Engineering properties and SEM analysis of eco-friendly geopolymer mortar produced with crumb rubber

2.1. Materials

Fly ash conforming to ASTM C311 [21] and regarded as class F according to ASTM C618 [22] was employed as an aluminosilicate-rich raw material in the production of geopolymer mortars. The chemical compositions and physical properties of fly ash, which was procured from the Çatalağzı thermal power plant in Zonguldak province of Turkey, are presented in Table 1. In addition, the X-ray diffraction (XRD) analysis result of fly ash is indicated in Figure 1a, and the image of its particles taken by scanning electron microscope (SEM) is demonstrated in

Figure 1b. The main themes presented in these figures are components constituting the fly ash and particle sizes and shapes of fly ash. In regard to Figure 1a, it can be stated that the peak point in the intensity occurred at quartz crystal since the fly ash highly consists of SiO2. On other hand, it can be seen in Figure 1b that the fly ash consists of generally particles smaller than 3 μm. Also, it should be emphasized that the particles of the FA are in a spherical shape. As an alkaline activator, NaOH and Na2SiO3 pair with a ratio of 1-to-2 was used. The chemical compositions and properties of NaOH in spherical pellet form in white color and Na2SiO3 in the liquid form in light yellow color

J Sustain Const Mater Technol, Vol. 7, Issue. 2, pp. 95–107, June 2022

Figure 2. River sand and crumb rubber: (a) general photographic views, (b) photographic view of particles, (c) SEM images, and (d) sieve analysis results. are presented in Tables 2 and 3, respectively. The NaOH pellets used to prepare the NaOH solution with a 12-M concentration had a purity of more than 98%. In addition to this, Na2SiO3 solution comprising of liquid and solid materials with respectively roughly 61.5% and 38.5% had a silica-to-sodium oxide ratio (SiO2/Na2O) of about 2.5. As the last material, the superplasticizer with a commercial name of MGlenium 51 having a specific gravity of

1.07. was used to achieve geopolymer mixtures with sufficient flowability.

In the production of the geopolymer mortars of the present study, river sand (RS) with a maximum particle size of 4 mm and specific gravity of 2.74 was used. However, the waste tire rubber aggregate named crumb rubber (CR) having a specific gravity of 0.55 was substituted in some mixtures with river sand. The general photographic views of these two fine aggregates are given in Figure 2a whereas their particle shapes and sizes are shown in Figure 2b and their SEM images are indicated in Figure 2c. In addition, sieve analysis results (determined in accordance with ASTM C136 [23]) of RS and CR compared with the upper and lower limits for fine aggregate (proposed by ASTM C33 [24]) are shown in Figure 2d. The fineness

2.06. On the other hand, the water content and absorption

values of RS were respectively about 0.94 and 2.31, whereas that of the CR were 0 which means the CR used in this study did not contain water and had no water absorption capability. As can be seen in Figures 2a and 2b, the CR consists of flaky and elongated particles. As the size of the particles decreases, the CR particles become angular and/ or partly rounded. Besides, the CR particles are substantially softer than the RS particles, thus the elastic modulus of CR is less than that of the RS, which makes such materials structural unstable material. However, such characteristics may be useful for absorbing the energy such as impact energy, sound energy, heat energy, etc.

2.2. Mixture Proportions and Production

The mixtures in the present study were designed at a Na2SiO3-to-NaOH ratio of 2.0 and alkaline activator-to-aluminosilicate-rich raw material ratio of 0.5. The alkaline activator content and powder material (aluminosilicate-rich raw material) dosage were designated as 300 kg/m3 and 600 kg/m3, respectively. The superplasticizer content for the control mixture with sufficient flowability

J Sustain Const Mater Technol, Vol. 7, Issue. 2, pp. 95–107, June 2022

Table 4. Mixture proportions for the rubberized geopolymer mortars, kg/m3

FA: Fly ash; NH: Sodium hydroxide; NS: Sodium silicate; RS: River sand; CRSL: Crumb rubber substitution level; CR: crumb rubber; SP: superplasticizer.

was determined as 3.5% of powder material by mass after many trial batches. Since the rheological behavior of such materials is significantly influenced by variation in the superplasticizer content, all the mixtures in the present study were produced at a superplasticizer content of 3.5%. In the production of the plain geopolymer mixture, only the RS was used as fine aggregate; however, in the production of the geopolymer mixture containing crumb rubber, the RS was partially replaced with the CR at the levels of 10%, 20%, 30%, 40%, and 50% by volume. In this way, 6 geopolymer mortar mixtures were produced in total. Table 4 presents the detailed proportions of the geopolymer mortar ingredients. Alkaline activator-to-fly ash ratio (Na2SiO3-toNaOH), alkali activator molarity, fly ash dosage, and alkali activator content selected within the scope of the study carried out by Ekmen et al. [25], which is one of the most comprehensive studies in this context, in which the effects of the above-mentioned parameters on the fresh and hardened properties of geopolymer mortars were taken into consideration. Ekmen et al. [25] investigated the fresh and hardened properties of fly ash-based geopolymer mortars with 2 activator-to-fly ash ratios, 3 Na2SiO3to-NaOH ratios, and 3 alkali activator molarities. Considering the consistency and strength findings obtained from this study, the mixture proportions in the current study were determined. The same mixing procedure was applied to all geopolymer mortars during the production. This procedure consisted of two states: in the first stage, the alkaline activator solution was prepared and in the second stage, the geopolymer mortar was produced. First, the NaOH solution with 12-M concentration was prepared, and then, it was mixed with Na2SiO3 solution before almost 24-h of the beginning of the second stage. This solution was kept in a beaker till it was used in the production of geopolymer mortars. The second stage of the production procedure began by mixing the alkaline activator solution with fly ash in a mixer for about 3 minutes. Subsequently, the superplasticizer was poured into the mixer, and it was allowed to revolve for 2 minutes more. Thereafter, the RS in the plain mortar mixture and RS and CR mixture in the other mortars were gradually added to the mixer,

and it was permitted to rotate the mixer extra 3 minutes after all fine aggregates were poured. Thereby, the production process of the geopolymer mortars was completed. The fresh-state properties of the geopolymer mortars were determined in terms of flowability measured by the flow table test, and fresh unit weight once the production process finished. After the fresh-state properties were determined, the mixtures were poured into the steel molds in two layers, of each which was vibrated for 30 seconds using a vibrating table. A heat curing of 60 °C for 24 hours was applied to the specimens taken and then, the demoulded specimens were kept in the laboratory condition where the temperature was 23±2 °C for the following 2 days. The specimens in the steel molds were put into the oven to be exposed to the heat curing regime and all the specimens were covered with a plastic bag to prevent water evaporation in the alkali solution during the heat curing. After the 3-day curing regime, the hardened-state tests were performed. Herein, it should be noted that one of the most important issues to be resolved regarding the applicability of geopolymer mortars is the high-heat curing required for the geopolymerization process and its application time. The main purpose of recent studies is to manufacture geopolymer mortar or concrete at lower curing temperatures as much as possible with application time. In terms of energy efficiency, 60°C as the curing temperature and 24 hours as the curing duration were selected according to research in the literature on the issue.

2.3. Testing Methods

The test apparatus shown in Figure 3a was employed to measure the flowability of the geopolymer mortar mixtures and ASTM C1437 [26] was followed during the application of the flow table test. During the determination of both fresh and hardened (1-day and 3-day) unit weights of the mixtures, ASTM C138 [27] was followed. A closed-loop testing machine shown in Figure 3b was used to perform the 3-day flexural strength test in regard to the ASTM C348 [28]. The 3-day compressive strength of the geopolymer mortar mixtures was determined per ASTM C109 [29] and ASTM C597 [30] was followed during measuring the 3-day UPV values of the mixtures as typically indicated in Figure 3c.

J Sustain Const Mater Technol, Vol. 7, Issue. 2, pp. 95–107, June 2022

Figure 3. (a) Test apparatus of flow table test and measuring the flow diameter, (b) test device used in the measurement of flexural strength, and (c) UPV test device.

3.1. Flowability Results

Based on the flow table test, the flowability of the mortars can be described in terms of mm or percentage. It should be noted that when the flowability is presented in mm the perfect flowability is 200 mm while when it is given in percentage this 200-mm flowability is defined as 100% flowability. The plain geopolymer mortar produced in this study had a flowing diameter of 210 mm which means 110% flowability. On the other hand, it was observed that substituting the RS with CR did not influence the flowability of geopolymer mortars. All the mixtures produced in the current study had a flowing diameter of 210 mm (namely 110% flowability). Since the CR particles have almost no water absorption capability, incorporating the CR into the geopolymer mortar did not affect the flowability characteristics of the mixtures. Besides, the CR particles that can be considered flaky and elongated may play a blockage role during the flowing but since the table is rammed 25 times in 15 seconds during performing the test, this role of the CR particles may have been minimized and/or eliminated. However, it should be stated that this finding is not consistent with the results reported in the literature. For example, when Aslani et al. [16] increased the crumb rubber replacement level from 10% to 20%, they observed a small decrease in the slump flow diameter values. In addition, Wongsa et al. [19] investigated how the crumb rubber incorporation influences the flowability, mechanical, and thermal properties of geopolymer mortar. They found that adding crumb rubber to geopolymer mortar mixtures made them less workable. On the other hand, Moghaddam et al. [20] reported a minor improvement in the flowability of geopolymer concrete after replacing natural sand with spherical crumb rubber derived from old tires. Similarly, Zhong et al. [31] achieved an enhancement in the workability of the geopolymer mortar from a general perspective. For this reason, the findings presented in the current study will contribute to the literature in this context.

Figure 4. The variation in both fresh and dry unit weights of rubberized geopolymer mortars

3.2. Unit Weight Results

The change in both fresh and dry unit weight of the geopolymer mortars with respect to the CR substitution level is indicated in Figure 4. The fresh unit weight of the plain geopolymer mortar was about 2228 kg/m3, whereas a systematic decrease in the fresh unit weight was observed by gradually substituting the RS with the CR. The fresh unit weights ranging between 2195 kg/m3 and 1924 kg/m3 were obtained for the geopolymer mortars containing CR. As expected, the CR incorporation into the geopolymer mortar led to a decrease in the unit weight of the mixtures because of its lighter weight than the RS. The highest decrease of 14% was observed when the RS was substituted with the CR at 50%. The drying of the geopolymer mortar mixtures resulted in weight loss as is expected. There were decreases to be ranging from about 1.0% to 5.0% in the unit weights of the mortars after 1 day. On the other hand, after 3 days, re-

J Sustain Const Mater Technol, Vol. 7, Issue. 2, pp. 95–107, June 2022

Figure 5. The variations in (a) flexural strength and (b) compressive strength of rubberized geopolymer mortars. ductions between about 2.5% and 6.5% in the unit weights were observed. The lowest unit weight after the 3 days was obtained in the mixture produced with 50% CR as about 1820 kg/m3 and on the same day, the control mixture had a unit weight of 2172 kg/m3. TS EN 206-1 [32] regards the concrete having a unit weight (but oven-dry) of less than 2000 kg/m3 and more than 800 kg/m3 as lightweight concrete. Similarly, ACI Committee 213R-03 [33] classified the concretes as lightweight concrete when their unit weight (but air-dry) is less than 1950 kg/m3. Since the geopolymer concretes (or mortars) have not been covered by any standard yet, these criteria can be taken into consideration to describe the geopolymer concretes (or mortars) based on the unit weight. In this context, it can be stated that the geopolymer mortar mixtures produced with 30%, 40%, and 50% CR can be definitely considered the lightweight geopolymer mortar when the 3-day unit weights were taken into consideration. Aslani et al. [16] and Wongsa et al. [19] characterized the geopolymer mortars incorporating crumb rubber manufactured in their studies as lightweight. Azmi et al. [34] reported a systematical reduction in the unit weight of the geopolymer concretes when the crumb rubber replacement level gradually increased from 0% to 20%. Besides, it should be stated that there are a limited number of studies in which the unit weight of the geopolymer mortar incorporating crumb rubber was investigated.

3.3. Flexural and Compressive Strengths

The variations occurring in flexural and compressive strengths of the geopolymer mortars because of the CR

incorporation are shown in Figures 5a and 5b, respectively. The 3-day average flexural strength value of the control geopolymer mortar was about 5.41 MPa, and there was observed to be a systematical decrease in the average flexural strength of the mortars by substituting the RS with the CR and increasing the substitution level. The lowest 3-day average flexural strength value of 2.42 MPa was observed in the geopolymer mortar produced with a 50% CR substitution level. The reduction rates in the 3-day average flexural strength due to the CR incorporation were about 14%, 23%, 34%, 43%, and 55% when the substitution levels were 10%, 20%, 30%, 40%, and 50%, respectively. Wongsa et al. [19] investigated the properties of two types of geopolymer mortars: one was fully produced with RS and the other one was fully produced with CR. They reported a more than 75% reduction in the 28-day flexural strength when the RS was fully replaced with the CR. However, Zhong et al. [31] reported a slight increase in the flexural strength of the geopolymer mortar when the RS was replaced with the CR at the substitution level of 5% and a decrease after this substitution level. Similarly, the compressive strength results revealed that incorporating the CR into the geopolymer mortar led to a decrease in the compressive strength. Also, increasing the CR substitution level yielded a systematic decrease in the compressive strength, see Figure 5b. The 3-day average compressive strength of the control geopolymer mortar was about 35.5% and incorporating 10% CR into the geopolymer mortar resulted in a nearly 18% decrease in the 3-day average compressive strength. Increasing the CR

J Sustain Const Mater Technol, Vol. 7, Issue. 2, pp. 95–107, June 2022

Figure 6. Correlation between (a) flexural and compressive strengths and (b) dry unit weight and compressive strength of rubberized geopolymer mortars. substitution level from 10% to 50% diminished the compressive strength to 11.2 MPa. The results achieved from the study herein comply with the limited number of studies reported in the literature. For example, both Wongsa et al. [19] and Zhong et al. [31] reported a decrease in the 7-day and 28-day compressive strengths of the geopolymer mortar by substituting the natural sand with CR. In a similar manner, Azmi et al. [34] observed a nearly 65–75% decrease in the 7-day and 28-day compressive strengths at the CR substitution level of 20%. One of the main reasons for the decrease in the strengths (both compressive and flexural) is that the interfacial transition zone occurring between geopolymer paste and rubber aggregate is larger than that occurring between the natural aggregate. Besides, the soft structure of the rubber aggregate will exhibit higher strain performance under loading compared to hardened geopolymer paste and natural aggregate. This means that in places where rubber aggregate is present, hardened geopolymer paste and/or natural aggregates will carry the most of the load, and the rubber aggregate will participate in the load-bearing role at even high strain levels, but at these strain levels, the hardened geopolymer paste has already cracked and the integrity of the mortar has deteriorated. Therefore, incorporating the rubber aggregate into the geopolymer mortar negatively influences its compressive strength. Moreover, in Figure 6a, the correlation between the flexural and compressive strengths of the geopolymer mortars produced in the current study is presented. By taking

into consideration the coefficient of determination (R2) value given in this figure, it can be stated that there is a strong relationship and correlation between the flexural and compressive strengths of the geopolymer mortars of this study. A similar correlation was observed between the dry unit weight and compressive strength of the geopolymer mortars produced in this study as shown in Figure 6b. When the R2 value given in this figure is considered, it can be stated that there is a directly proportional and strong relationship between the 3-day dry unit weight and compressive strength values.

3.4. UPV Values

The change in the UPV values of the geopolymer mortars depending on the CR substitution level is indicated in Figure 7a. The control mixture, which does not contain CR, had a UPV value of 3027 m/s. Since there is no qualifying scale for the cement-based and/or geopolymer mortars, the classification for the concrete given by Hwang et al. [35] can be used in this context to get an idea about the quality of the mortars produced in this study and to establish a relationship between the other engineering properties and UPV values. According to this classification, the geopolymer mortar produced in this study can be regarded as moderate-quality mortar. However, incorporating the CR into the geopolymer mortar caused decreases in the UPV values, correspondingly in the quality of mortar. The main reason behind this situation is the macro-scale porosity formations in the geopolymer mortar. Such porous structures in the

J Sustain Const Mater Technol, Vol. 7, Issue. 2, pp. 95–107, June 2022

Figure 7. (a) Variation in UPV values of the geopolymer mortars due to the CR incorporation and the correlation between the UPV values and (b) flexural strength, (c) compressive strength, and (d) dry unit weight. hardened mortar prevent and/or delay the transferring of the ultrasonic waves from the transmitter to the receiver. For this reason, there is to be a decrease in the UPV values. But it should be also stated that the decrease in UPV values of geopolymer mortar due to the CR incorporation may be also caused by the nature of the rubber. As it is well-known, the rubber material is a perfect isolation material for waves

having any frequency. Therefore, the ultrasonic pulse waves may have been absorbed by the rubber particles during the transfer from the transmitter to the receiver. As can be seen in Figure 7a, the geopolymer mortars produced with CR had UPV values of less than 3000 m/s and more than 2100 m/s. About 30% reduction in the UPV value of geopolymer mortar was observed when the 50% of RS was substi-

J Sustain Const Mater Technol, Vol. 7, Issue. 2, pp. 95–107, June 2022

Figure 8. SEM images of (a) control geopolymer mortar mixture and (b) and (c) rubberized geopolymer mortar mixtures. tuted with CR. In this context, the rubberized geopolymer mortars produced in the present study can be considered poor-quality mortars with respect to the classification presented by Hwang et al. [35]. On the other hand, in order to establish the relationship between the engineering properties of the geopolymer mortars and their UPV values, a binary correlation was used. In this regard, Figures 7b, 7c, and 7d indicate the correlations

between UPV values and flexural strength, compressive strength, and dry unit weight, respectively. When the correlations between the UPV values and the aforementioned properties of the geopolymer mortars are investigated, it will be seen that there are strong relationships with the R2 values of more than 0.93 between the UPV values and the flexural strength, compressive strength, and dry unit weight of the geopolymer mortars. This does not only show how

J Sustain Const Mater Technol, Vol. 7, Issue. 2, pp. 95–107, June 2022

Figure 9. Cross-sections of hardened geopolymer mortar mixtures. the interactive relations between these properties are well but also indicates that the experimental program conducted in this research has been properly carried out.

3.5. SEM Analysis and Visual Observation

Here, the SEM images of both control and rubberized geopolymer mortars are presented and discussed. Two SEM images (x100 and x250 zoom-in) of the control geopolymer mortar are given in Figure 8a. Similarly, the SEM images presented in the same zoom-in of the rubberized geopolymer mortar are given in Figure 8b; however, in addition to this, three SEM images (x50, x100, and x250 zoom-in) of another rubberized geopolymer mortar are presented in Figure 8c. When the SEM images of the control mixture are investigated, a distinct ITZ between the geopolymer paste and RS particle (as shown with blue arrows) can be noticed. Besides, there are detected to be microcracks (as shown with green arrows) and unreacted and partially reacted fly ash particles on the geopolymer matrix (as indicated with pink arrows). In a similar way, when the SEM images of rubberized geopolymer mortars are investigated, more distinct ITZs between the geopolymer paste and CR particles will be sighted. The ITZ between the CR and the geopolymer paste is more distent in the SEM pictures than it is between the RS and the geopolymer paste. Such distinct ITZ formations decrease the interlock between the particle and paste, thus resulting in lower mechanical performances. Furthermore, SEM scans revealed certain gap areas around the CR particles (Fig. 8c). Such gap regions around the fine particles reduce the permeability resistance of the geopolymer mortars. In addition to the observation based on the microscopic scale, the CR distributions on the cross-section of the geopolymer mortars were visually observed. One of the difficulties faced during the production of such types of mortars is the bleeding in the mortar and relevantly the inhomogeneous distribution of rubber particles and segregation issues. Since the CR particles are lighter than the geopolymer paste, they have a tendency to move up, leading to the separation of the geopolymer paste and CR particles. As a result, in such types of materials, there occurs a layer consisting of CR particles and paste at the top of the cross-section, and there occurs another paste layer immediately below

this layer. Such problems need a very sensitive mixture design. Figure 9 is presented to display the bleeding detection and distribution of rubber particles on the cross-section of mortars produced in the present study. When the cross-sections of the geopolymer mortars illustrated in this figure are investigated, it will be seen that these layers did not occur in any mortar mixtures, and besides, the visually detected CR particles are homogeneously distributed on the mortar cross-sections. In other words, it can be stated that no bleeding and relevant segregation problems were observed in the mortar mixtures produced in the current study.