Development of Green Geopolymer Using Agricultural and Industrial Waste Materials with High Water Absorbency

: Geopolymer is a promising porous material that can be used for evaporative cooling applications. Developing a greener geopolymer using agricultural and industrial wastes is a promising research area. In this study, we utilize rice husk (RH), rice husk ash (RHA), metakaolin (MK), ground granulated blast furnace slag (GGBS), and palm oil fuel ash (POFA) to prepare geopolymer pastes, with alkali liquid as an activator. Many geopolymer samples have been prepared as per the Design of Experimental software (DOE), and its corresponding response surface mode and central composite design and later they were characterized. The samples were cured in an oven for 2 h at 80 ◦ C, and thereafter stored at room temperature (~25–30 ◦ C) prior to being tested for its water absorption and compressive strength. The effect of the different composition of precursors on water absorption, density, porosity, and the compressive strength of the prepared geopolymers have been investigated. The results showed that the compressive strength of geopolymers is directly proportional to the ratio of the alkali liquid. Post-optimization, the best geopolymer paste mixture was conﬁrmed to contain 10% of RH, 15% RHA, 35% MK, 10% POFA and 30% of GGBS, with 72% desirability for maximum water absorption (~38%) and compressive strength (4.9 MPa). The results conﬁrmed its applicability for evaporative cooling.


Introduction
Employing passive evaporative cooling technology as an energy-efficient approach to partially decreasing energy consumption in buildings is becoming popular due to improvements in materials and designs [1].
Geopolymers, or inorganic polymers, have been touted as unique engineering materials that have the potential to form elements that are environmentally sustainable for the construction industry due to their respectable mechanical properties, low shrinkage, fire resistance, and low energy consumption [2][3][4]. Geopolymers are regarded as an environmentally friendly material, owing to

Materials
Rice husk (RH) was sourced from a local rice mill during the milling season in Malaysia. Rice husk ash (RHA) was prepared in the concrete and structure laboratory in University Kebangsaan Malaysia (UKM) using a specialized furnace [13]. Metakaolin was purchased from Burgess, Greenville, SC, USA. Palm oil fuel ash (POFA), which includes fibers, nut shells, and empty fruit bunches, were gathered from a local plant called Seri Ulu Langat Palm Oil Mill Sdn. Bhd., Dengkil, Selangor, Malaysia. It was then dried in an oven for 24 h at 100 • C to remove any remaining moisture and then sieved using a 300-µm sieve. The ground granulated blast-furnace slag (GGBS), which is a byproduct of steel plants, was collected from Slag Cement Sdn. Bhd., Pulau Indah, Selangor, Malaysia. Afterwards, RH, RHA, and POFA were grounded using a ball mill until the mass of the fine particles were retained on a sieve size No. 325 (an aperture of 75 µm) at 1-3%. The MK and slag were used as is. Sodium silicate solutions (Na 2 SiO 3 ) (10.6% Na 2 O, 26.5% SiO 2 , and 62.9% H 2 O) were purchased from Sigma-Aldrich (St. Louis, MO, USA), while sodium hydroxide and pellet with analytical reagents were purchased from ACS and used as alkali activators. Sodium silicate solution or water glass (WG) and sodium hydroxide were used in the form of a solution to make a part of the mixture. WG was used as received, while different concentrations of NaOH were prepared to study its effect on the prepared geopolymer samples. The chemical compositions of RH, RHA, MK, POFA, slag, WG, and the specified geopolymer sample before/after casting (dry mix before and after activation) were determined using X-ray fluorescence (XRF) (S8 Tiger, Bruker, Billerica, MA, USA) ( Table 1). Loss on ignition (LOI), which is an indicator for carbon content, was measured according to ASTM C 114-04 [14]. The XRD patterns for raw materials and geopolymer were determined using Bruker DB-Advance X-ray Diffractometer (Bruker), while the functional groups of the samples were determined using FTIR spectrophotometer (Perkin Elmer-400, PerkinElmer, Waltham, MA, USA). The specific surface area, pore size, pore volume, and pore diameter of the samples were determined via the Brunauer-Emmet-Teller (BET) method using a nitrogen adsorption instrument (Micrometics ASAP 2010, Micrometrics instrument, Norcross, GA, USA). The samples were degassed at 150 • C for 6 h prior to the analyses. Pore size distribution was calculated from the adsorption-desorption of the isotherms using the Barret-Joyner-Halenda (BJH) model. Table 1. Chemical composition of rice husk (RH), rice husk ash (RHA), metakaolin (MK), palm oil fuel ash (POFA), Slag, water glass (WG), and selected geopolymer sample before and after casting (dry mix before and after activation) using XRF (mass %).

Geopolymer Preparation and Characterization
Taking into account the ratio of alkali liquid to solid, the ratio of sodium silicate to sodium hydroxide (Na 2 SiO 3 /NaOH), and the concentrations of sodium hydroxide on geopolymers performance, in the first step, the effects of these factors on a selected sample (10% RH, 15% RHA, 35% MK, 10% POFA and 30% Slag) based on literature review and trial-and-error was investigated. In this step, the Design of Experimental software (DOE) version 7 (Stat-Ease cooperation, Minneapolis, MN, USA, 2015), response surface mode, and central composite design software were employed for studying these factors and determining water absorption, density, and compressive strength for the purpose of obtaining the optimal molar ratio and preparing the geopolymer paste. As per the literature, the ratios of alkali liquid to solid were set at 0.3, 0.4, 0.5; the ratios of sodium silicate to sodium hydroxide (Na 2 SiO 3 /NaOH) were set at 1, 1.5, and 2.5; the concentrations of sodium hydroxide (NaOH) were set at 8, 10, 14 M; all of which were within range of the DOE. The results are presented in Table 2, while the analysis of variance (ANOVA) table is tabulated in Table 3 (Figures S1-S3).   ANOVA values confirmed that all main factors significantly influence the three responses. The model's F-value of 5.38 implies that it is significant. There is only a 0.13% chance that a "Model F-Value" this large could occur due to noise. The "Lack of Fit F-value" of 2.261 implies that the Lack of Fit is not significant relative to pure error ( Table 3).
The determination of the optimum value for control variables (factors) is one of the main objectives of RSM, which can have maximum (or a minimum) response over a specific region of interest. Having a "good"-fitting model is necessary to adequately represent the mean response because such a model can be utilized to determine the optimum value [15].
In the second step, a total of 30 mixtures were prepared at different percentages of RH, RHA, MK, POFA, and GGBS using NaOH (alkali activation) based on Design of Experimental software (DOE) version 7, response surface mode, and central composite design, by introducing each component's limitations and factors ( Table 4). The samples were manually mixed, and the pastes were poured into a 20 × 20 × 20 mm 3 molds to set, forming the specimens that will be characterized. For each geopolymer paste mixture, eighteen specimens were cast in a 20 mm 3 cube to determine its compressive strength, density, porosity, and water absorption capabilities. The samples in the molds were tightly wrapped in plastic films to prevent any loss of moisture. All of the samples were heated in an electric oven at 80 • C for 2 h after casting. Later, the prepared samples were kept at an ambient temperature of 30 ± 2 • C prior to being characterized. The water absorption and porosity of the specimens were calculated based on ASTM D 570-98 and ASTM C 642-06, respectively. The density of the samples was measured using an Electronic Densimeter (model MD-300S, Alfa Mirage, Miyakojima-ku, Osaka, Japan). The capillary rise of the samples was measured by immersing one side of the prepared cube in water at room temperature at various times, as per ASTM 1585. The compressive strength of mortar is one of the principal parameters that represent the quality of the binder, where its used to gauge the potential success of geopolymerization. The test is regarded as being cost effective and simple to administer, and the development of strength is used as a primary measure of its utility for application in the construction industry [16]. It was determined using a 20 mm 3 cube sample, as per ASTM C 109 (2009) testing standards, with a PC base automatic SSM-AJ-10KN compression testing machine. An average was taken from three values for samples after 3, 7, and 28 days post-curing.
To investigate the effect of the concentration of NaOH on compressive strength in the prepared geopolymers, two samples were randomly selected and prepared at different concentrations of NaOH (8, 10, and 14 M). The compressive strength of the samples has been tested after curing at room temperature for 28 days.
The samples' components and their water absorption, porosity, density, and compressive strengths are tabulated in Table 3.
As per Table 4, the residual compressive strength of the specimens is inversely proportional to both water absorption and porosity. Sample 27 had a residual strength of 4.99 MPa, corresponding to the lowest water absorption (30.11%). Contrarily, sample 14, at 48.4% water absorption, reported the lowest residual compressive strength of 2.43 MPa.

NaOH Concentration Effect
The results from Table 2 (validated with ANOVA) showed that the best ratio was 0.5 for alkali liquid to solid, and 2.5 for sodium silicate to sodium hydroxide. The concentration 10 M for NaOH agrees with [9,11,[17][18][19][20].
According to the results, the compressive strength is maximum when NaOH concentration is at 10 M. This could be due to the higher dissolution of initial solid materials at high concentrations of NaOH, which accelerates the geopolymerization reaction, leading to the higher compressive strength of geopolymers [21]. The solubility of aluminosilicate is directly proportional to the concentration of NaOH [12,22]. Increased additions of NaOH accelerated the chemical dissolution but suppressed the formation of ettringite and CH in the course of binder hydration [23]. Reduction in the CH content resulted in superior strength and durability [24]. However, compressive strength is inversely proportional to the concentration of NaOH at 10-14 M. This could be due to the presence of excess hydroxide ion concentration [25] and excess Na + ions in the framework [22,26], which precipitates aluminosilicate gel during the early stages of the reaction, thus preventing geopolymerization, culminating in lower compressive strengths. The higher the alkalinity of the hydration water, the slower the rate of the hydration [27].

Curing Time Effect
The effect of curing time on the compressive strength of the prepared geopolymers samples (using 10 M NaOH) has been studied.
The compressive strength for geopolymer samples that were cured at room temperature for 3, 7, and 28 days are depicted in Figure 1. As can be seen, increasing the curing time from 3 to 28 days increased the compressive strength. The results agree with [22,28]. It is believed that the increment in compressive strength alongside curing time is related to the improvement of the geopolymerization process. NaOH, which accelerates the geopolymerization reaction, leading to the higher compressive strength of geopolymers [21]. The solubility of aluminosilicate is directly proportional to the concentration of NaOH [12,22]. Increased additions of NaOH accelerated the chemical dissolution but suppressed the formation of ettringite and CH in the course of binder hydration [23]. Reduction in the CH content resulted in superior strength and durability [24]. However, compressive strength is inversely proportional to the concentration of NaOH at 10-14 M. This could be due to the presence of excess hydroxide ion concentration [25] and excess Na + ions in the framework [22,26], which precipitates aluminosilicate gel during the early stages of the reaction, thus preventing geopolymerization, culminating in lower compressive strengths. The higher the alkalinity of the hydration water, the slower the rate of the hydration [27].

Curing Time Effect
The effect of curing time on the compressive strength of the prepared geopolymers samples (using 10 M NaOH) has been studied.
The compressive strength for geopolymer samples that were cured at room temperature for 3, 7, and 28 days are depicted in Figure 1. As can be seen, increasing the curing time from 3 to 28 days increased the compressive strength. The results agree with [22,28]. It is believed that the increment in compressive strength alongside curing time is related to the improvement of the geopolymerization process.   The compressive strengths of geopolymer are dependent upon the source of the aluminosilicate and reaction conditions (the type and concentration of the alkali used, curing time, and temperature). Rice husk (RH), Rice husk ash (RHA), and Palm oil fuel ash (POFA) contain a high percentage of silicon dioxide or alumino-silica in amorphous form and can be used as pozzolanic material [29] to prepare geopolymers. The compressive strength could be tuned by controlling the strength of the gel phase, the ratio of the gel phase to undissolved Al-Si particles, the nature of the amorphous phase, the degree of crystallinity, and surface reactions between the gel phase and undissolved particles [30,31]. The OH − in the activators plays a catalytic role for reactivity, while the metal cation helps form a structural element, and balance the negative framework of tetrahedral aluminum [32] to accelerate the precipitation and crystallization of the siliceous and aluminous species. In order to develop a more enhanced version of geopolymer, we took into account the influence of the aforementioned parameters on geopolymer's compressive strength, while the organic materials utilized as precursors should be studied in greater detail.
The sample contains RHA 20%, POFA 10%, MK 35%, while slag 35% without raw rice husk was investigated for comparison purpose, which reported a compressive strength of 5.48 MPa (28 days curing) and water absorption of 30.8% (24 h soaking). It was confirmed that although the addition of raw rice husk decreases compressive strength, it increases the percentage of water absorption percentage.

Water Absorption
The effect of soaking time on water absorption of the prepared samples has been investigated. The results showed that water absorption of the specimens increases from 1 to 24 h of being soaked ( Figure 2). The reported results are the average of three samples. aforementioned parameters on geopolymer's compressive strength, while the organic materials utilized as precursors should be studied in greater detail. The sample contains RHA 20%, POFA 10%, MK 35%, while slag 35% without raw rice husk was investigated for comparison purpose, which reported a compressive strength of 5.48 MPa (28 days curing) and water absorption of 30.8% (24 h soaking). It was confirmed that although the addition of raw rice husk decreases compressive strength, it increases the percentage of water absorption percentage.

Water Absorption
The effect of soaking time on water absorption of the prepared samples has been investigated. The results showed that water absorption of the specimens increases from 1 to 24 h of being soaked ( Figure 2). The reported results are the average of three samples. Meanwhile, the relation between porosity, water absorption, and compressive strength showed that increment in porosity increases water absorption, while the residual compressive strength of the specimens is inversely proportional to both water absorption and porosity (Table 4).
Taking into account the necessity of having high water absorption and compressive strength in the prepared geopolymers for evaporative cooling application, the DOE software selected sample 16, which reported a ~38% of water absorption and ~5 MPa compressive strength.
Despite the fact that porous material such as porous ceramic and geopolymer was previously Meanwhile, the relation between porosity, water absorption, and compressive strength showed that increment in porosity increases water absorption, while the residual compressive strength of the specimens is inversely proportional to both water absorption and porosity (Table 4).
Taking into account the necessity of having high water absorption and compressive strength in the prepared geopolymers for evaporative cooling application, the DOE software selected sample 16, which reported a~38% of water absorption and~5 MPa compressive strength.
Despite the fact that porous material such as porous ceramic and geopolymer was previously fabricated as evaporative cooling materials by [33][34][35], this research developed a greener geopolymer by utilizing a significant amount of industrial and agricultural waste as secondary raw materials, resulting in excellent waste management and sustainable production ( Table 5). The development of geopolymer materials is in accordance with the requirements of sustainable development; it not only require lower processing temperatures, but can also utilize significant amounts of industrial and agricultural waste materials as a secondary raw material, and convert them to a new product. The creation of a new material from different industrial and agricultural waste and by-products might result in reduced energy consumption, reduction of waste production, reduction of global CO 2 emissions, as well as the reduction of exploitation of natural resources. Geopolymers can be customized to a wide variety of properties and characteristics, depending on its constituents or processing methods. Some of these properties include compressive strength, shrinkage, setting, acid resistance, fire resistance, and thermal conductivity. However, these properties are not inherent in all of the geopolymetric formulations. In this case, inorganic polymers should be regarded as the solution to problems associated with the material selection, but as a solution that can be tailored by a correct mix and processing design that will optimize properties and/or deduct the costs for a given application.

Capillary Rise Properties
The capillary rise results of the geopolymer paste samples are shown in Figure 3. Equation (1) has been employed.
where I = the absorption, m t = the change in specimen mass in grams at the time t, a = the exposed area of the specimen, in mm 2 , and d = the density of the water in g/mm 3 .
where I = the absorption, mt = the change in specimen mass in grams at the time t, a = the exposed area of the specimen, in mm 2 , and d = the density of the water in g/mm 3 . As seen in Figure 3, the capillary rise increases in tandem with time. Despite the tremendous increase taking place from the initial immersion time up till 300 s, there is only a slight increase recorded in that span of time. However, our results are not the highest [36,42] but are promising, due to its high water absorption and reasonable mechanical properties.

Microstructure and Morphology of Materials
The morphological features of the particles of the materials were examined using the SEM Carl Zeiss-Merlin compact (A Carl Zeiss SMT AG, Oberkochen, Germany) ( Figure 4). Differences in microstructures could be seen in the raw and product material. As seen in the images, the outer epidermis of rice husk is organized in structures that resemble rolling hills (Figure 4a), while RHA was in the form of crushed shaped particles and cellular grains (Figure 4d). However, metakaolin exhibited layered structure, while POFA exhibited irregular-shaped particles that are porous. The slag particles are squared diamonds. The SEM images of sample 16 before and after casting (dry mix before and after activation) shows that the geopolymer became more uniform after casting. As seen in Figure 3, the capillary rise increases in tandem with time. Despite the tremendous increase taking place from the initial immersion time up till 300 s, there is only a slight increase recorded in that span of time. However, our results are not the highest [36,42] but are promising, due to its high water absorption and reasonable mechanical properties.

Microstructure and Morphology of Materials
The morphological features of the particles of the materials were examined using the SEM Carl Zeiss-Merlin compact (A Carl Zeiss SMT AG, Oberkochen, Germany) ( Figure 4). Differences in microstructures could be seen in the raw and product material. As seen in the images, the outer epidermis of rice husk is organized in structures that resemble rolling hills (Figure 4a), while RHA was in the form of crushed shaped particles and cellular grains (Figure 4d). However, metakaolin exhibited layered structure, while POFA exhibited irregular-shaped particles that are porous. The slag particles are squared diamonds. The SEM images of sample 16 before and after casting (dry mix before and after activation) shows that the geopolymer became more uniform after casting.

X-ray Diffraction (XRD)
Inorganic polymers, including geopolymers, are often described as X-ray amorphous [2], since the main characteristic of their XRD spectra is a featureless bump centered at~2θ of 20-30 • (Figure 5).
metakaolin used in this study shows a product of low crystallinity due to the presence of an amount of quartz (SiO2), which is identified by a peak at ~25° (2θ) [51,52]. The XRD pattern for sample 16 before and after casting belongs to the crystalline SiO2 phase, which proves that traces of quartz are already present in MK precursor in MK-based geopolymers [53,54]. The reason for the small shift from 26.84° (2θ) to 26.62° (2θ) could be due to the reaction of SiO2 during the casting and forming of another silica phase (Folder S1).  Figure 6 shows the IR spectra of the raw materials and geopolymer sample 16 before and after casting. According to Davidovits [55], networks of silico-aluminate-based geopolymers consist of SiO4 and AlO4 tetrahedral linked alternately via oxygen sharing. The main band of geopolymer falls within the region of 900-1200 cm −1 , corresponding to the Si-O-T (where T is Al or Si) linkages. The absorptions at 996 and 797 cm −1 are assigned to Al (IV)-OH (6 fold coordinated) and Al (IV)-O (6-fold coordinated), respectively. As seen in Figure 6, the IR peaks of geopolymer before casting appeared in 1052 cm −1 instead of 996 cm −1 . This means that the asymmetric bending of the bonds O-Si-O and O-Al-O are within lower frequencies, which is in agreement with Alonso et al. [56]. The broad bands at 3000-3500 cm −1 and 1650-1655 cm −1 , which can be clearly seen in the geopolymer sample after casting, are attributed to the stretching and deforming vibrations of OH and H-O-H groups of water Rice husk, which contains cellulose, hemicellulose, and lignin, only exhibits the crystalline structure of cellulose. The peaks~16 • , 22 • , and 35 • are attributed to cellulose [43,44], which exhibits the crystalline structure due to hydrogen bonding interactions and Van der Waals forces between adjacent molecules [45,46]. The RHA shows a broad peak~22 • (2θ) for amorphous silica and cellulose [44,47,48]. The major crystalline phase in slag corresponds to CaCO 3 [49].

Fourier Transform Infrared Spectroscopy (FTIR)
The POFA peaks are related to SiO 2 (Q-Quartz; C-Cristobalite) [50]. The XRD pattern of the metakaolin used in this study shows a product of low crystallinity due to the presence of an amount of quartz (SiO 2 ), which is identified by a peak at~25 • (2θ) [51,52]. The XRD pattern for sample 16 before and after casting belongs to the crystalline SiO 2 phase, which proves that traces of quartz are already present in MK precursor in MK-based geopolymers [53,54]. The reason for the small shift from 26.84 • (2θ) to 26.62 • (2θ) could be due to the reaction of SiO 2 during the casting and forming of another silica phase (Folder S1). Figure 6 shows the IR spectra of the raw materials and geopolymer sample 16 before and after casting. According to Davidovits [55], networks of silico-aluminate-based geopolymers consist of SiO 4 and AlO 4 tetrahedral linked alternately via oxygen sharing. The main band of geopolymer falls within the region of 900-1200 cm −1 , corresponding to the Si-O-T (where T is Al or Si) linkages. The absorptions at 996 and 797 cm −1 are assigned to Al (IV)-OH (6 fold coordinated) and Al (IV)-O (6-fold coordinated), respectively. As seen in Figure 6, the IR peaks of geopolymer before casting appeared in 1052 cm   Table 6 shows the textural properties of raw materials and geopolymer sample 16 before and after casting (the dry mix before and after activation). The pore size distribution was calculated from the adsorption-desorption of the isotherms using the Barret-Joyner-Halenda (BJH) model [57]. The decrease in the surface area of the geopolymer after casting could be due to the completion of geopolymerization reaction and particle growth after casting.

Conclusions
This study investigated the use of certain agricultural, and industrial wastes byproduct to prepare environmentally friendly geopolymers for evaporative cooling system applications. The results showed that the concentration of alkali activator influenced the compressive strength of geopolymeric specimens. The ratio of sodium silicate solution-to-sodium hydroxide solution ratio in terms of mass was directly proportional to the compressive strength of the geopolymer paste. This correlation was also evident for the compressive strength, water absorption, and apparent porosity of the geopolymer paste specimens; with specimens exhibiting lower water absorption and porosity resulting in higher residual compressive strength. The percentage of water absorption and pore volume of the specimens increased via the usage of RH, MK, and RHA. In this study, a green geopolymer employing a sustainable approach was prepared with excellent comprehensive strength, water absorption, and capillary rise using different industrial and agriculture waste and byproducts. To the best of our knowledge, there is no work that utilizes Rice husk (RH), rice husk ash (RHA), metakaolin (MK), ground granulated blast furnace slag (GGBS) and palm oil fuel ash (POFA) in geopolymer binder for evaporative cooling applications. However, the compressive strength of prepared geopolymers is not as high when compared to the other geopolymers; it is within the accepted range for use as evaporative cooling materials.  Table 6 shows the textural properties of raw materials and geopolymer sample 16 before and after casting (the dry mix before and after activation). The pore size distribution was calculated from the adsorption-desorption of the isotherms using the Barret-Joyner-Halenda (BJH) model [57]. The decrease in the surface area of the geopolymer after casting could be due to the completion of geopolymerization reaction and particle growth after casting.

Conclusions
This study investigated the use of certain agricultural, and industrial wastes byproduct to prepare environmentally friendly geopolymers for evaporative cooling system applications. The results showed that the concentration of alkali activator influenced the compressive strength of geopolymeric specimens. The ratio of sodium silicate solution-to-sodium hydroxide solution ratio in terms of mass was directly proportional to the compressive strength of the geopolymer paste. This correlation was also evident for the compressive strength, water absorption, and apparent porosity of the geopolymer paste specimens; with specimens exhibiting lower water absorption and porosity resulting in higher residual compressive strength. The percentage of water absorption and pore volume of the specimens increased via the usage of RH, MK, and RHA. In this study, a green geopolymer employing a sustainable approach was prepared with excellent comprehensive strength, water absorption, and capillary rise using different industrial and agriculture waste and byproducts. To the best of our knowledge, there is no work that utilizes Rice husk (RH), rice husk ash (RHA), metakaolin (MK), ground granulated blast furnace slag (GGBS) and palm oil fuel ash (POFA) in geopolymer binder for evaporative cooling applications. However, the compressive strength of prepared geopolymers is not as high when compared to the other geopolymers; it is within the accepted range for use as evaporative cooling materials.