Introduction

According to a 2017 United Nations prediction, the global population is projected to reach 9.8 billion by the year 2050, indicating a 30% growth compared to the population in 2017. The demand for resources also increases as the population increases[1]. The increased demand for food, infrastructure, healthcare, and education is a result of population growth, leading to enhanced demand for more cement, steel, and bricks. Bricks are the primary building blocks for fulfilling the demand for infrastructure. China is the largest producer of bricks, with a total share of more than 50%, followed by India with a share of 11% [2]. Fired clay bricks are still the most widely adopted and promoted bricks, which are highly energy-consuming processes that make use of fossil fuel burning to modulate the clay structure for strength; moreover, they are rapidly exhausting the agriculture fertile soil[3, 4]. On the other hand, cement consumption climbed to 355.46 million tons in 2022 and is projected to reach 450.78 million tons by 2027 due to an increase in infrastructure demand. With more than 7% of global capacity, India is the world's second-largest cement manufacturer, according to cement industry research. Domestic cement production in India will increase from 296 million tons in 2021 to 356 million tons in 2022 [5]. One ton of ordinary Portland cement (OPC), a conventional binding material, makes use of almost 1.7 tons of source raw material and produces approximately one ton of CO2 in the atmosphere[6]. Zhang et al. [3] have classified the brick-making process based on material and process orientation, which includes fired brick, cement-based, geopolymer-based, and others. The use of waste as a raw material, especially waste that is rich in alumina and silica, like different kinds of slag [7,8,9], tailings mine waste [10,11,12], RHA [13,14,15], cotton waste [16], oyster shell [17], and sawdust from wood [18], for brick production could provide an economical and environmentally friendly solution. Waste materials may be used as inputs for the production of composites, which helps to tackle waste management problems and enhance the qualities of the composites, hence replacing non-environmentally friendly materials [19]. Several researchers have tried to provide a sustainable solution for waste utilisation in making bricks. This research truly focuses on the circular economy principle and explains its contribution to circular economy in Table 1

Table 1 Overview of circular economy principles contribution
Full size table

Aluminosilicate waste materials are primarily used as precursors for making geopolymer binders along with alkaline activation. The problem of effective industrial waste disposal continues to be a significant concern, particularly in developing countries. Additionally, pollution caused by plastic waste is a major environmental concern globally [20, 21]. Plastic manufacturers in Europe estimated plastic production at 370 million metric tons globally, and at the current pace, it will reach 11 billion tons by 2025. Further, only a small portion, as small as 9%, is recycled, and the remaining 80% moves to landfills or is dumped in the marine environment, causing endangered species [21,22,23]. In the recent decade, geopolymers, due to their decreased energy demands [24] and lower carbon emissions[25, 26] have attracted interest as a more environmentally friendly option for building materials and as a binder alternative to cement[27, 28]. Researchers in the past have used plastic waste for different applications and emphasised the significance of PET [29] and polypropylene PP [30] for providing environmentally friendly solutions. The selection of PET plastic waste for incorporation into geopolymer bricks is strategically grounded on several crucial criteria, as outlined in Table 2.

Table 2 Reasons for choosing PET plastic waste in geopolymer bricks
Full size table

The production and use of waste materials for making geopolymers, including fly ash, a by-product obtained from coal burning[39], GGBS [40, 41] produced from iron manufacturing, and RHA [14, 42, 43] a by-product of milling rice and red mud [44,45,46] produced during the processing of aluminium, are industrial by-products that are merely waste materials that help through the preservation of natural resources and the reduction of greenhouse gas emissions. These waste materials not only lessen the burden associated with trash disposal but also support sustainable building practices. These precursor materials are rich sources of silica and alumina, which favour the polymeric reaction required to make a geopolymer binder. The sialate monomers that are recapping units (Si–O-Al-O) are formulated during geopolymerization to form a geopolymer binder[47,48,49]. For the geopolymer binder to solidify, the geopolymer reaction essentially consists of six phases. In the first, aluminosilicate materials that are rich in alumina and silica are mixed in an alkaline solution of NaOH and Na2SiO3 to aid and accelerate their dissolution and depolymerization, resulting in free [SiO4]- and [AlO4]-tetrahedral units that get separated and are unrestricted in the solution. These unbound molecules [SiO4]- and [AlO4]-tetrahedral units fascinate oxygen atoms, which results in the development of polymeric precursors in the arrangement of (-Si–O-Al-O-)-sialates linkage[50,51,52]. Six major phases are involved in the geopolymer solidification process including reticulation, polycondensation, gel formation of oligosialates, alkalinisation, and depolymerization of silicates[53]

The scientific literature reports that bricks made with GGBS possess higher compressive strength when compared with traditional red clay bricks. Researchers have used lime and GGBS in this investigation, and compaction was done at 4.9 MPa, cured for 270–272 °C with 95% humidity, and allowed to mature for 28 days. compressive strength of 20 MPa was recorded [54, 55]. Lazorenko et al. investigated the characteristics of coal fly ash-based GP mortars with increasing percentages (20–100%) of PET particles substituting quartz sand. The findings reveal that increasing plastic aggregate concentration corresponds with lower compressive and flexural strengths, whereas splitting tensile strength somewhat rises to a 40% sand substitution. In particular, GP mortars made from recycled PET are lighter (up to 15% lighter), better at absorbing water (up to 26%), and better at keeping heat in (up to 59%) when they are completely replaced [56]. This could be a way to make sustainable building materials that are good for the environment and the economy. Zhang et al. reviewed the different geopolymer bricks formed with different methods and materials. Further, he reported that the energy-intensive manufacture of burnt clay bricks has spurred research on sustainable solutions to reduce the brick industry's carbon impact. This Table 3 summarizes key findings, materials, and outcomes from recent studies on geopolymer bricks, highlighting various aspects such as compressive strength, water absorption, durability, and environmental considerations.

Table 3 Summary of key findings, materials, and outcomes in recent studies on geopolymer bricks
Full size table

Geopolymerization provides a sustainable alternative to fired clay bricks[3]. Maaze et al. conducted intensive research on geopolymer brick and made geopolymer brick mixes with brick waste powder (BWP) from building and demolition debris. The study found that 12M2.5AS50 °C was the best mix after using multi-response optimization to change the molarity (4 M–12 M), the alkaline solution ratio (1.5–2.5), and the curing temperature (40–60 °C). Molarity, alkaline solution ratio, and curing temperature affected physical and mechanical qualities. XRD, SEM–EDX, FTIR, and TGA were used to study phase structure, morphology, and heat resistance, verifying the experimental variables' impact on brick mixtures [62, 63].

This research study primarily focuses on the utilisation of almost 100 percentage waste material in the form of industrial by-products, i.e. RHA, GGBS, Red Mud, C&D waste, and plastic waste, as mixed ingredients for making sustainable geopolymer bricks. The novelty of this research lies in the use of three different aluminosilicate materials in developing a geopolymer binder with the use of PET plastic waste and the implementation of machine learning algorithms to check the efficacy of geopolymer bricks in strength development at different levels of plastic waste. Additionally, this research has made use of microstructural studies to correlate the results obtained during physical–mechanical testing of the geopolymer bricks.

Materials and mix design

Source materials

The raw materials were collected from the local sources available in the Punjab and Chandigarh regions. Rice husk ash is one of the major concerns in the region of Punjab and is abundantly available, it was used as the major proportion for making geopolymer binder due to its richness in silica, as revealed in morphological analysis. In order to compensate for the alumina deficiency, GGBS, which is rich in alumina compounds, is used along with red mud for the tintering effect. Geopolymer binder performance is greatly influenced by the choice of the appropriate material; therefore, detailed microstructural analyses were performed in order to get insights related to their chemical and physical nature, which were achieved by SEM, XRD, XRF, and FTIR. The WD-XRF was performed on a spectrometer Model-S8 TIGER Bruker, and the results are shown in Table 4. WD-XRF results of RHA, GGBS, and Red Mud, show the elemental composition of the oxide in these materials. A scanning transmission electron microscope (SEM) model JSM6100 (Jeol) was used to perform SEM of the raw material, which is shown in Fig. 1 RHA has sharper and irregular particles, whereas GGBS and red mud have round particles, which might be responsible for different behaviours in strength, water absorption, and setting time of the geopolymer binder.

Table 4 WD-XRF results of RHA, GGBS, and Red mud
Full size table
Figure 1
figure 1

a SEM image of RHA b GGBS c Red mud

Full size image

Small-angle X-ray scattering by ANTON PAAR model SAXSPACE with Panalytical's X.Pert Pro The XRD peak was measured between 5° and 90° 2θ, with a step size increase of 0.05. The generator set was modulated at 30 mA and 40 kV, and X'pert Highscore software was used to evaluate the XRD peak data.

The most common minerals found were quartz, mullite, cristobalite, goethite, hematite, and kaolinite, which are shown in Fig. 2. These materials' unique chemical fingerprints can be seen in FTIR spectrophotometer model RZX (Perkin Elmer) measurements, as shown in Fig. 3 GGBS shows silicate and aluminate bands in the 800–1200 cm−1 and 400–600 cm−1 ranges, reflecting its composition. Rice Husk Ash (RHA) exhibits strong Si–O-Si bands around 1000–1300 cm−1, while Red Mud displays iron oxide bands at 400–600 cm−1 and alumina-related bands.

Figure 2
figure 2

XRD analysis of RHA, GGBS, and Red mud

Full size image
Figure 3
figure 3

FTIR analysis of RHA, GGBS, and Red mud

Full size image

The alkaline solution was prepared using sodium silicate (Na2SiO3) containing 27.74% SiO2 and 8.33% Na2O, manufactured by Loba Chemie Pvt. Ltd., India, and sodium hydroxide NaOH with 97.78% purity in the form of flakes was used to prepare the 12 M alkaline solution in the deionized water. Recycled sand obtained from the C&D waste plant setup by municipal cooperation Chandigarh (MCC) was used as filler for making the geopolymer brick. The sand falls in Zone II as per IS 383–1786. Plastic waste is obtained from waste generated by polyethylene terephthalate (PET) bottles, which are shredded in the form of fine powder [64]. The bottles were collected, washed, and air dried, after which they were cut into small pieces with the help of scissors and finally shredded into small, fine aggregates and sieved so that they almost match the size of recycled sand and lie in the range of less than 100% particles passes the 4.75 mm sieve and retained on 15 mm sieve size. The particles of recycled sand are granular and round while PET plastic waste has a sharp irregular shape. The particle size distribution curve for PET plastic waste and recycled sand is shown in Fig. 4.

Figure 4
figure 4

Particle size curve for recycled sand and PET plastic waste

Full size image

Mix design

The detailed microstructural understanding of the raw material helps us to develop the optimum mix design with three aluminosilicate waste materials. The production process for making geopolymer mix is quite different from conventional concrete mix design. First, the alkaline solution is prepared with the combination of 12 M NaOH and Na2SiO3, where they are mixed and kept for 24 h as the mixing of NaOH and deionized water is exothermic, and the temperature during mixing can reach as high as 80–90 °C as recorded and For further information on the process of mix design preparation, please refer our previously published article [65] (Fig. 5).

Figure 5
figure 5

The research methodology adopted for this study

Full size image

The ratio of sodium silicate to sodium hydroxide was fixed at 2. Three aluminosilicate waste materials were utilised as precursors, with 60% of RHA, 20% of GGBS, and 20% of red mud for making geopolymer brick. The alkaline-to-binder ratio was fixed at 0.45. The ratio of precursor to recycled sand was set at 1:3. The plastic waste obtained from PET was used as a partial replacement of sand, with a replacement percentage of 10% at each level by weight of sand. The detailed mix design is presented in Table 5.

Table 5 Nomenclature and details of mix design for geopolymer bricks
Full size table

Preparation of geopolymer brick samples

Firstly, all the dry source material, which includes precursor and recycled sand in a ratio of 1:3, is mixed in the pan with a capacity of 100 kg with a 7.5 HP motor for 3–5 min for proper mixing at a rate of 25 RPM. According to the mix design, the percentage of plastic waste is added to the different lots and mixed for another 2–3 min for proper mixing. Once the dry mixing is completed, an alkaline solution is added, and again, it is mixed for another 3–5 min to produce a proper workable mix. After the mix has been prepared, it is carried by the conveyor belt of 18 feet with a 2 HP motor attached to it and placed in a fully automatic brick-making machine with a compression capacity of 80 tons, producing 8 bricks per stroke. The step-by-step procedure for making geopolymer bricks is shown in Fig. 6 once geopolymer bricks are produced by a fully automatic machine, they are kept for ambient curing and allowed to dry properly. After the bricks were fully matured for 28 days, their physical, mechanical, and microstructural properties were tested and reported in Sect. "Result and analysis" of this research paper.

Figure 6
figure 6

The process involved in making geopolymer bricks

Full size image

Physical–mechanical testing procedure

The physio-mechanical testing methodology for this research paper involves a systematic approach to conduct the different tests required to assess the properties of geopolymer bricks. The standard procedure adopted for brick testing is mentioned in Table 6. A flow table test was conducted to assess the fresh behaviour of geopolymer brick as per the ASTM C1437 [66]. The main aim of conducting the flow table test was to assess the consistency of mix and workability. The Experimental methodology adopted is shown in Fig. 7. Compressive strength one of the most basic strength testing parameters was conducted as per the IS 3495-(Part 1)-1992 [67]. Three specimens were used to calculate the mean strength at each level as per the design mix for 28 days. The frog in the sample is filled with mortar in a 1:1 ratio, thus three days of curing are needed before testing can begin. The automated compressive testing equipment, supplied by Aimil Ltd., is employed with a capacity of 3000 KN, and the rate of loading applied is 140 kg/cm2. Equation 1 may be used to compute the compressive strength.

$${\text{Compressive}}\;{\text{Strength}} = \frac{{{\text{Maximum}}\;{\text{Load}}\;{\text{at}}\;{\text{failure}}\;{\text{in}}\;N}}{{{\text{Average}}\;{\text{area}}\;{\text{of}}\;{\text{Loading}}\;{\text{face}}\;{\text{in}}\;{\text{mm}}^{2} }}$$
(1)
Table 6 Standard codes used for physio-mechanical testing
Full size table
Figure 7
figure 7

Experimental methodology adopted for making geopolymer brick

Full size image

Water absorption and dry density were carried out as per the IS 3495-(Part 2)-1992 and IS 2185 (Part 1): 2005. The brick samples are initially kept in the oven at 105 °C for 24 h, then allowed to cool slowly until they reach room temperature. They are then weighed using a sensitive weighing balance and recorded as (W1). Subsequently, the samples are placed in water and allowed to boil for two hours. Once again, they are allowed to cool down and reach room temperature. At this point, we measure the suspended weight of the brick as (W2), which is used to measure the apparent porosity. Lastly, the samples are submerged in water for 24 h, with any excess water removed to obtain the submerged weight of the sample (W3). Equation (2)may be used to determine water absorption.

$${\text{Water}}\;{\text{Absorption}}\; = \;\frac{{W_{3} - W_{1} }}{{W_{1} }} {\text{X}} 100$$
(2)

W1 is the weight of the brick after oven drying, W2 is the suspended weight of the brick and W3 is the weight of the brick noted after 24 h of submergence. An impact resistance test is conducted for geopolymer brick to evaluate the impact energy. A 2.5 kg steel ball with a diameter of 65 mm and a 4.5 kg cast iron ball was utilised in the drop weight test. The steel ball was positioned such that it rested on the base plate within the positioning lugs and was held in the centre of the specimen. The test included repeatedly dropping a cast iron ball from a predetermined distance of 450 mm onto a steel ball. N1 was the number of drops for the first visible fracture, and N2 represented the number of drops for the specimen failing. This kind of test has been performed by several researchers in the past [68,69,70]. Equations (3) and (4) were used to determine the impact energy at the initial crack (Ei) and the final crack (Ef).

$$E_{{\text{i}}} = N_{1} mgh$$
(3)
$$E_{{\text{f}}} = N_{2} mgh$$
(4)

The purpose of the thermal conductivity test is to evaluate a material's capacity to transmit heat, which is an essential quality for applications requiring heat transfer or thermal insulation. Bricks are essential components of a building that play a crucial role in the thermal performance of a building. Thermal conductivity was conducted as per ASTM C177 with the help of EIE-7013 hot guard plate apparatus thermal conductivity apparatus. The average value of thermal conductivity was recorded by measuring three specimens. For the ultrasonic pulse velocity (UPV) test, a fresh geopolymer brick specimen was used and tested length-wise by using a TICO ultrasonic testing instrument. The distance between two ultrasonic transducers (transmitter and receiver) mounted on the surface of the geopolymer specimen was 190 mm. The test was conducted as per IS 516-(Part 5)-2020.

Microstructural analysis

A brick microstructural study offers important information about the characteristics and composition of the material. A sample was taken from the core of the brick with a size of 1× 1 × 1 cm and was thoroughly polished with silicon carbide paper having a range of 500–2000 μ with a disc polisher. High-resolution imaging is possible using the scanning electron microscopy (SEM) model JSM6100 (Jeol), which shows the porosity, grain boundaries, and surface morphology of bricks. By identifying the crystalline phases that are present in the brick, X-ray diffraction (XRD) helps to clarify its mineralogical makeup. Small-angle X-ray scattering by the ANTON PAAR model SAXSPACE with Panalytical's X.Pert Pro was used for XRD analysis. The XRD peak was measured between 5 and 90º 2θ, with a step size increase of.05. The generator set was modulated at 30 mA and 40 kV, and X'pert Highscore software was used to evaluate the XRD peak data. By providing details on molecular vibrations, Fourier transform infrared spectroscopy (FTIR) aids in the identification of certain functional groups within the structure of bricks. An FTIR spectrophotometer model RZX (Perkin Elmer) was used, measuring a wavenumber from 400 to 4100 cm−1. Combining SEM, XRD, and FTIR analyses enables a comprehensive exploration of the brick's microstructure, aiding in the characterisation and quality assessment of the material. This multidimensional approach is crucial for understanding the physical and chemical aspects that influence the performance and durability of bricks in various applications.

Statistical analysis using machine learning algorithms

To predict and evaluate our findings, we have employed five different machine learning algorithms, i.e. linear regression, decision tree, random forest, support vector machine (SVM), and k-nearest neighbours (KNN). We used three main parameters to evaluate performance, which include R-squared (R2), mean squared error (MSE), and mean absolute error (MAE). R2 measures the percentage of variation in the dependent variable explained by the model, which is computed as the explained variance divided by the total variance. MSE calculates the average squared difference between expected and actual values, giving a complete picture of prediction accuracy. MAE is the average absolute difference between anticipated and actual values, and it provides information on model accuracy. These parameters, taken together, allow us to evaluate the predictive capabilities and overall effectiveness of each machine learning algorithm in our investigation. R2, MSE and MAE can be calculated by Eqs. (5), (6) and (7).

$$R^{2} = 1 - \frac{{\mathop \sum \nolimits_{i = 1}^{n} \left( {z_{i} - \hat{z}} \right)^{2} }}{{\mathop \sum \nolimits_{i = 1}^{n} \left( {z_{i} - \overline{z}} \right)^{2} }}$$
(5)

n is the number of observations,

\({z}_{i}\) is the actual value of the dependent variable for observation i,

\(\widehat{z}\) is the predicted value for observation i,

\(\overline{z }\) is the mean of the actual values.

$${\text{MSE}} = \frac{1}{n}\mathop \sum \limits_{i = 1}^{n} \left( {z_{i} - \hat{z}_{i} } \right)^{2}$$
(6)

n is the number of observations,

\({z}_{i}\) is the actual value of the dependent variable for observation i,

\(\widehat{{z}_{i}}\) is the predicted value for observation i.

$${\text{MAE}} = \frac{1}{n} \mathop \sum \limits_{i = 1}^{n} \left| {z_{i} - \hat{z}_{i} } \right|$$
(7)

n is the number of observations,

pedicle \({z}_{i}\) is the actual value of the dependent variable for observation i,

\({\widehat{z}}_{i}\) is the predicted value for observation i.

A robust analysis of compressive strength in geopolymer materials involves a diverse set of machine learning algorithms, as evidenced by the studies presented in Table 7. Each research effort contributes unique insights, combining different algorithms and outcome measures to advance our understanding of geopolymer performance.

Table 7 Overview of machine learning algorithms used in analysing compressive strength of geopolymer materials
Full size table

Result and analysis

Workability analysis

Workability is defined as the ease with which the material can be moulded into the desired form. Figure 8 shows the effect of waste plastic replacement from 0 to 100% on workability. The workability is visibly affected after the replacement of recycled washed sand plastic waste, especially after 30% of replacement. Further, only a 4.43% change was recorded up to 30% replacement after the addition of more waste plastic the flow value decreases due to the mechanical adhesion and bonding between crushed PET plastic waste as the edges are sharp and rough making it suitable for greater adhesion. A similar fall in workability with the addition of plastic waste in the form of PET aggregate was observed by Lazorenko et al. [56] and Bhogayata et al. [80] also reported a decrease in workability with the addition of packaging waste plastic was incorporated in geopolymer concrete both studies supported their decrease in workability by strong frictional attachment of plastic waste in geopolymer matrix as seen by SEM images. In this research study, a fall of 14.75% was noticed when recycled washed sand was completely replaced by 100% PET plastic waste aggregates. A flow fluctuated between 305 and 260 mm flow diameter. The SEM images of Fig. 15 show that with increased percentage of plastic waste in the Geopolymer matrix provides a rough adhesion between the surface which became the main cause of a decrease in workability.

Figure 8
figure 8

Workability measurement showing the effect of the usage of plastic waste

Full size image

Compressive strength

The compressive strength of brick is one of the most viable properties to be tested to determine the strength-carrying ability. Figure 9 shows the compressive strength along with the percentage change labelled at the top. The decrease in compressive strength is quite significant when recycled washed sand was replaced with plastic waste from 0 to 100% it can be seen clearly that almost 75% strength was reduced when a complete replacement was performed, however sudden reduction of strength was noted after 10% of replacement up to 10% only 6% of strength was lost. The geopolymer brick made with the control mix has a compressive strength of 27.32 MPa while at 50% replacement, the compressive strength is 16.79 MPa which falls to 6.8 MPa when 100% replacement by volume of plastic waste was done with recycled washed sand. The decrease in the compressive strength can be inferred due to the low strength of plastic aggregate when compared with sand. The SEM, XRD, and FTIR analysis also projected the weak bonding with plastic waste increment in the mix. The same decrease in compressive strength by the addition of different plastic waste is reported by many researchers in their study [56, 81, 82]. The trend of decreased compressive strength in geopolymer is similar to cement concrete which has been studied by a significant number of researchers in the past [83, 84]. This decrease in strength is mainly due to the decrease in the bulk density of the composites, due to the decrease in adhesive strength between the surfaces of the plastic particles and geopolymer paste. this behaviour can be attributed to the different interactions between the hydrophobic organic material particles of polyethylene terephthalate when compared to sand particles. The hydrophobic organic nature of the PET particles prevents good adhesion with the strongly polar aluminosilicate matrix. However, each brick passed the minimum criteria of compressive strength which is 3.5 MPa as per IS 1077 (1992) [85].

Figure 9
figure 9

Compressive strength measurement shows the effect of the usage of plastic waste

Full size image

Dry density

Figure 10 shows the results obtained for dry density when recycled washed sand was replaced with PET plastic waste. The maximum dry density was noted by the control mix containing 0% plastic aggregate as 1932 kg/m3 with 50% replacement it was 1434.35 kg/m3 showing a decline of almost 25% and at 100% replacement the dry density was 936.71 kg/m3 with a total decline of 51%. The decrease in dry density can be explained by an increment of lightweight aggregate in the form of plastic waste increment in the mix which leads to a fall in dry density. The light weight of geopolymer brick is an advantageous aspect from a structural point of view. The decrease in dry weight comes in the form of one of the most positive impacts of this research study. The decrease in dry density with an increased percentage of plastic waste is noted by Atoyebi et al. [86] and Lazorenko et al. [56]. The decreasing trend of dry density is observed in cement concrete as recorded and reported by Shiuly et al. [87].

Figure 10
figure 10

Dry density measurement shows the effect of the usage of plastic waste

Full size image

Water absorption

Water absorption is another essential and elementary test conducted in the usage of brick. The geopolymer bricks perform well in water absorption when compared with conventional brick due to the internal structure of geopolymer matrix formed by the polymeric reaction of sialate formation –(–Si–O–Al–O–)- as reported by Shilar et al. [88]. The water absorption for geopolymer brick containing plastic waste as volume replacement of recycled washed sand is shown in Fig. 11 which depicts that as the plastic waste replacement percentage increases the amount of water absorption also increases with an overall change of almost 140% when compared with the control mix containing 0% plastic waste. The water absorption percentages at 0, 50, and 100% are 5.68, 9.7 and 13.73% the increase can be explained by the increase in void space structure in the mix which has been reported by Kunthawatwong et al. [89]. However, the water absorption percentage laid by Indian standard code IS 3495 (part 1)-1992[90] of 20% for first-class brick has been passed by each sample which indicates its practicality to be used in field applications.

Figure 11
figure 11

Water absorption measurement showing the effect of usage of plastic waste

Full size image

Impact resistance

Figure 12 shows the combined impact resistance with the impact energy at the initial crack (Ei) and Ef impact energy at the final cracking with the replacement of sand with plastic waste it shows the better impact resistance capacity. The Ei and Ef are calculated in terms of the Number of drops (N1) and (N2) as mentioned in detail in Sect. "Physical–mechanical testing procedure" of this research paper. As plastic percentage vary from 0 to 100% replacement the value of Ei varies from 258.25 to 2443.43 J while Ef varies from 437.04 to 2622.21 J. This increased energy resistance is due to the number of drop increments that happened to cause cracks which happen due to the better elastic behaviour of plastic waste aggregate when replaced with recycled washed sand. The number of drops for 0, 50, and 100% replacement in the case of N1 is 13, 68 and 123 while for N2 it is 22,77 and 132. The increased percentage of plastic aggregate in the mix ultimately amplified the impact resistance which has been also recorded by previous researchers Bhogayata et al. [80]. The same trend of increased impact resistance was observed in conventional cement concrete when replaced with plastic waste as investigated by Saxena et al. [68]. The higher elastic and crack-resisting properties of raw plastic aggregate slow down crack generation and final specimen failure. The ductility behaviour of the concrete specimens was enhanced by the flexibility of the plastic waste aggregate, increasing energy absorption capacity.

Figure 12
figure 12

Impact measurement showing the effect of the usage of plastic waste

Full size image

Ultrasonic pulse velocity test

During surface sounding, the ultrasonic pulse velocity test that was employed in this study consisted of monitoring the amount of time it took for an ultrasonic pulse to travel from an emitter to a receiver inside a geopolymer specimen. It is one of the most widely adopted Non-destructive testing methods. UPV enables us to understand the nature and quantify the amount of void space in the samples [91]. All samples of geopolymer brick were tested by UPV instrument by the direct method which is the most significant method of testing by placing the emitter and receiver facing each other. Figure 13 shows the results obtained by the ultrasonic pulse velocity test. The test results revealed that with an increased percentage of replacement of plastic waste with recycled sand, the UPV decreased from 4760 to 2420 m/s with an overall percentage change of 50% the decremental trend can be explained by increasing cavity in the internal structure leading to an increase of acoustic impedance for UPV. The results of UPV testing show a strong compliance with compressive strength. The same has been investigated by multiple researchers [56, 68, 92].

Figure 13
figure 13

UPV measurement showing the effect of the usage of plastic waste

Full size image

Thermal conductivity

Thermal conductivity is one of the major properties that are tested in the bricks as it provides the picture of the thermal insulation of the building by showing the amount and pace with which it can lose the heat in and out of the buildings. The normal thermal conductivity of brick varies approximately from 1.0 ± 0.4 W/(m∙K) as reported by Kubiś et al. [93]. Figure 14 shows the results obtained by the thermal conductivity test on geopolymer brick showing the decrease in the thermal conductivity when percentage replacement increases from 0 to 100%. The value of thermal conductivity at 0, 50 and 100% are 1.3, 0.95 and 0.55 W/(m∙K) showing and overall percentage change of 58%. The reason for the decrease in thermal conductivity due to the increasing percentage can be understood by the fact that plastic waste has a relatively low thermal conductivity which varies from 0.13 to 0.24W/(m∙K) as stated by Poonyakan et al. [94] The results of low thermal conductivity are positive in sense of utilising these geopolymer brick for thermal insulation and low level of heat loss in the buildings.

Figure 14
figure 14

Thermal conductivity measurement showing the effect of usage of plastic waste

Full size image

Microstructural analysis

Microstructure analysis was performed on five samples at percentages of 0, 20, 50 80, and 100% of plastic waste replacement with sample ID named GW100P0, GW80P20, GW50P50, GW80P20, and GW0P100. A microstructural study was conducted with the help of SEM, XRD, and FTIR techniques in order to evaluate and correlate the results obtained in physio-mechanical analysis. SEM is the most widely adopted technique for understanding the changes happening at the microscopic level leading to changes in mechanical properties. SEM images are shown in Fig. 15 and it can be analysed that with a percentage increase in plastic waste, the microcracks and voids are increasing causing defects at the micro surface which is causing the decrease in the compressive strength and other properties of nature. Some microcracks are due to the internal stresses that samples have undergone during physical–mechanical testing. For GW100P0 containing 0% plastic, there is a homogenous surface representing a continuous gel-like formation for the Si-Al matrix. It can also be seen in Fig. 15 that as the plastic waste increases the interfacial transition zone (ITZ) becomes weaker and develops cracks and voids creating more porous which aligns with the results obtained in physical mechanical parameters and some unreactive precursors can be seen in geopolymer matrix.

Figure 15
figure 15

Surface image and SEM analysis of 5 samples selected for microstructural analysis

Full size image

Further microstructural study was carried out using XRD. Figure 16 shows the combined XRD analysis of all five samples that were selected. Recycled sand which is mostly unreactive and does not contribute to geopolymerization featured large peaks as depicted in Fig. 17 and as the plastic percentage increases the presence of quartz decreases as replacement of sand which results in a decrease in the peak reflected in the peaks. At a smaller angle of 10–25°, the formation of CASH calcium aluminium silicate hydrate is seen and the same is reflected by different researchers [63, 95]. N–A–S–H gel formation can be seen in samples by shifting of peak towards the right which is one of the main causes of strength development and binding in the geopolymer. GW100P0 shows the highest no. of N-A-S–H peak which is responsible for the maximum compressive strength and subsequently sample that has a higher percentage of plastic waste has lower N–A–S–H peaks and similar results are reported by researchers [1, 62, 96].

Figure 16
figure 16

XRD analysis of a selected sample of geopolymer brick at 28 days

Full size image
Figure 17
figure 17

FTIR analysis of a selected sample of geopolymer brick at 28 days

Full size image

FTIR was conducted for the 5 samples selected, analysed, and represented in Fig. 17 H–O–H bands are observed in the range of 3300–3600 cm−1 wavenumber as reported by [28, 31] and Si–O–Si or Al–O–Si bond [32, 33] were found in the range of 997–1066 cm−1 as plastic percentage increases the shift of peak is noted and intensity of peak also increase which results into lower strength as stretching of bonds are observed with higher percentage of plastic waste and also random Si-Al were found at lower wavenumber of 450–490 cm−1 as plastic percentage increases.

Correlation matrix

The correlation matrix with the help of a parametric map is one of the most effective tools for understanding the relation between different performance parameters that were chosen to test the geopolymer bricks. Figure 18 shows the correlation between Flow Diameter, Dry Density (kg/m3), Water Absorption (%), Compressive Strength, Ei (Joules) and Ef (Joules), Ultrasonic Pulse Velocity (m/s), Thermal conductivity W/mK. The value that is as close as one depicts there is a direct correlation between the parameters while the negative sign indicates the inverse or negative correlation. From Fig. 18. it can be easily validated that compressive strength has an almost direct relation with flow diameter, dry density, UPV, and thermal conductivity showing a value near 1 which validates the fact that as with the increase in plastic waste percentage replacement from 0 to 100% the compressive strength decreases and other value also fall either due to incremental void surface with cavity or due to low elastic modules of plastic waste and it is visible by the microscopic analysis also as discussed in Sect. "Microstructural analysis".

Figure 18
figure 18

Correlation parametric map showing the relation between the different properties of geopolymer brick

Full size image

Performance by machine learning algorithms

In order to get a deeper understanding and behaviour of our results with a modern applicative approach, five different kinds of machine learning algorithms were used such as Linear Regression (LR), Decision tree, Random Forest, Support vector machine (SVM), and K-Nearest Neighbour (KNN). The data set obtained by laboratory experiment was trained and tested and evaluated three useful performance indicators which are mean square error (MSE), mean absolute error (MAE), and R-Square (R2). MSE & MAE depict the error between the actual value and predictive value to assess this it should be as minimal as possible for the better performance of the algorithms. From Fig. 19. it can be seen clearly that Linear regression has a minimum value of 0.12 and 0.31 for MSE and MAE while the maximum value of R-square (R2) is 99.81% while the least performing machine learning algorithms were support vector machine (SVM) where the MSE& MAE was highest with a value of 42.83 and 5.39 and R2 was 33.56% which is quite low and showing that the SVM will produce large error between the experimental value and predictive values.

Figure 19
figure 19

Performance of different machine learning algorithms

Full size image

This analysis enables us to give a picture that plastic waste replacement in geopolymer bricks has a linear relationship between the dependent variables and independent variables. Table 8 basically enables us to predict the compressive strength at different percentages of plastic waste replacement. The prediction is made at 5% of plastic waste replacement on compressive strength by linear regression and support vector machine (SVM). These techniques are quite helpful in predicting the percentage variation at different levels very accurately. Overall error found in predictive the compressive strength by linear regression was found to be 1.13% and that by SVM is 7.92%. Similarly, we can predict the compressive strength of geopolymer bricks at any percentage for the replacement of plastic waste and other parameters.

Table 8 Prediction table of compressive strength at different percentages of polyethylene waste replacement
Full size table

Discussion on advantages, disadvantages, and application scopes

In this section, a comprehensive discussion on the advantages, disadvantages, and potential application scopes arising from the incorporation of PET plastic waste in geopolymer bricks is discussed below. Figure 20 visually summarizes the comprehensive analysis of the advantages and disadvantages associated with integrating PET plastic waste in geopolymer bricks.

Figure 20
figure 20

Advantages and disadvantages of PET plastic waste in geopolymer bricks

Full size image

Advantages of geopolymer bricks with PET plastic waste

  • Environmental sustainability: The incorporation of PET plastic waste in geopolymer bricks provides an eco-friendly alternative by recycling widely available plastic waste, contributing to the reduction of environmental pollution.

  • Improved impact resistance: The study indicates a significant improvement in impact resistance, ductility, and energy absorption capacity of geopolymer bricks with increased plastic waste. This enhanced property can be advantageous in construction applications where impact resistance is crucial.

  • Thermal insulation: Geopolymer bricks with PET plastic waste exhibit better thermal conductivity, making them suitable for applications requiring insulation. The decrease in thermal conductivity aligns with the trend of increased plastic waste percentage, making these bricks energy-efficient.

  • Utilisation of PET waste: The research demonstrates a practical use for PET plastic waste, addressing concerns related to the disposal of this specific type of plastic. This utilisation aligns with sustainability goals and waste reduction initiatives.

Disadvantages and considerations

  • Decrease in workability and compressive strength: The study identifies a decrease in workability and compressive strength as the percentage of PET plastic waste increases. This reduction is attributed to increased voids and poor interaction between plastic waste and the geopolymer matrix.

  • Dry density reduction: The dry density of geopolymer bricks decreases with the incorporation of plastic waste, potentially impacting structural considerations. However, this decrease is identified as advantageous for certain applications, providing lightweight properties.

  • Water absorption increase: Geopolymer bricks with higher percentages of PET waste show an increase in water absorption. While this may affect certain applications, all samples meet the water absorption standards for practical use in the field.

  • Microstructural changes: Microstructural analysis reveals the formation of microcracks, voids, and weaker interfacial transition zones with increased plastic waste. These changes contribute to the observed alterations in physical–mechanical properties.

Application scopes

  • Construction: Geopolymer bricks with PET plastic waste can find applications in construction where enhanced impact resistance and thermal insulation are desired, such as in non-load-bearing walls, partitions, or facades.

  • Thermal Insulation Applications: The low thermal conductivity of these bricks makes them suitable for applications where thermal insulation is a primary consideration, contributing to energy-efficient buildings.

  • Sustainable Building Practices: The utilisation of PET plastic waste aligns with sustainable building practices, appealing to environmentally conscious construction projects.

Conclusion

  • In this research study, we have investigated the impact of PET plastic waste in geopolymer bricks and studied the physical–mechanical and microstructural properties of the bricks, along with the use of machine learning algorithms to correlate and predict the performance of geopolymer bricks. The following conclusions were drawn, as stated below:

  • As the percentage of plastic waste increases in the geopolymer brick from 0 to 100%, a total decrease of 14.75% is recorded in the workability, and almost 75% of the reduction is recorded in the compressive strength due to an increase in the voids and poor interaction of plastic waste with the geopolymer matrix.

  • The same trends of decrement were observed in the dry density as well, and the water absorption of bricks increased with the increase in the percentage of plastic waste. Water absorption increased from 5.68 to 13.73% with complete replacement of sand.

  • The impact resistance of the bricks improves with the incorporation of plastic waste percentage due to the improvement in ductility, energy absorption capacity, and thermal conductivity, which is also enhanced with 100% replacement, showing a total improvement of 57%, making it a better choice for thermal insulation.

  • The combination of SEM, XRD, and FTIR enables us to understand the actual cause of change in the physical–mechanical properties by knowing the real cause of change in the properties. With an increase in plastic waste, there is an increase in pore size, voids, microcracks, and fissures, as revealed in SEM analysis, and lower NASH was seen in the XRD at a higher percentage of plastic. An increase in intensity in peaks at different wave numbers was seen with an increase in plastic waste showing weaker bonds aligning with the results obtained in physiochemical testing.

  • Machine learning algorithms were effectively implemented for the prediction of strength parameters, and linear regression was found to be the most effective among all five machine learning algorithms, with a maximum efficacy of almost 100% and the highest R-square of.998.

  • Despite the lower load capacity of geopolymer bricks with the incorporation of polyethylene terephthalate waste, they show increased impact resistance and ductility and can be effectively used for better thermal insulation, above all, they can be implemented as a sustainable substitute for environmentally friendly bricks.