Incorporation of Glass and Plastic Waste into Alkali-Activated Mill Residue Bricks

Abstract

Recycling of glass and plastic waste has been increasingly attracting the attention of researchers worldwide. Relevant studies have been conducted to prove the feasibility of incorporating glass and plastic wastes into cement-based concrete and fired bricks. However, the high embedded energy and large carbon footprint of these materials have hindered the achievement of sustainable goals. Hence, this study attempts to diversify the recycling pathways for glass and plastic waste via a low carbon route. The brick clay mill residue has been used as a precursor to prepare alkali-activated bricks containing plastic and glass fines with a specific curing regime. The compressive strength, water absorption, linear shrinkage, and microstructure were investigated with varied content of glass and plastic content. The results showed that the maximum acceptable ratio of glass fines was around 55 wt.% for samples with the glass waste solely, achieving the compressive strength of 22 MPa. While foror samples incorporating plastic (PET) waste only, the maximum allowable ratio was only 2 wt.%, because excessive plastic resulted in the spalling of the sample surface. When both the glass and plastic waste were added to the samples, the maximum substitution ratio was 25 wt.% of glass and 2 wt.% of plastics. Scanning Electron Microscope images indicates that the plastic particles had more adverse effects on the microstructure of the alkali-activated samples than the glass particles. There was little or no bonding between plastic waste and alkali-activated mill residues. In contrast, the bonding between glass particles and alkali-activated mill residues was captured. The effect of the addition of glass and plastic samples on the durability of alkali-activated mill residue material needs to be further investigated, such as dimension stability, resistance to salt attack, freeze and thaw, and so on.

1. Introduction

Resources consumed by human beings have increased significantly since consumerism became prevalent in the twentieth century [1]. On the one hand, hundreds of thousands of products have been manufactured from these resources to fulfill and facilitate people; on the other hand, these products have left substantial refuse on the planet. Glass and plastic waste are two of the major types of all. most glass waste itself is not hazardous, except for some specific types of glass products, such as cathode ray tube [2], fluorescent lamp [3], and so on. However, inappropriate disposal of glass waste which is not biodegradable can cause severe damage to the landscape, ecology, and communities. According to Pickin et al. [4], 1.1 million tons of glass waste were generated in Australia during 2016–2017; although 57% of the material was recycled, close to half of the waste was landfilled directly.

For plastic waste, its lightweight and durability enable it to travel through space generating ecological problems on the planet and affecting the local environments such as ocean life. According to Peng et al. [5], it was estimated that there were at least 14 million tons of plastic debris ending up in the ocean every year. Such enormous accumulation of plastic waste in the ocean has begun to threaten the survival of marine wildlife severely. Studies have found that organic pollutants can be absorbed by plastic particles and then transfer into the tissues and organs of wildlife through ingestion and entanglement [6,7,8]. These infectious animals can make up their way through the food chain harming human health. A fact that is worse than the above negative consequences is that the scenario of plastic waste was much worse than glass waste as lightweight plastic waste is easier to spread in the environment and harder to re-collect than glass.

To reduce the impact of plastic waste on the environment, it needs to be managed through recycling, encapsulation, and reuse. In Australia, the recycling rate of plastics was only 12% during 2016–2017, with a total quantity of 2.5 million tons generated [4]. Even in Europe, where the strictest recycling regulation is in place, the recycling rate of the collected plastic waste was only 31.1% in 2016 [9].

Researchers attempted to upcycle the two waste streams in construction applications. One major field that has been widely researched is the incorporation of glass and plastic wastes into concrete [10,11]. It was found by some researchers that adding crushed glass waste at relatively low replacement ratios had the potential to improve the compressive, flexural strength, and modulus of elasticity of concrete [12,13,14,15]. These improvements can be attributed to the large surface area and pozzolanic compositions of the crushed glass waste [16,17]. Despite the positive results, there were studies reporting that the properties of concrete deteriorated after introducing glass waste. One of the most significant adverse effects was that crushed glass waste tended to reduce the workability of concrete mix, especially when the substitution ratio of glass waste was relatively high [16,18]. Researchers stated that the morphological features of glass waste, such as smooth surfaces and sharp edges, resulted in weak workability of the concrete [17,19].

Researchers have also examined the effect of adding plastic waste on the properties of concrete. All the existing research found that plastic waste hindered the strength development of concrete and weakened its integrity [20]. The weak bond between plastic surfaces and cement paste and the hydrophobic property of plastic waste leads to this result [21]. Despite the drawbacks, there are studies reporting that the addition of plastic waste increases the ductility of the concrete, enabling it to carry loading for an extended period after failure [22]. Overall, concrete containing plastic waste has been proven to be a viable option with some drawbacks, though the practical application is still very limited.

In addition to concrete, fired bricks are another major field for reusing glass and plastic waste. Since the main chemical composition of glass waste is silicon dioxide, this renders glass waste a favorable material to be incorporated in the fired bricks [23]. Some studies have reported that adding glass fines enhanced the mechanical performance of bricks and reduced their water absorption [24,25,26,27,28]. This was attributed to the re-vitrification of glass particles during high-temperature firing [27]. Additionally, some researchers found that glass waste had the potential to reduce the required temperature for the sintering process because of the presence of sodium oxide and calcium oxide in the glass [29,30]. Moreover, due to the extremely high temperature triggering the crystalline restructuring, fired bricks become a desirable article to immobilize the glass waste containing hazardous substances [2,24,31]. As for plastic waste, since its ignition point is much lower than the firing temperature, the primary purpose of adding it is to increase the porosity of bricks thus reducing the weight, and few studies were found in this field [32].

Although a wide range of studies have been conducted to ascertain the feasibility of incorporating glass and plastic wastes into concrete or fired bricks, it must be considered that both concrete and fired bricks possess rather high embedded energy and carbon footprint [33,34]. Thus, some researchers have started to investigate the effect of the inclusion of glass and plastic waste in the geopolymer or alkali-activated materials to recycle the waste more sustainably [35,36]. The existing literature on geopolymer or alkali-activated materials incorporating glass and plastic waste mainly focused on the use of fly ash [35,37], slag [38], and metakaolin [39,40] as precursors. Most of these studies were undertaken to produce waste-added geopolymer or alkali-activated concrete [41,42]. Limited research was conducted to use other precursors, to prepare the geopolymer or alkali-activated bricks or masonry units containing glass and plastic waste. Thus, it would be able to diversify a new recycling pathway to explore the use of another material, brick clay mill residues, as the precursor for glass- and plastic-waste-added alkali-activated bricks.

Brick clay mill residues are an industrial by-product generated by the pulverization process of raw clay and shale into desirable sizes for fired brick manufacturing. Its ultrafine size of particles is unsuitable for conventional bricks but has the potential to be used as the binder material in alkali activation. previous studies have demonstrated the feasibility of the alkali-activated mill residue bricks, determining the best introduction ratio and the composition of the alkaline activator, as well as the curing regime [43,44].

The present research, therefore, examines the viability of incorporating glass and plastic waste into the alkali-activated bricks from brick clay mill residues. The effects of glass and plastic waste on the compressive strength, linear shrinkage, and water absorption of the alkali-activated bricks are investigated. The microstructures of glass- and plastic-added samples were observed. Lastly, a comparison between the waste-incorporated mill residue bricks and conventional fired bricks is provided.

2. Results and Discussion

2.1. Compressive Strength Evaluation

2.1.1. Glass Waste

The experimental results of the compressive strength of samples incorporating glass waste (<0.4 mm) are illustrated in Figure 1. In general, the substitution of mill residues with glass waste weakened the compressive strength of the masonry units. While the strength of samples declined as the glass content increased, the samples containing up to 55 wt.% glass waste satisfied the compressive-strength requirement of 20.7 MPa for the severe weathering (SW) grade bricks in ASTM C62 [45].

The results shown in Figure 1 illustrate that the introduction of glass waste at 5 and 15 wt.% notably deteriorated the compressive strength of samples, reducing the strength from 40 MPa (0 wt.%, 28d) to 35 MPa (15 wt.%, 28d). However, the samples incorporating 25 and 35 wt.% did not exhibit a further significant decrease in compressive strength compared with the samples with 15 wt.% glass waste. The compressive strength of 28-day-old samples comprising 35 wt.% glass waste was 33 MPa, which was merely less than the 35.3 MPa of the 15 wt.%-glass-waste samples slightly. When the content of glass waste exceeded 35 wt.%, the compressive strength of samples weakened rapidly again. This phenomenon can be explained by the three structures of the alkali-activated mill residue bricks comprising filler materials, as shown in Figure 2. When the ratio of the filler material, glass waste, is low, the particles of glass waste are suspended in the body of alkali-activated mill residues. This kind of microstructure is called suspension mode (Figure 2A). The strength of the suspension structure is significantly affected by the coherence between the binder material and filler material. In the present study, the filler material, soda-lime glass, is a relatively-low reactive material with smooth surfaces, attributed to the stability of Silicon-Oxygen tetrahedron. Both the weak reactivity and smooth surface lead to weak coherence between the alkali-activated mill residues and glass waste. As a result, the compressive strength of samples with suspension structures would decrease as the content of glass waste increased. Therefore, the samples incorporating 5 and 15 wt.% glass waste (<0.4 mm) showed a decrease in strength.

As the content of filler material increased, its particles suspended in the binder material would adhere and interlock with each other, forming a skeleton to undertake loading with the binder material mutually. This kind of microstructure is called the compact skeleton mode (Figure 2B). In the current study, when the ratio of glass waste (<0.4 mm) increased to 25 and 35 wt.%, the compressive strength of samples did not decline notably compared with the samples containing 15 wt.%. This result indicated that the dominated structure of 25 and 35 wt.% glass-added samples had moved from suspension to compact skeleton mode. Although the increasing content of glass waste would weaken the samples, the skeleton formed due to interlocked glass particles undertook part of the loading, reducing further deterioration. Hence, the compressive strength of samples incorporating 25 and 35 wt.% glass waste (<0.4 mm) did not decrease significantly compared with the 15 wt.% glass samples.

When the substitution ratio of glass waste (<0.4 mm) further increased to 45 and 55 wt.%, the structure of samples shifted from the compact skeleton to the void skeleton mode (Figure 2C). The difference between the compact and void skeleton is that the void skeleton possesses higher porosity than the compact skeleton. Although the framework of the interlocked glass particles still existed in the samples to bear loading, the binder material, alkali-activated mill residue, was not sufficient to fill the gaps between glass particles, resulting in considerable voids within the samples. These voids significantly weakened the compressive strength of the samples due to the reduction in cross-sectional area and potential stress concentration [46]. Therefore, the strength of samples continued to decrease as the content of glass waste (<0.4 mm) was more than 35 wt.%, as shown in Figure 1.

Moreover, it was observed that the compressive strength of the samples reduced as the age increased (the gap between the curves of 7-day and 28-day compressive strength), but the addition of glass waste leaned to mitigate this effect (Figure 1). The drop in strength decreased as the glass content rose. There was basically no difference in compressive strength between the 7-day and 28-day samples when the substitution ratio reached 55 wt.%.

A previous study revealed that the decrease in strength from the age of 7 to 28 days was mainly attributed to the absorption of moisture from the atmospheric environment [43]. There are two major reasons for the observation that adding glass waste mitigated the decrease in strength due to short-term aging. The first reason is that the relative ratio of the alkaline activator to the mill residues increased as the substitution ratio of glass waste rose. Since the usage of the alkaline activator relating to the dry weight of total solid materials (mill residues and glass and/or plastic waste) was fixed at 20 wt.%, the increased glass content represented the higher ratio between the mill residues and alkaline activator. As a result, the degree of activation of the mill residues would increase as the amount of glass waste rose. The high-degree activation would lead to dense binding between mill residue particles, thereby low porosity of the alkali-activated mill residues. The samples containing high content of glass waste would absorb less moisture from the atmosphere than the samples incorporating glass waste at low ratios.

The second reason for the mitigation effect of the glass waste on the strength weakening during aging is the existence of the interfacial transition zone between glass waste and alkali-activated mill residues. Figure 3 shows the results of SEM-EDX analysis about the interfacial transition zone in the sample containing 25 wt.% glass waste. It can be observed that a rather thin layer formed between the glass waste and alkali-activated mill residues. This indicated that the alkaline activator dissolved glass waste to a limited depth of around 200 nm from the surface. The dissolved glass and mill residues combined into another material, forming the interfacial transition zone. According to the EDX result, the material synthesized in the zone is postulated to be sodium aluminosilicate gel dominated by silicon. Although the thin layer of sodium aluminosilicate did not contribute to the strength gain, it played a role in sealing the surface of the samples so that relatively less moisture vapor would enter the samples. Considering the two reasons above, the reduction in strength of samples due to short-term aging decreased as the glass content rose.

Figure 4 shows the effect of the size of crushed glass waste on the compressive strength of samples. The data indicated that the coarse particles of crushed glass had more adverse effects on the strength of samples than the fine glass waste, provided that the range of the size of glass waste was below 1.5 mm. The samples incorporating <0.4 mm of glass waste possessed higher strength than the samples containing 0.4–0.8 mm and 0.8–1.5 mm of glass waste. The reason is that the finer glass fines possess higher specific surface area, more easily being dissolved in alkaline environment, leading to more interfacial bonding between glass and alkali-activated mix, thereby higher strength.

Another finding from the data is that the difference in strength between the samples comprising the glass waste with the three sizes enlarged as the substitution ratios increased. For example, when the substitution ratio was 5 wt.%, the compressive strength of the samples incorporating <0.4 mm, 0.4–0.8 mm, and 0.8–1.5 mm were 38.2 MPa, 37.3 MPa, and 36.5 MPa, respectively. The difference between these samples was within the range of 2 MPa. When the content of glass waste increased to 55 wt.%, the strength of samples using <0.4 mm, 0.4–0.8 mm, and 0.8–1.5 mm of glass waste recorded 22.1 MPa, 16.2 MPa, and 10.8 MPa, respectively. The difference in strength between the 55 wt.%-glass samples with different glass sizes increased to around 5–6 MPa.

The increment of compressive strength differences amongst the samples with the three sizes of glass waste can also be explained by the concept of microstructures as described in Figure 2. When the ratio of glass content was low, as, in the 5 wt.%, the microstructure of the samples would be suspension (Figure 2A), regardless of the particle sizes of glass waste. In this circumstance, the coarser glass waste would generate more impact on the integrity of samples than the finer glass waste, thereby resulting in a lower strength. However, the impact of the overall strength is limited when the difference in size is little. As the content of glass waste increased, the microstructure of samples would change from the suspension to the void skeleton. The samples incorporating a larger size of glass waste would turn into the void skeleton at a lower ratio of glass waste because the bulk density of a material decreases as its particle size increases. Provided the same weight of glass waste in a certain shape, the larger particle size would lead to larger voids. As a consequence, there would be more voids in the samples with larger sizes of glass waste than finer ones. Therefore, the difference in strength between the samples with the three sizes of glass waste was increased as the substitution ratio of glass increased.

2.1.2. Plastic Waste

The effects of the addition of plastic waste on the compressive strength of samples are shown in Figure 5. It is clear that substituting part of the mill residues with PET waste deteriorated the compressive strength of the samples. When the PET ratio reached 5 wt.%, the strength of the samples at the age of 28 days was unable to satisfy the minimum requirement for the Grade SW bricks in ASTM C62 [45]. It should be noted that this result did not represent that the content of PET waste can be added into the mill-residue samples up to 4 wt.%. As shown in Figure 6, there was no obvious PET piece found on the surface of samples containing 2 wt.% plastic waste. However, when the content of plastic waste increased to 3 wt.%, noticeable PET pieces appeared on the surface of samples and could be peeled off easily from the sample body. This outcome suggests that the immobilization of PET waste by alkali-activated mill residues was ineffective when the substitution ratio is equal to or above 3 wt.%, provided that the size of PET pieces was around 3 mm. Thus, the appropriate incorporation ratio of plastic waste was determined at 2 wt.% other than 4 wt.%.

Another finding from Figure 5 is that adding PET waste did not reduce the strength reduction during the short-term aging as compared with the glass waste. This observation can be attributed to little or no coherence between PET waste and alkali-activated mill residues. Figure 7 presents the microstructures of samples containing glass waste and plastic waste, obtained through SEM. The failure crack found on the particle of glass waste indicated that the glass and alkali-activated mill residues undertook the load mutually, suggesting there was appropriate coherence between the two materials (Figure 7A). In contrast, notable gaps between the PET piece and alkali-activated mill residues can be identified in the plastic-added samples. These gaps allowed atmospheric moisture vapor to enter the core of the sample easily, thereby resulting in a decrease in the strength of samples after short-term aging. The deformation of PET pieces was due to the heating temperature of 155 °C which also contributed to the presence of these gaps.

Comparing Figure 1 and Figure 5, the incorporation of plastic waste deteriorated the strength of the samples more than the glass waste. In addition to the gap formed between PET pieces and alkali-activated mill residues mentioned above, the degradation of the PET pieces was another key reason for the weak strength observed. As shown in Figure 8A,B, it was evident that the PET piece added to the samples was significantly eroded by the alkaline activator after curing at 155 °C. By contrast, the glass particle was still stable in the alkaline environment, and there were interlockings formed between the surface of the glass and alkali-activated mill residues. The degradation significantly weakened the capacity of plastic waste to withstand loading, so the strength of the sample containing PET was rather low.

2.1.3. Combined Glass and Plastic Waste

The acceptable substitution ratio of using both the glass and plastic waste was determined when introducing these two materials into samples together. The content of plastic waste was fixed at 2 wt.%, which was the most appropriate introduction ratio of PET waste identified in the previous experiments. Figure 9 illustrates the compressive strength of samples containing both glass waste (<0.4 mm) and plastic waste (fixed at 2 wt.%). The more addition of glass waste led to lower compressive strength. When the ratio of PET waste was fixed at 2 wt.%, the maximum acceptable ratio of glass waste was 25 wt.% to satisfy the requirement of ASTM C62 [45].

2.2. Linear Shrinkage

Figure 10 illustrates the linear shrinkage of samples incorporating three sizes of glass waste, plastic waste, and both glass and plastic waste simultaneously. As expected, the addition of glass waste decreased the shrinkage of samples, and the higher ratios of glass waste led to a lower linear shrinkage (Figure 10A). Since the shrinkage of the alkali-activated samples was mainly due to the loss of moisture, the minor moisture absorption of glass was the first reason for the decreasing shrinkage with an increase in glass content. In addition, the mitigating effect of glass on shrinkage was also associated with the microstructure formed by glass waste particles inside the samples. The skeleton of interlocked glass particles was the second reason why adding glass waste can resist shrinkage. Furthermore, provided void-skeleton samples with a fixed ratio between glass waste and mill residues, the content of mill residues within the sample would decline as the size of glass waste increases. Hence, the differences in shrinkage between samples using the three sizes of glass waste were enlarged with the increase in glass content.

Similarly, the shrinkage of samples incorporating PET waste also decreased as the content of plastic waste increased (Figure 10B). The first reason for this result is that PET is a water-repellent material, so it would not deform noticeably due to gain or loss in moisture content. The second reason is the deformation of PET pieces attributed to the elevated temperature curing, especially during the early stage of curing when the strength of samples had not developed. This is also the reason why there was an aesthetic issue with the surface of the plastic-added samples (Figure 6B). In addition, the samples containing both glass and plastic waste had lower shrinkage compared with the samples incorporating glass or plastic waste only.

2.3. Water Absorption

2.3.1. Glass Waste

The results of water absorption tests for the glass-added samples are shown in Figure 11. The figure illustrates that the water absorption of glass-added (<0.4 mm) samples decreased as the glass content increased when the substitution ratio of glass was less than 35 wt.%. However, when the substitution ratio was more than 35 wt.%, the water absorption started to increase as the quantity of glass waste rose. Such observations can be explained by the microstructure of glass-incorporated samples. When the content of glass waste was reduced, the microstructure of the samples was in the suspension mode, in which the glass waste particles had no contact with each other. Due to little or no absorption of glass, the samples containing greater glass waste would possess lower water absorption ratios after immersion.

As the glass content increased, the microstructure of samples would shift to the compact skeleton mode, in which the binder materials, alkali-activated mill residues, would exactly fill the voids within the skeleton formed by the interlocked glass particles. The samples with the compact-skeleton microstructure would possess the lowest water absorption ratio as they had the maximum content of glass and the minimum open voids within a given volume. This ratio of glass waste was the turning point of the water absorption curve. Beyond the ratio leading to the lowest water absorption, the microstructure of samples changed to the void skeleton mode, in which there were voids that can trap the moisture content after immersion. Thus, the water absorption ratios of samples began to increase as the ratio of glass waste increased again.

The maximum hot water absorption ratio regulated by ASTM C62 [45] is 17 wt.%, so all the substitution ratios of glass waste (<0.4 mm) except 75 wt.% satisfied the requirement. In addition, the data showed that the difference between cold and hot water absorption ratios of the glass-added samples was minor and was less than the control samples without the addition of glass waste. This indicated that the addition of glass waste had positive effects on improving the stability of samples under the hydrothermal environment. The reason for this finding can be attributed to the thin interfacial transition zone formed between the glass particles and alkali-activated mill residues identified in Figure 3. The zone and glass particles jointly functioned as a barrier to reduce the paths that the moisture content could seep through into the core of the sample, therefore it was hydrothermally stable.

Moreover, the particle size of glass waste influenced the water absorption of samples as well. As shown in Figure 12, the samples containing 0.8–1.5 mm glass waste possessed the lowest hot water absorption ratio at the glass content of around 20 wt.%. For the glass waste with the particle size of less than 0.4 mm, the lowest water absorption appeared at the ratio of 35 wt.%. The lowest water absorption ratio of the samples using 0.4–0.8 mm glass waste was between the values of samples with 0.8–1.5 mm and less than 0.4 mm glass waste. This trend in water absorption was consistent with the trend in compressive strength determined in Figure 1. The turning points of the compressive strength trends for the samples using <0.4 mm, 0.4–0.8 mm, and 0.8–1.5 mm were also around 15, 25, and 35 wt.%, respectively. These results reinforced the argument that the use of a larger size of glass waste rendered the samples achieved the compact-skeleton microstructure at a low substitution ratio of glass. Furthermore, it can be found that at the low ratios that led to the suspension microstructure of samples, the sizes of glass in the ascending order of water absorption were 0.8–1.5 mm, 0.4–0.8 mm, and <0.4 mm, though the magnitude of the difference was small. However, for the samples in the void-skeleton microstructure with relatively high ratios of glass waste, this order regarding the size of glass completely reversed, and the magnitude of the difference was enlarged. This was because the coarse and fine particles had similar effects on the samples with the suspension microstructure, whereas the coarse particles influenced the formation of the void-skeleton microstructure more than the fine particles.

2.3.2. Plastic Waste

Figure 13 shows the water absorption ratios of the samples containing plastic waste up to 5 wt.%. The curve suggests that the water absorption of samples slightly increased as the content of PET waste rose. This finding supported the argument that PET waste weakened the microstructure of samples. Assuming the microstructure of samples remained unchanged, the water absorption ratios would decrease as the content of plastic waste increased due to the water repellence of the PET. However, the result observed contradicted the above reasoning. Although the PET pieces hardly absorb any moisture content, their deformation due to heating and little or no coherence to alkali-activated mill residues disturbed the microstructure, increasing the porosity of samples. Thus, the water absorption of plastic-added samples was in an ascending trend. Moreover, the notable increase in water absorption after immersion in boiling water indicated that the microstructures of PET-added samples were further destroyed because the gaps between PET pieces and alkali-activated mill residues enabled boiling water to seep into the core of samples. This kind of phenomenon was not observed in the samples with glass waste.

2.3.3. Combined Glass and Plastic Waste

Figure 14 shows the results of water absorption ratios of samples containing both glass and plastic waste. It was found that the water absorption ratios did not show a similar decreasing trend as identified in the samples incorporating glass only when the content of glass waste was low. This reflected that the addition of PET waste tended to degrade the microstructure of the samples. Nevertheless, the water absorption ratios of the samples with up to 55 wt.% glass waste and 2 wt.% plastic waste satisfied the maximum water absorption ratio of 17 wt.% required by ASTM C62 [45].

2.4. Comparison of Fired Bricks and Waste-Added Alkali-Activated Bricks

The above results have demonstrated the potential use of glass and plastic waste as the filler material in the alkali-activated mill residue bricks. It is necessary to compare the glass- and plastic-added alkali-activated bricks with the conventional fired bricks now.

Table 1. Comparison between the alkali-activated mill residue bricks incorporating glass and plastic waste and the conventional fired bricks.

Phases	Alkali-Activated Bricks Incorporating Glass and Plastic Waste	Conventional Fired Bricks
Raw materials sourcing	Mill residues Glass waste Plastic waste Sodium silicate Sodium hydroxide	Clay Shale
Mixing	Requiring some modifications on the ordinary mixing system	Ordinary mixing system
Shaping	Pressing/extruding	Pressing/extruding [47]
Drying/Curing	50 °C and 90% RH for 48 h 155 °C and 80% RH for 24 h	38–204 °C with moderate humidity for 24–48 h [47]
Firing	None	Up to 1316 °C for 10–40 h [47]
Cooling	No specification required	Gradually cooling for several hours [47]

illustrates each stage of the production of the two types of bricks. In the first stage of raw material sourcing, the conventional fired bricks cost less than the alkali-activated bricks, because the latter requires alkaline activators and additional transportation costs for the waste materials. Nevertheless, the glass- and plastic-added bricks possess the crucial advantage of diverting the two common types of waste from landfills and the natural environment, improving sustainability. Particularly, considering the fact that the fired bricks are not suitable to incorporate plastic waste due to the extremely high temperature and generating more carbon emissions.

The second stage of brick production is mixing. Due to the use of the alkaline activator, manufacturing alkali-activated bricks requires brickmakers to modify the existing mixing system of their production lines. This would be another extra cost of the alkali-activated bricks. As for the shaping process, there would not be much difference between the alkali-activated mill residue bricks and fired bricks. The alkali-activated bricks can be shaped via the existing facilities used by the conventional fired bricks.

The largest difference between the alkali-activated bricks and fired bricks is at the drying/curing and firing stages. Although both types of bricks need to be placed at temperature ranging from 50 to 200 °C and in a humid environment, the purposes are completely different. For the fired bricks, the drying process is simply to eliminate the moisture content within the bricks before firing so that cracks would appear on the brick body. It does not directly contribute to the development of the strength of fired bricks. In contrast, for the alkali-activated bricks, the drying process is the elevated-temperature curing process, during which the bricks complete the development of strength. Therefore, the alkali-activated bricks do not require the firing process. This characteristic significantly contributes to sustainability, because the energy consumption and carbon emission of the firing process account for more than 80% of the total consumption and emission of the whole production process of conventional bricks [48]. Despite the additional embedded energy consumption and carbon emissions from the use of the alkaline activator, one previous study has proven that the consumption and emissions pertaining to the activator used in the alkali-activated bricks are significantly less than the quantity required for the firing process of conventional bricks [43]. Overall, the alkali-activated mill residue bricks incorporating glass and plastic waste are one possibility to improve the sustainability of the brick industry.

3. Materials and Methods

3.1. Materials

Brick clay mill residues used to produce samples were obtained from a local brick manufacturer in Melbourne, Australia. The particle size distribution of the raw mill residue is shown in Figure 15. Over 60% by weight (wt.%) of the mill residue particles are finer than 75 µm, and more than 30 wt.% of the particles are finer than 5 µm. The mill residues were sieved through 4.75 mm to eliminate large particles.

The crushed glass waste for sample making was derived from container glass made from the most common soda-lime glass waste. There were three categories of crushed glass waste according to the particle size: less than 0.4 mm, 0.4–0.8 mm, and 0.8–1.5 mm. The plastic waste used in this study was prepared from used polyethylene terephthalate (PET) bottles. The used bottles were collected from communities around the Swinburn University of Technology and then shredded into pieces less than 5 mm by a granulator with a screen sieve.

The alkaline activator is comprised of sodium hydroxide (SH) solution (8 M) and sodium silicate (SS) solution (14.7 wt.% Na₂O, 29.4 wt.% SiO₂, 55.9 wt.% H₂O). The two solutions were mixed for 1 min before sample preparation and then added immediately into the dry mix. The ratio between SH and SS was fixed at 1.0, which was the optimal ratio found by prior research [43].

3.2. Methods

The mill residues and glass waste were dried at 105 ± 5 °C for 24 h, and the plastic waste was dried at 50 ± 1 °C for 72 h before the sample-making process. The dry materials were placed in a mixer bowl at a designated ratio and then mixed for 2 min to achieve better uniformity. The alkali activator mentioned above was added gradually into the mixer bowl at the ratio of 20 wt.% relative to the total dry weight of mill residues, glass waste, and/or plastic waste. Water (2 wt.% relative to the same total dry weight) was added to the mixtures to maintain the total moisture content at 15 wt.% so that the plasticity of the mixture was suitable for shaping. The mixing of solid materials and liquid activator lasted for 3 min. The ready mixture was filled into a stainless-steel mold for compression molding. The load applied was fixed at 8 tons, and held for 1 min. The block samples de-molded had dimensions of 115 mm in length, 110 mm in width, and 76 mm in height. Another set of cylinder samples (19 mm in diameter and 127 mm in length) were prepared for linear shrinkage measurement according to ASTM C326-09 [49]. Following the curing regime identified in our previous study [43], all the samples were cured at 50 °C and 90% relative humidity (RH) for 48 h, subsequently at 155 °C for 24 h.

Table 2. Mix design of alkali-activated composite.

Specifications
Sodium hydroxide (SH) concentration	8 M
Sodium silicate composition	14.7 wt.% Na2O, 29.4 wt.% SiO2, 55.9 wt.% H2O
Alkaline activator ratio (in addition to the total dry mix)	20 wt.%
SH/SS ratio by mass	1.0
Equivalent water/solid ratio	15 wt.%

and

Table 3. The content of glass and plastic waste in all the samples.

Sample ID	Glass Waste		Plastic Waste	Mill Residues
	wt.%	Sizes (mm)	wt.%	wt.%
Control	--	--	--	100
G5	5	<0.4 0.4–0.8 0.8–1.5	--	95
G15	15	<0.4 0.4–0.8 0.8–1.5	--	85
G25	25	<0.4 0.4–0.8 0.8–1.5	--	75
G35	35	<0.4 0.4–0.8 0.8–1.5	--	65
G45	45	<0.4 0.4–0.8 0.8–1.5	--	55
G55	55	<0.4 0.4–0.8 0.8–1.5	--	45
G65	65	<0.4 0.4–0.8 0.8–1.5	--	35
G75	75	<0.4 0.4–0.8 0.8–1.5	--	25
P1	--	--	1	99
P2	--	--	2	98
P3	--	--	3	97
P4	--	--	4	96
P5	--	--	5	95
G5P2	5	<0.4	2	93
G15P2	15	<0.4	2	83
G25P2	25	<0.4	2	73
G35P2	35	<0.4	2	63
G45P2	45	<0.4	2	53
G55P2	55	<0.4	2	43

show information on all the sample mixes. Specifically, the sample group incorporating glass waste had three subgroups according to the particle size of the crushed glass.

The compressive strength of a sample was determined by Tecnotest 2000, and the loading speed was fixed at 0.5 MPa in accordance with ASTM C67 [50]. The linear shrinkage measurement was performed on the cylinder samples. Two reference lines 102 mm apart were marked on each cylinder sample using gauge blocks once the sample was de-molded after shaping. The varying distance between the reference lines was measured by a Vernier caliper (with a resolution of 0.01 mm) at a designated age of the sample after curing, recorded as L in mm. The linear shrinkage was then calculated by [(102 − L)/102] × 100%] [49]. The cold and hot water absorption ratios of samples were obtained by immersing oven-dried samples into cold water (15 ± 2 °C) for 24 h, followed by immersion in boiling water for 2 h. The initial weights (W_i) of dried block samples were measured by a lab balance (A&D EJ-4100—0.1 g resolution). The weights of the samples were measured after cold- and hot-water immersions, respectively, recorded as W_c and W_h. The cold and hot water absorption ratios were calculated by (W_c − W_i)/W_i and (W_h − W_i)/W_i, respectively. For the microstructures of samples, the cores of crushed samples from compressive strength tests were collected and placed into isopropyl alcohol for 24 h to stop possible reactions that occurred inside the samples. The core pieces were coated with gold using EMITECH K975X (40 mA, 60 s) for Scanning Electron Microscope—Energy Dispersive X-ray Spectroscopy (SEM-EDX) analysis. Gemini SUPRA 40VP was used for SEM-EDX at the electron high tension (EHT) voltages of 5 kV for SEM and 15 kV for EDX.

4. Conclusions

This research has examined the major properties of alkali-activated bricks containing glass and plastic waste. Up to 55 wt.% glass waste (<0.4 mm) can be incorporated into alkali-activated mill residue samples solely to meet the compressive strength requirement of the severe-weathering-grade bricks in accordance with ASTM C62 [45]. As the size of glass waste increased, the acceptable substitution ratios of glass waste decreased to 45 wt.% and 35 wt.% for 0.4–0.8 mm and 0.8–1.5 mm glass waste. This finding was attributed to the effect of the glass particle size on the microstructure of samples. As for the plastic-added samples, the acceptable substitution ratios were up to 2 wt.% considering both the compressive strength and surface appearance. The samples containing more than 2 wt.% PET waste had obvious aesthetic issues too. When introducing both the glass and plastic waste, the satisfactory samples were the ones with up to 25 wt.% glass waste and 2 wt.% PET waste.

Moreover, the glass and plastic waste were found to have effects on linear shrinkage and water absorption. Both the waste materials decreased the shrinkage of samples. This property may enable manufacturers to produce lightweight building materials. As to the water absorption, the plastic waste increased this ratio as it degraded the microstructure of samples. For glass waste, the water absorption of samples firstly decreased as the content of glass increased and then increased. This was also due to the effect of the size of glass particles on the microstructure of samples.

Furthermore, this study has compared the alkali-activated mill residue bricks incorporating glass and plastic waste to the conventional fired bricks. Diverting glass and plastic waste from landfills and the natural environment is one of the most important advantages of the bricks in the present study. It also tackles the problem that plastic waste cannot be recycled into conventional fired bricks. Although the use of an alkaline activator increases the embedded energy consumption and carbon emissions of the bricks, the corresponding conservation of energy and reduction in emission outweigh it, pertaining to the elimination of the firing process.