microstructural damage caused by ASR.

A PANalytical XPert PRO X-ray diffractometer (λ = 1.54 A^o) was used to carry out the X-ray diffraction (XRD) analysis of the powdered mortar mixtures. At the age of 35 days, mortar specimens with and without PWG were reduced to fine powder. All scans were carried out at diffraction angles of 2θ ranging from 0° to 100°. The step size was set at 2θ = 0.0170°

4. Results and discussion

4.1. Flow and setting time

Figure 4 shows the results of the flow test of control and PWG based mortar mixtures. These results show that as the percent replacement of cement with PWG increases, the flow of the mortar mixture increases. A 15% increase in the flow is recorded when 20% of OPC is replaced with PWG. This trend can be attributed to the increase in effective w/cm ratio of the mixture with an increase in the PWG content noting that PWG particles have a significantly lower water absorption compared to cement particles. The smoothness of the PWG particles further enhance this effect. Although the morphology of the PWG particles (Figure 2) does have the effect of reducing the flow of relevant mixtures, however the two opposing effects of increase in effective w/cm ratio and surface smoothness of PWG particles dominate this effect and the net effect appears in the form of increase in flow with increase in PWG content. These observations are in line with the findings of earlier researchers (Ramadan et al., 2020; Jiang et al., 2022; Schwarz et al., 2007). Figure 5 shows that both; the initial and final setting times of PWG paste mixtures increase with an increase in the percent replacement of OPC with PWG. The initial setting time is seen to increase by about 5%, 17%, and 24% when OPC is replaced by 10%, 15%, and 20% with PWG. Similarly, the final setting time is increased by about 6%, 9%, and 12% when OPC is replaced by 10%, 15%, and 20% with PWG. It is noted that the addition of PWG influences the initial setting time more than the final setting time. The increase in setting time of PWG-pastes may be attributed to the availability of more water (higher effective w/cm ratio) in the mixtures owing to the less water absorption of glass particles compared to that of cement particles. Furthermore, the dilution effect of PWG and its low reactivity at the early age of hydration causes delayed hydration of cement in the PWG mixtures. Similar effects of PWG on the setting time of cementitious mixtures have been reported by other researchers (Patel et al., 2019; Lu et al., 2021; Liu et al., 2019). It is to be mentioned that the reduced effect of PWG on the final setting time may be due to the offsetting effect caused by the condensation of paste brought about by the incorporation of finer PWG particles with the passage of time.

Figure 4. Flow test results of control and PWG mortar mixtures.

Figure 5. Setting time of cement paste with and without glass powder.

4.2. Rate of heat of hydration and cumulative heat of hydration

Figure 6 illustrates the rate of heat of hydration and cumulative heat of hydration per gram of cement of control and PWG paste mixtures. All the mixtures (including control) exhibit single peak flow of heat of hydration but with different intensities. The single peak of heat of hydration is believed to be the result of conversion of tricalcium silicate (C₃S) to C-S-H and CH and the formation of ettringite as a result of the hydration of tricalcium aluminate (C₃A) (Mehta & Paulo, 2014; Neville, 2006). Figure 6(a) shows that the PWG mixtures show lower heat of hydration at all times as compared to that of control (pure cement) mixture. At peak value of heat of hydration evolution, the rate of heat flow of C80G20, C85G15, and C90G10 mixtures are 29%, 14%, and 3%, respectively lower than the corresponding heat flow of the control mixture. There may be three possible reasons for this trend: (i) The inclusion of PWG as partial replacement of cement results in reduction of C₃A and C₃S contents of the unhydrated binder content which results in lower rate of heat evolution, noting that the two compounds account for about 77% of the heat of hydration within the first 72 hrs. of hydration (Mehta & Paulo, 2014); (ii) the pozzolanic reaction of glass with the hydrates of cement is a slow process compared to the hydration of cement and is delayed until sufficient quantity of Portlandite is available in the pore solution for its occurrence, hence the presence of PWG in the mixture results in slowing down the overall hydration which consequently causes reduction in hydration heat evolution; and (iii) the dilution effect brought about by the addition of PWG in the mixtures. Similar findings about the effect of PWG on the rate of heat of hydration have been reported in the literature (Lu et al., 2021; Liu et al., 2019; Nahi et al., 2020).

Figure 6. Heat of hydration test results: (a) rate of heat evolved (b) cumulative heat of hydration.

Figure 6(b) shows that at the cut off age of 100 hrs. of mixtures age, the C90G10, C85G15, and C80G20 mixtures showed 10%, 16%, and 25%, reduction in cumulative heat of hydration, respectively when compared with that of control mixture. This effect can be explained by the comparison of Figures 5 and 6. The addition of PWG results in dilution of the mixtures, that is a reduction in the cement content, which consequently reduces the rate of heat of hydration (and cumulative heat of hydration) and slows down the hydration process until the pozzolanic reaction of PWG sets in with CH after release of its sufficient quantity in the pore solution. Thus, a corresponding increase in the setting time of the PWG mixtures is recorded. Earlier studies have reported similar findings with regards to the cumulative heat of hydration of cementitious mixtures incorporating SCMs (Li et al., 2020, 2022). Test results presented in Figure 6 show a 15% reduction in the heat of hydration at 7 days of the cementitious mixture age at 20% of cement replacement. The value is closely comparable to the 12% reduction in heat of hydration reported by Chang et al. (2015). The 3% difference between the two studies may be attributed to the extra fine PWG used Chang et al.

4.3. Compressive strength of cementitious paste and mortar

Compressive strength test results of paste mixtures at various ages with and without PWG are shown in Figure 7, while Figure 8 presents similar results for the mortar mixtures. Both figures show a steady increase in compressive strength with curing age. It can be seen in Figure 7 that at early ages (3 and 7 days) compressive strength of paste mixtures incorporating PWG is less than that of control mixture. At 7 days of mixtures age, the compressive strength of C80G20 mixture is about 17% less than that of C100G0 (control) mixture. However, at 28 days of mixtures age this trend is reversed and the C80G20 mixture is about 10% higher than that of control mixture. At the age of 90 days this difference rises to about 19% in favor of C80G20 mixture. Mortar mixtures (Figure 8) show similar trends to that of paste mixtures. At 28 days of mixtures age, M80G20 mixture shows about 14% higher compressive strength than that of control mixture (M100G0). Whereas, at 90 days of mixtures age, the compressive strength of M80G20, M85G15, and M90G10 mixtures are 24%, 16%, and 5%, respectively, higher than control mixture. The lower initial compressive strength of PWG containing paste and mortar mixtures may be attributed to: (i) the higher effective w/cm ratio of these mixtures; and (ii) dilution effect of the PWG in these mixtures. However, at later ages (28 days and after) the pozzolanic reaction of PWG is believed to reverse the equation in favor of PWG-based mixtures which show a higher compressive strength compared to the relevant control mixtures. The pozzolanic reaction of PWG with hydrates of cement (chiefly Portlandite), is brought about by the dissolution of silica and alumina in the pore solution after the glass structure breaks down in the highly alkaline environment of cementitious mixtures (Zheng, 2016; Taylor, 1997). Through its pozzolanic reaction, the PWG not only imparts densification at the microstructural level that results into strength enhancement, but it also refines pore system of the relevant mixtures. Findings of this wok with regards to the increase in compressive strength of PWG-based mixtures are in conformity with the results of earlier researchers (Liu et al., 2019; Khmiri et al., 2013; Carsana et al., 2014).

Figure 7. Compressive strength of cementitious paste mixtures.

Figure 8. Compressive strength of mortar mixtures.

4.4. Water sorption

Sorption test results for control and PWG containing mixtures are displayed in Figures 9 and 10. While Figure 9 shows the time^1/2 vs sorption plots for all mixtures, Figure 10 shows the 8 days cumulative sorption for these mixtures. Figure 9 indicates significantly less water sorption of PWG-mortar mixtures as compared to control mixture. This difference increases with an increase in the curing age of mixtures. Plots of Figure 10 show that after 8 days of continuous exposure to water, the M80G20 mixture shows about 32% less cumulative sorption as compared to the control mixture (M100G0). In the case of M85G15 and M90G10, this reduction is 27% and 18%, respectively. Reduction in sorption of PWG containing mixtures is thought to be the result of (i) filler effect of PWG and (ii) pozzolanic reaction of PWG with the hydrates of cement resulting into pore filling and pore refinement effect of the relevant mixtures which cause reduction in their porosity with consequent reduction in their sorption attribute (Nahi et al., 2020). The sorption test results of this study correlate well with that of compressive strength test results, pointing at the microstructure enhancement of the mixtures caused by the inclusion of PWG. Lu et al. (Lu et al., 2019) concluded significant reduction in water absorption when 20% of cement is replaced with glass powder in concrete mixture which is used for the production of precast members. Similarly, Letelier et al. (Letelier et al., 2019) reported reduction in the coefficient of capillarity of mortar mixtures incorporating glass powder. They concluded that the beneficial effect towards reduction in the water absorption characteristics of relevant mixtures is caused by the filler effect of the 38 µm size glass particles.

Figure 9. Sorption test results of mortar mixtures.

Figure 10. 8 days cumulative sorption of mortar mixtures.

4.5. Microstructural analysis of aged and unaged mortar specimens

Figures 11(b) and 12(b) show the EDX plots of control mortar (M100G0) and the one produced with 20 wt.% replacement of cement with PWG (M80G20) at the locations shown in Figures 11(a) and 12(a), respectively. Comparison of the two EDX plots show that the addition of PWG has marked effect on the composition of the relevant mixture. An enlarged Si peak and appearance of Na peak in the EDX plot of M80G20 mixture is seen. It is noted that calcium oxide to silicon oxide (C/S) ratio which is 12.73 in case of control (M100G0) mixture, drops to 4.22 for M80G20 mixture, which is a typical indication of the occurrence of secondary pozzolanic reaction resulting into microstructural enhancement with a consequent increase in strength characteristics (Mindess et al., 2012). Earlier studies have reported similar findings (He et al., 2019; Li et al., 2012).

Figure 11. (a) SEM micrograph of M100G0 mixture (b) EDX spectrum at the identified location.

Figure 12. (a) SEM micrograph of M80G20 mixture (b) EDX spectrum at the identified location.

Figure 13 shows the SEM image of M100G0 mixture while Figure 14 shows the SEM image of M80G20 mixture. The heterogenous microstructure of M100G0 is seen to comprise of large crystals of CH, microcracks, scattered capillary pores and significant amount of Ettringite needles. The microstructure of M100G0 mixture is altered to a dense, continuous, less heterogenous, and refined porosity microstructure of M80G20 mixture as a result of the incorporation of PWG. Comparison of the two micrographs show that the incorporation of PWG results into pore filling effect, significant reduction Portlandite content, and elimination of cracks from the microstructure of the cementitious mixture. These effects are attributed to the pozzolanic reaction of PWG (as evidenced in Figures 10 & 11) resulting into formation of additional C-S-H which enhances the microstructure of the cementitious mixture. Furthermore, pore filling effect of the unreacted PWG particles also results into densification of the microstructure of the cementitious matrix. The observations of Figures 13 and 14 verify the narrative of Section 4.3, that is the improvement in microstructure brought about by the inclusion of PWG improves the macrolevel mechanical characteristics of the cementitious mixtures. Jiang et al. (2022) reported the formation of C-S-H rim around glass particle when PWG was used an SCM in cement paste mixture. Authors further observed continuous decrease in CH content with an increase in PWG content in the mixture, confirming the occurrence of pozzolanic reaction. Zheng (2016). found that the reduction in the portlandite level over time is an indication of occurrence of pozzolanic reaction of glass with the hydrates of cement. The author further reported the production of two types of C-S-H by the glass; the inner product formed within the original boundary of glass grains and the precipitated outer product.

Figure 13. SEM micrograph of M100G0 mortar mixture.

Figure 14. SEM micrograph of M80G20 mortar mixture.

Figures 15 and 16 present the SEM images of M80G20 mixture produced with normal (nonreactive) sand and M80G20 mixture produced with reactive sand known to cause ASR, respectively, after aging in NaOH solution at 80 °C for 6 and 20 days. While Figures 15(a) and 16(a) show the microstructure of the two types of mixtures after 6 days of aging, Figures 15(b) and 16(b) show similar images after 20 days of aging in NaOH solution. Comparison of the two figures show no signs of the occurrence of the ASR, that is there is no evidence of formation of deleterious ASR gel. These figures suggest that up to 20 wt.% replacement of cement with PWG does not possibly release sufficient amount of alkalis to trigger the ASR in presence of reactive fine aggregate (Shayan & Xu, 2004). Another reason could be occurrence of the pozzolanic reaction which results in suppressing the deleterious ASR, given that pozzolanic reaction is faster than ASR (Taylor, 1997). Zheng (2016) has reported that glass grains mitigate the ASR due to the increase in the Al concentration in the pore solution which is brought about by the inclusion of PWG in the mixture. Fanijo et al. (2021) observed that the presence of large size (> 300 µm) particles of PWG may result into the formation of ASR gel, while smaller particles only undergo pozzolanic reaction, thereby nullifying the occurrence of ASR.

Figure 15. SEM micrograph of M80G20 mortar mixture produced with normal sand: (a) after 7 days of exposure to NaOH solution; (b) after 20 days of exposure.

Figure 16. SEM micrograph of M80G20 mortar mixture produced with reactive sand: (a) after 7 days of exposure to NaOH solution; (b) after 20 days of exposure.

4.6. XRD analysis of mortar specimens with and without PWG

The XRD patterns of all mortar mixtures are shown in Figure 17. Characteristic peaks of various types of hydration products are shown in these spectra. Figure 17(a) (XRD pattern of M100G0 mixture) shows peaks of CH and unreacted C₃S besides peaks of gypsum and SiO₂ (quartz). It also shows a minor peak of C-S-H. Figure 17(a) to (d) show that the replacement of cement with PWG results in the reduction of CH peaks with corresponding increase in the C-S-H peaks. Besides, Ettringite peak appears in the spectra of M85G15 and M80G20 which may be attributed to the increase in the Al content in the mixture caused by the inclusion of PWG. These observations are in conformity with the findings described in the Sections 4.3 and 4.5 above. That is the pozzolanic reaction of PWG results into production of additional C-S-H compound by consuming CH, which results into enhancement of microstructure of corresponding mixtures with consequent increase in compressive strength and enhanced moisture barrier characteristics.

Figure 17. XRD patterns of 35 days old mortar mixtures with and without PWG.

5. Conclusions

This study investigated the effects of PWG as partial cement replacement in paste and mortar mixtures on rheological, mechanical, hydration kinetics and microstructural properties. Following conclusions were drawn based on results of the experimental investigations.

• Flow of the mortar mixtures has a proportional increase with PWG content as partial replacement of cement. At 20 wt.% replacement of cement with PWG, 15% increase in the flow is recorded. Use of PWG has a delaying effect on the setting time of cementitious paste. At 20. wt% replacement of cement with PWG the initial and final setting times were increased by 24% and 12%, respectively as compared to that of control mixture. These effects are believed to favor mortar and concrete placement in hot weathering conditions.

• Significant reduction in the rate of heat of hydration and cumulative heat of hydration is recorded for PWG mixtures. At 20 wt.% replacement of cement with PWG, the rate of heat flow reduces by 29% and cumulative heat of hydration reduces by 25% when compared to the control mortar mixture. Thus, the incorporation of PWG is seen to slow down hydration process of cementitious mixtures. Concrete mixture incorporating PWG with lower heat of hydration is believed to very suitable for use in mass concrete pours where the rise in temperature is a significant issue.

• The later age (post 28 days) compressive strength of PWG incorporated paste and mortar mixtures show higher strength than that of control mixtures. At the age of 28 days, the PWG mixtures show comparable compressive strength to that of control mixtures. At 90 days of age the blended paste and mortar mixtures with 20 wt.% PWG show 19% and 24% higher compressive strength than that of corresponding control mixtures. The increase in later age strength of the blended mixtures is due to the pozzolanic reaction of PWG.

• SEM analyses indicate significant improvement in microstructure of mortar mixture incorporating 20 wt.% PWG as compared to that of control mixture. The densification and refinement in the microstructure are brought about by the conversion of CH into secondary C-S-H through the pozzolanic reaction of PWG with the hydrates of cement. The XRD and EDX analyses confirm the pozzolanic reaction of PWG. The enhancement in microstructure consequently results into 32% reduction in the 8 days cumulative water sorption of 20% PWG incorporated mixture.

• SEM analyses of the microstructure of PWG mixture produced with reactive sand and aged in NaOH solution at 80 °C for 20 days shows no signs of ASR. The microstructure of these mixtures is seen to be stable and similar to that of related mixtures produced with nonreactive sand. Based on this analysis it is concluded that up to 20 wt.% replacement of cement with PWG, is innocuous from the standpoint of ASR.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research project was funded by the American University of Ras Al Khaimah (AURAK) through Seed Grant No. ENGR/003/19. Support provided by AURAK Office of Research and Community Service is gratefully acknowledged.

Notes on contributors

Roz-Ud-Din Nassar

Roz-Ud-Din Nassar is an Associate Professor in the Department of Civil and Infrastructure Engineering at American University of Ras Al Khaimah, United Arab Emirates. His research interests include the study of strength and durability characteristics of sustainable cementitious materials and long-term field testing of various types of concrete.

Danish Saeed

Danish Saeed is a Lecturer of Civil Engineering at Khwaja Fareed UEIT Rahim Yar Khan, Pakistan. Study of the use of recycled materials in concrete is his research area.

Tewodros Ghebrab

Tewodros Ghebrab is an Assistant Professor in the Department of Civil, Environmental, and Construction Engineering at Texas Tech University, Lubbock, USA. His research interest includes the study of performance of cement-based materials under adverse environments and the effect of mineral/chemical admixtures on the structure-property relationships of cement-based materials.

Shah Room

Shah Room is a Lecturer in the Department of Civil Engineering and Technology, University of Technology Nowshera, Pakistan. His research interest comprises of the study of environmentally-friendly cementitious mixtures under exposure to aggressive environment.

Ahmed Deifalla

Ahmed Deifalla is a Professor n the Department of Structural Engineering and Construction Management, Future University in Egypt, New Cairo City 11835, Egypt. His research areas comprise of the study of lightweight concrete, high strength concrete, fiber reinforced concrete, and the strengthening of reinforced concrete elements subjected to variety of loadings.

Kadhim Al Amara

Kadhim Al Amara - Assistant Professor Department Mechanical and Industrial Engineering, Liwa College of Technology, Abu Dhabi, United Arab Emirates. His research interests include the study of thermal spray coating and surface technologies, additive manufacturing, and sustainable energy.