Geopolymeric Composite Materials Made of Sol-Gel Silica and Agroindustrial Wastes of Rice, Barley, and Coffee Husks with Wood-Like Finishing

: Geopolymers have been mainly utilized as structural materials; their chemical structure and morphologies have been explored for their potential as a high-performance material in emerging applications. Geopolymer composites reinforced with materials based on agro-waste are attracting interest in engineering applications due to their easy processing, low cost, low density, and high strength-to-weight ratio. This investigation pursues an experimental methodology that consists of a scheme to make composites with a geopolymer matrix and agro-waste (rice husk, barley, and coffee) as aggregate material, that can be applied in various ﬁelds of construction. The study was intended to determine the inﬂuence of adding various agro-wastes on the mechanical properties of the geopolymer. According to the respective ASTM standards, the materials obtained were prepared and analyzed to determine their compressive strength, ﬂexural strength, hardness and scanning electron microscopy (SEM)-determined characteristics. The results revealed that, for the compression tests, the composites formed by a sol-gel matrix and barley husk showed a better yield, obtaining the highest value of 3.5 N/mm 2 . Concerning hardness testing, the composites with a geopolymer matrix and coffee husks obtained higher values compared to the other composites. For the ﬂexural tests, the compounds with the sol-gel/ﬂy ash matrix obtained the highest yield stress value, which was 5.25 MPa with an elastic modulus of 7.59 GPa. The results of the microstructural analysis showed good husk-matrix interaction, together with failure mechanisms. The conformation of such waste-based compounds may enable them to replace natural wood in some applications, such as in the ﬁnishing of interiors of homes, during the ﬁnal stages in the construction of buildings, or in the decoration of inhabited houses, as well as in ﬁnishing in the manufacture of furniture.


Introduction
The need for new, more efficient, and highly sophisticated materials has contributed to the rapid development of various materials for the construction and building sectors [1]. These materials can significantly influence the environment throughout the entire process, ing artificial fibers is an energy-consuming process compared to growing and harvesting natural fibers [37].
In the literature, several studies have investigated improvements in the physical and mechanical properties of polymeric composites reinforced with natural fibers. Zhou et al. [38] evaluated the influence of cotton stem fibers. These authors applied a pretreatment to the cotton stem to improve the fiber-matrix bond performance and the geopolymer mechanical properties. Bending and compression tests were carried out. The results showed that the alkaline treatment was effective in creating a fiber-matrix union and the compressive and flexural strength values were improved by 4.8% and 11.5%, respectively, compared to untreated composites.
Wongsa et al. [39] evaluated two types of natural fibers (sisal fiber and coconut fiber) in geopolymer mortars with variable proportions of 0%, 0.5%, 0.75%, and 1.0% of total volume. They evaluated the mechanical, thermal, and physical properties. The results indicated that the addition of natural fiber as a reinforcing material resulted in a significant improvement in tensile strength and flexural performance similar to that obtained by the use of fiberglass.
Silva et al. [40] evaluated the effect of the fiber content of jute (between 0.5 and 2.0% by weight) and sisal (between 0.5 and 3.0% by weight) on the mechanical properties of geopolymers. Performance in terms of compression, traction by division, and bending at three points was assessed. The addition of 2.5% sisal fibers increased the compressive, split tensile, and flexural strengths by up to 76%, 112%, and 270%, respectively, compared to control samples. Geopolymers with 1.5% (wt%) jute fiber reinforcement showed increases of up to 64%, 45%, and 222% in compressive, split tensile, and flexural strength, respectively.
However, the search continues for new construction materials with geopolymeric matrices based on natural fibers that can show high resistance to compression, flexion, and traction to replace traditional materials based on Portland cement.
The objective of the present study was to evaluate the incorporation of agro-industrial residues (rice, barley, and coffee husks) as a reinforcing material in geopolymeric matrices, to support the use of agro-industrial wastes in the development of an alternative material to wood, with potential primary applications in the manufacture of furniture and in the interior finishing of homes.

Preparation of Composite Materials with Geopolymeric Matrix-Husks (Agro-Waste)
In this investigation, four different matrices were used for the conformation of matrix/agrowaste composites. Following the sol-gel process methodology described above, a matrix composed mainly of SiO 2 was formed and labeled SG. Another matrix was the geopolymer, consisting mainly of metakaolin (GP), produced according to the established methodology for manufacturing the geopolymer [30]. For the sol-gel and geopolymer matrix, denoted SG/GP, the geopolymer was first mixed according to the methodology for the GP matrix, except for the incorporation of metasilicate, which was replaced by a sol-gel solution. For the sol-gel and fly ash matrix (SG/FA), sol-gel was used and fly ash added. Table 1 shows the components and amounts used to prepare the four matrixes. The mixture was made with the husks once the matrices were elaborated. This was performed manually at room temperature ( Figure 1). According to previous studies [21,32], it was determined that a 4:1 (80/20) matrix-too-husks weight ratio allows the husks to be fully integrated into the matrix. The size and shape of the husk particles significantly affect the mixture and characteristics of the composite materials.
The samples derived from the combination of the different matrices and the husks were placed in cylindrical molds with the following dimensions: 5 cm in diameter by 10 cm high ( Figure 2). They were allowed to dry at room temperature on a dry surface and covered on top. After a pre-drying stage, the samples were heat-treated at 60 • C for 3 h. This was performed to complete the drying process and reaction of the matrix. Figure 2 shows the sol-gel silica solution, the mixtures with husks, and the cylindrical probes. Figure 3a shows the metakaolin powder and Figure 3b the silica sol-gel solution. In the case of the geopolymeric mixtures, prior to the molding stage, the dough was allowed to set for about 1-3 h (Figure 3c,d), covered by a polypropylene bag to avoid humidity loss. This setting time ensured complete hydrolyzation and polymerization of the admixture so that it was uniform without lumps.
Under the same procedures as used for the samples for the compression tests (Figure 3e-g), other samples with a rectangular shape of about 5 × 24 × 1.5 cm, following the ASTM C947-99 standard [46], were prepared to perform the flexural tests (Figure 3h,i). The samples were allowed to dry. After a pre-drying stage on a flat surface at room temperature, the clapboards were heat-treated in a furnace at 60 • C for 3 h (Figure 3j).

Mechanical Tests
The cylindrical samples were subjected to a force and speed of 2.5 mm/min in a Galdabini 2000 Universal Machine of 50 ton. Calculations were performed using the formulas in Table 2 for compressive strength, deformation, elastic modulus, and strain.
The hardness test measures the surface resistance of a material to penetration by a hard object. The cylindrical samples used for the compression test were evaluated with a Shore D Durometer, Traverstool. A minimum of six measurements were performed for each side of the composites; then, the average value of these measurements was obtained. Rectangular samples were prepared to be subjected to the bending test in a Tinius Olsen Universal Machine, in accordance with ASTM C947-99 [46].
Microstructural images of the geopolymers were obtained using a scanning electron microscope (SEM, JEOL JSM-6510 LV) with an energy-dispersive X-ray spectrometer (EDS, Bruker; Fahrenheitstraße 4, Bremen, Germany). The images were obtained from thick wafers cut from the cylindrical samples. Their surfaces were polished to obtain a smooth finish. The samples derived from the combination of the different matrices and the hus were placed in cylindrical molds with the following dimensions: 5 cm in diameter by cm high (Figure 2). They were allowed to dry at room temperature on a dry surface an   In the case of the geopolymeric mixtures, prior to the molding stage, the dough was allowed to set for about 1-3 h (Figure 3c,d), covered by a polypropylene bag to avoid humidity loss. This setting time ensured complete hydrolyzation and polymerization of the admixture so that it was uniform without lumps.   Under the same procedures as used for the samples for the compression tests ( Figure  3e-g), other samples with a rectangular shape of about 5 × 24 × 1.5 cm, following the ASTM C947-99 standard [46], were prepared to perform the flexural tests (Figure 3h,i). The samples were allowed to dry. After a pre-drying stage on a flat surface at room temperature, the clapboards were heat-treated in a furnace at 60 °C for 3 h (Figure 3j).

Structure of Composite Materials with Husks
Preliminary drying is an important factor in the preparation of composite materials. The shape the material acquires during its drying is directly related to the mold into which the mixture has been poured. When the geopolymer was dried at room temperature, the mold was provided with as much ventilation as possible and the geopolymer position was alternated. Uniform ventilation was achieved using a mold by shaping a metallic net. This was to achieve a product with a symmetric shape as uneven drying deforms the samples. As a final step in drying, the pieces were heat-treated at 60 • C for 3 h.
The composite materials were studied using SEM micrographs to observe and analyze the microstructure and morphology of the geopolymers. Figure 4 shows the images obtained by scanning electron microscopy of the rice husk composites for the four types of matrices. Figure 4a shows an image of a sol-gel silica matrix with fractures and some needle-shaped effloresce with no observation of a husk on the surface. White needle-like formations are associated with crystals of sodium silicate, which arise as efflorescence. Figure 4b shows a mostly agglomerated and low-porous surface with an uncovered fragment of rice husk embedded in the matrix on its right side. In Figure 4c, rice husks are shown to be entirely covered by the matrix, and a rougher, homogeneous, and low-porous surface is observed. These images indicate that the rough surface of the husk improved the bond with the geopolymeric matrix. Figure 4d shows the cenospheres of the fly ash, which indicates that, in the sample preparation process, more sodium hydroxide and time were required for the dissolution to be completed and more raw material was required to be provided to conform the matrix of the composite material. Figure 5 shows micrographs of composites using four different matrixes and barley husks. Figure 5a shows the surface of a sample of composite material with the SG matrix. Since the matrix completely covers the husks, no fragments of husks are observed without coverage of the matrix; however, in this case, the matrix is composed of particles of about 3 µm, and, in turn, does not show fractures. In Figure 5b, the GP matrix is observed to have an agglomerated structure. In the case of the matrix composed of SG/GP (Figure 5c), a column of piled fragments was observed where the husks were left uncoated. In the same way, efflorescences corresponding to sodium silicate were present. Figure 5d shows no cenospheres and abundant fractures of the matrix material. Amplification in the image allows dendritic crystalline formations of sodium silicate under the surface level to be seen.    Figure 5a shows the surface of a sample of composite material with the SG matrix. Since the matrix completely covers the husks, no fragments of husks are observed without coverage of the matrix; however, in this case, the matrix is composed of particles of about 3 µm, and, in turn, does not show fractures. In Figure 5b, the GP matrix is observed to have an agglomerated structure. In the case of the matrix composed of SG/GP (Figure 5c), a column of piled fragments was observed where the husks were left uncoated. In the same way, efflorescences corresponding to sodium silicate were present. Figure 5d shows no cenospheres and abundant fractures of the matrix material. Amplification in the image allows dendritic crystalline formations of sodium silicate under the surface level to be seen. In Figure 6, the micrographs show the different matrices covering the coffee husks. In Figure 6a, the fractures and efflorescence of the sodium silicate crystals are presented. In Figure 6b, needle-shaped crystals can be observed (about 5 µm), associated with the sodium silicate being in excess at the composite and its solubility in water. In Figure 6c, the SG/GP matrix shows an agglomerated structure, similar to the previous image, where the material reacted completely, resulting in a more homogeneous surface. Figure 6d no longer shows cenospheres, but has fractures associated with the loss of moisture and a rigid structure; aluminum bonds cause the breakage without allowing the inorganic polymeric chains of the gel to adjust to volume changes.

Hardness and Compression Strength Tests
The mechanical behavior of composite materials is one of their important properties since some organic-based materials have very weak mechanical properties and require particular processing for use as a construction material [26,47].
The twelve combinations of the materials were processed in triplicate. The composite matrices were SG, GP, SG/GP, and SG/FA; the aggregates were: rice, barley, and coffee husks. The drying process significantly affected the final shapes of the samples. When the drying was not uniform, these samples were deformed, making it difficult to use some of the samples for the different tests. In Figure 6, the micrographs show the different matrices covering the coffee husks. In Figure 6a, the fractures and efflorescence of the sodium silicate crystals are presented. In Figure 6b, needle-shaped crystals can be observed (about 5 µm), associated with the sodium silicate being in excess at the composite and its solubility in water. In Figure 6c, the SG/GP matrix shows an agglomerated structure, similar to the previous image, where the material reacted completely, resulting in a more homogeneous surface. Figure 6d no longer shows cenospheres, but has fractures associated with the loss of moisture and a rigid structure; aluminum bonds cause the breakage without allowing the inorganic polymeric chains of the gel to adjust to volume changes. Hardness and compression strength tests were performed on the samples. In accordance with the results shown in Figure 7a, it was observed that the GP-based composite had higher hardness compared to the other three matrices and for the three different types of husks. The highest hardness value was obtained by combining the GP matrix with the coffee husk reinforcement.
The hardness values obtained in the SG-based composites were similar for the three types of husks. The indentation penetration for the hardness tests was relatively small. When there is a low reinforcement/matrix ratio, it is to be expected that the matrix is mainly being evaluated based on a low contribution by the reinforcement. In this study, the matrix is a binding material for the husks, which were in a high proportion and acted more like an adhesive. This indicates that the matrix covered the husk particles and its characteristics were evident in the hardness measurements. In these measurements, the shape of the husks did not have a significant effect.
For the composites made with the SG/GP matrix, a notable difference was observed between the hardness values obtained with the rice and barley husks, with the rice husks exhibiting higher values. In the case of the SG/FA matrix, the rice and barley husks were minimally different compared to the coffee husks. For this matrix, the coffee husks showed a similar value to that obtained with the SG matrix, with the lowest values obtained among the different matrices. Consistent with Figure 7a, it was observed that, for hardness strength, the GP matrix presented higher values. This was attributed to the presence of aluminosilicate compounds containing Si(IV) and Al(III), which are associated with hard materials.
However, such materials are fragile when subject to load, as evidenced by the compressive strength tests. The same applied to the matrix formed with SG/FA. Aluminosilicates were present in this matrix, providing hardness to the composite. However, the material had lower hardness than SG, which did not include aluminum as a constituent.

Hardness and Compression Strength Tests
The mechanical behavior of composite materials is one of their important properties since some organic-based materials have very weak mechanical properties and require particular processing for use as a construction material [26,47].
The twelve combinations of the materials were processed in triplicate. The composite matrices were SG, GP, SG/GP, and SG/FA; the aggregates were: rice, barley, and coffee husks. The drying process significantly affected the final shapes of the samples. When the drying was not uniform, these samples were deformed, making it difficult to use some of the samples for the different tests.
Hardness and compression strength tests were performed on the samples. In accordance with the results shown in Figure 7a, it was observed that the GP-based composite had higher hardness compared to the other three matrices and for the three different types of husks. The highest hardness value was obtained by combining the GP matrix with the coffee husk reinforcement. The compressive strength test was performed on a series of twelve samples. The best compressive strength value was obtained for the SiO 2 sol-gel matrix, with an average resistance of about 1.92 N/mm 2 . The results showed that the composite having the best compressive strength was the SiO 2 sol-gel matrix with barley husks. In this composition, a resistance of about 3.49 N/mm 2 was reached, a value more than double and triple that obtained with respect to rice and coffee husks, respectively.
The average obtained in the samples elaborated with the GP matrix was about 1.27 N/mm 2 . The rice husk composites achieved the highest compressive strength of about 2.51 N/mm 2 . The sample with barley husk had a compressive strength value of 1.10 N/mm 2 and that prepared with coffee husk a value of about 0.20 N/mm 2 . The enhanced performance with the rice husk may be associated with an increased bond between it and the geopolymeric matrix.
The test results for the samples formed by GP/SG/FA showed an average compressive strength of about 0.68 N/mm 2 . It should be noted that, in these composites, the samples produced with coffee husks did achieve a suitable consistency to be tested on the different occasions they were manufactured; therefore, they are not shown in Figure 7b The hardness values obtained in the SG-based composites were similar for the three types of husks. The indentation penetration for the hardness tests was relatively small. When there is a low reinforcement/matrix ratio, it is to be expected that the matrix is mainly being evaluated based on a low contribution by the reinforcement. In this study, the matrix is a binding material for the husks, which were in a high proportion and acted more like an adhesive. This indicates that the matrix covered the husk particles and its characteristics were evident in the hardness measurements. In these measurements, the shape of the husks did not have a significant effect.
For the composites made with the SG/GP matrix, a notable difference was observed between the hardness values obtained with the rice and barley husks, with the rice husks exhibiting higher values. In the case of the SG/FA matrix, the rice and barley husks were minimally different compared to the coffee husks. For this matrix, the coffee husks showed a similar value to that obtained with the SG matrix, with the lowest values obtained among the different matrices. Consistent with Figure 7a, it was observed that, for hardness strength, the GP matrix presented higher values. This was attributed to the presence of aluminosilicate compounds containing Si(IV) and Al(III), which are associated with hard materials. However, such materials are fragile when subject to load, as evidenced by the compressive strength tests. The same applied to the matrix formed with SG/FA. Aluminosilicates were present in this matrix, providing hardness to the composite. However, the material had lower hardness than SG, which did not include aluminum as a constituent. The combination with SG/FA produced an average compressive strength result of about 1.75 N/mm 2 . The specimen that presented the highest resistance value was that made with barley husk with a value of about 2.43 N/mm 2 , with rice husk next at about 2.11 N/mm 2 , and then coffee husk at about 0.72 N/mm 2 . Etuk et al. [48] obtained values of 0.152 ± 0.006 N/mm 2 for the compressive strength of an insulating panel developed from sugarcane leaf. Therefore, the compounds obtained in this investigation showed higher resistance before fracturing or breaking than the sugarcane leaf panel.
Finally, Figure 7b shows that the barley husk samples provided satisfactory results for compression resistance in three of the four types of matrices (SG, SG/GP, and SG/FA). These samples had higher values compared to those corresponding to the other two husks. This was attributed to the shape of the barley husks which created a higher degree of contact with the matrix on all its faces, reinforcing it. Therefore, the values obtained for the different compounds differed from each other, mainly due to the fiber-binder bondage, the geometrical configuration, the density, the micro-fibril angle and the moisture absorbing capacity [49].
The rice husk composites also exhibited favorable properties in terms of the compressive strength of the material. Again, this was attributed to the shape of the husks, which, unlike the entire barley husks, comprised small fine particles. The increased surface area enabled stronger adherence with the matrix and formed a more resistant product. The coffee husk composites obtained the lowest results in the different matrices, with poor performance when mixed with the SG/GP matrix. This means that the samples could not be measured because they crumbled in the form obtained under the same conditions and material concentrations for the other husk samples. This indicates certain limitations in the use of this kind of husk to form compounds. The husk geometric structure is an important factor that improves the properties of the developed compound, depending on the desired characteristics of the composite material.
The capacity of the barley husk composites to withstand higher loads was attributed to the fact that they covered a larger volume in the three dimensions, which facilitated the distribution of the load. In the case of the rice husks, having the smallest particles, they facilitated the compaction of the material, resulting in higher compressive strength than for the coffee husk composites. The coffee husk composites showed lower resistance than the other two composites, except for the SG/FA matrix. Its geometric structure favored the resistance of the compound when evaluated in a direction perpendicular to the plane of the flakes. The compounds were weaker when the loads were applied parallel to the plane of the flakes. Similarly, it was found that, in the composite material of the SG matrix and coffee husks, the husks were aligned, forming precisely stacked layers of husks. Figure 8 shows the fractures in the composite test cylinders made with rice (a-d), barley (e-h), and coffee (i-k) husks. The coffee husk sample with a matrix of geopolymer/solgel/fly ash was adequately molded. This combination required a higher amount of sodium hydroxide in the preparation process. Figure 8i shows the stacked husks distributed in the composite, forming aligned layers of the husks, which caused anisotropic behavior in the horizontal and vertical axes.  Figure 8 shows the fractures in the composite test cylinders made with rice (a-d), barley (e-h), and coffee (i-k) husks. The coffee husk sample with a matrix of geopolymer/sol-gel/fly ash was adequately molded. This combination required a higher amount of sodium hydroxide in the preparation process. Figure 8i shows the stacked husks distributed in the composite, forming aligned layers of the husks, which caused anisotropic behavior in the horizontal and vertical axes.

Flexural Properties
The appearance most like natural wood was achieved with the compositions formed with the sol-gel matrix, especially for the compositions formed with rice husks. Figure 9 shows clapboards prepared for the flexural tests. They show the appearance of these composites with Serie I (sol-gel) and II (sol-gel/fly ash) resembling natural wood.

Flexural Properties
The appearance most like natural wood was achieved with the compositions formed with the sol-gel matrix, especially for the compositions formed with rice husks. Figure 9 shows clapboards prepared for the flexural tests. They show the appearance of these composites with Serie I (sol-gel) and II (sol-gel/fly ash) resembling natural wood. Serie III samples with the composite made with a silica sol-gel/geopolymer (SG/GP) matrix were similar to clay or mortar. Figure 10a,b present a close-up of the Serie III sample. This kind of composite is of interest for some finishings. The soft texture at the touch, the appearance after polishing, and the consistency were evident when implemented on a building (Figure 10b). Figure 10c-e show some application stages on some surfaces of a building. This material was preferred as a polished surface for appearance and better display of the yellow rice husk components.   Serie III samples with the composite made with a silica sol-gel/geopolymer (SG/GP) matrix were similar to clay or mortar. Figure 10a,b present a close-up of the Serie III sample. This kind of composite is of interest for some finishings. The soft texture at the touch, the appearance after polishing, and the consistency were evident when implemented on a building (Figure 10b). Figure 10c-e show some application stages on some surfaces of a building. This material was preferred as a polished surface for appearance and better display of the yellow rice husk components. Figure 10f,g show a close-up view of the silica sol-gel/coffee husks with a porous structure. The flat flakes of coffee husks made the structure pilled laminas with many void spaces.
The component ratios for the SG/GP require adjustment for the intended use, such as in Figure 10a-e. Thus, only the samples made with the SG, GP, and SG/FA matrixes were tested with the mixtures of rice, barley, and coffee husks. Figure 11 shows the flexural tests performed on the clapboards of rice husks (Figure 9a), barley husks (Figure 9b), and coffee husks (Figure 9c). In this figure, three double columns show the clapboards before and after conducting the tests. The higher values were observed in the sample with rice husks, which was above double the value obtained for the other two husk samples. Depending on the use intended for the material, these values may be sufficient. This is why defining the application first and modifying the mixtures is recommended to obtain products with adjusted characteristics. Table 3 summarizes the values obtained in the flexural tests for each sample and Figure 12 show the elastic and rupture modules. Ceramics are materials with limited mechanical toughness; because of this, the flexural strength values observed in the tests were, relatively, not high. In the case of the samples made with the sol-gel matrix, the composite material with the best yield strength and elastic modulus was that made using barley husks. This exceeded the values obtained for the other two combinations.
Serie III samples with the composite made with a silica sol-gel/geopolymer (SG/GP) matrix were similar to clay or mortar. Figure 10a,b present a close-up of the Serie III sample. This kind of composite is of interest for some finishings. The soft texture at the touch, the appearance after polishing, and the consistency were evident when implemented on a building (Figure 10b). Figure 10c-e show some application stages on some surfaces of a building. This material was preferred as a polished surface for appearance and better display of the yellow rice husk components.   The component ratios for the SG/GP require adjustment for the intended use, such as in Figure 10a-e. Thus, only the samples made with the SG, GP, and SG/FA matrixes were tested with the mixtures of rice, barley, and coffee husks. Figure 11 shows the flexural tests performed on the clapboards of rice husks ( Figure  9a), barley husks (Figure 9b), and coffee husks (Figure 9c). In this figure, three double columns show the clapboards before and after conducting the tests. The higher values were observed in the sample with rice husks, which was above double the value obtained for the other two husk samples. Depending on the use intended for the material, these values may be sufficient. This is why defining the application first and modifying the mixtures is recommended to obtain products with adjusted characteristics.  Table 3 summarizes the values obtained in the flexural tests for each sample and Figure 12 show the elastic and rupture modules. Ceramics are materials with limited mechanical toughness; because of this, the flexural strength values observed in the tests were, relatively, not high. In the case of the samples made with the sol-gel matrix, the composite material with the best yield strength and elastic modulus was that made using barley husks. This exceeded the values obtained for the other two combinations.

Yield
Yield De-Ultimate Ultimate Elastic   In the case of the geopolymer matrix, the combination with the highest yield stress and modulus of elasticity was that of the mixture with rice husks, though inferior to the values obtained using the SG/FA matrix. It was not possible to calculate the stresses for this matrix in combination with coffee husks since the material immediately showed fractures.
The SG/GP matrix samples could not be measured because they crumbled in the obtained form as the parameters were similar to those of the other composites.
For the matrix integrated using sol-gel/fly ash, the combination with rice provided better results for the composite. The last registered value with this combination was higher than 5 MPa and the modulus of elasticity was 7595 GPa. These were the highest values for all the samples tested. Malaszkiewicz et al. [50] observed flexural strength values between 0.91 and 2.46 MPa for fiberboard production. Figure 13 shows the behavior of the different compounds tested corresponding to the SG, GP, and SG/FA matrices. Figure 13a shows similar graphs with respect to the strain stress behavior for the three types of husks. However, the yield points had low values for these matrices, confirming their fragile behavior. In the case of the geopolymer matrix, the combination with the highest yield stress and modulus of elasticity was that of the mixture with rice husks, though inferior to the values obtained using the SG/FA matrix. It was not possible to calculate the stresses for this matrix in combination with coffee husks since the material immediately showed fractures.
The SG/GP matrix samples could not be measured because they crumbled in the obtained form as the parameters were similar to those of the other composites.
For the matrix integrated using sol-gel/fly ash, the combination with rice provided better results for the composite. The last registered value with this combination was higher than 5 MPa and the modulus of elasticity was 7595 GPa. These were the highest values for all the samples tested. Malaszkiewicz et al. [50] observed flexural strength values between 0.91 and 2.46 MPa for fiberboard production. Figure 13 shows the behavior of the different compounds tested corresponding to the SG, GP, and SG/FA matrices. Figure 13a shows similar graphs with respect to the strain stress behavior for the three types of husks. However, the yield points had low values for these matrices, confirming their fragile behavior. With respect to the GP matrix, the behaviors of the three husks were very different The resistance of the rice husks was relatively high compared to all other specimens. This can be attributed to the particle size since this enabled better interaction with the geopolymer matrix. However, from the beginning, the probes showed fractures but the samples took longer to collapse. Figure 13b shows the linear behavior of the probes made of coffee husk composite This occurred because the material was fractured in the second measurement, so the elas- With respect to the GP matrix, the behaviors of the three husks were very different. The resistance of the rice husks was relatively high compared to all other specimens. This can be attributed to the particle size since this enabled better interaction with the geopolymer matrix. However, from the beginning, the probes showed fractures but the samples took longer to collapse. Figure 13b shows the linear behavior of the probes made of coffee husk composite. This occurred because the material was fractured in the second measurement, so the elastic modulus values were not determined. In the barley husk composites, more uniform behavior was obtained without high values being achieved due to the ceramic characteristics of the matrix. In Figure 13c, it can be seen that, for the SG/FA matrix, the value obtained for the samples with rice husks was more than twice that of the other husks.

Adobe-Like Composite Prototypes Made with Husks
Adobe-like samples, 10 × 15 × 30 cm in size, were made based on the composition of the tested pieces. The mixtures were created in a mechanical mixer and the materials were compressed in an Adopress 1000 device. They were heat-treated at 60 • C for 8 h in a QL Model 30 GC oven. Figure 14 shows the different combinations that were made based on the procedures used for of the other samples shown. Coffee husks were used with the sol-gel matrix. Samples were made using barley husks for the GP, FA/SG, and SG/GP/FA matrices.

Adobe-Like Composite Prototypes Made with Husks
Adobe-like samples, 10 × 15 × 30 cm in size, were made based on the composition of the tested pieces. The mixtures were created in a mechanical mixer and the materials were compressed in an Adopress 1000 device. They were heat-treated at 60 °C for 8 h in a QL Model 30 GC oven. Figure 14 shows the different combinations that were made based on the procedures used for of the other samples shown. Coffee husks were used with the solgel matrix. Samples were made using barley husks for the GP, FA/SG, and SG/GP/FA matrices.  . These samples were lighter than a mudbrick or adobe [51][52][53], with a typical value for density of about 1800 kg/m 3 .
Another composite was made with the rice husks and the geopolymer matrix. This material was allowed to dry at room temperature. The samples were more compact, and the appearance was more similar to wood than the others. The pressures were adjusted according to the humidity content of the mixture and the husk amount in the samples because, frequently, the mixture overflowed the mold. In this case, the excess of water affected the mechanical behavior. Figure 15 shows the composite geopolymeric bricks after heat treatment. These pieces, molded under pressure, 10 × 15 × 30 cm in size, took a long time to dry. They were also heavy to carry and set. Moreover, they could break (as shown in Figure 15b). They required further adjustments to the composition ratio, water content, and heating ramp.  . These samples were lighter than a mudbrick or adobe [51][52][53], with a typical value for density of about 1800 kg/m 3 .
Another composite was made with the rice husks and the geopolymer matrix. This material was allowed to dry at room temperature. The samples were more compact, and the appearance was more similar to wood than the others. The pressures were adjusted according to the humidity content of the mixture and the husk amount in the samples because, frequently, the mixture overflowed the mold. In this case, the excess of water affected the mechanical behavior. Figure 15 shows the composite geopolymeric bricks after heat treatment. These pieces, molded under pressure, 10 × 15 × 30 cm in size, took a long time to dry. They were also heavy to carry and set. Moreover, they could break (as shown in Figure 15b). They required further adjustments to the composition ratio, water content, and heating ramp. The water content was substantially lower when using an adobe device that applied pressure, which can be beneficial, as in the cement/water ratio for making concrete bricks.

Composite Characteristics Using Sol-Gel Silica Binder
A difference of the present study from others undertaken is that sol-gel silica was used as the matrix. This solution has a colloidal solution of nanoparticles 30 wt/% in solid and sodium silicate. The pH was >11.
The key factor for the conformation of these composites was the proportion of NaOH. As can be seen in Table 1, there was a unique sol-gel solution with a unique quantity of NaOH. Nonetheless, each type of husk and fly ash (FA) required different quantities for mercerization or FA dissolution.
We did not change the NaOH content to compare the composites directly. So, the geopolymer-rice husk did not produce favorable results. However, this composite was applied in a building because of its aesthetic finish.
This investigation involved a comparison, under similar conditions, of the tested composites. It was not about determining the ideal quantities of components to achieve optimal results in terms of mechanical test performance or desirable external aesthetic finishing.
It is well known that fly ash requires much water for preparing concrete, which adversely affects its mechanical properties. Fly ash requires higher amounts of NaOH to dissolve the cenospheres to enable the material to be used in a new composite structure.

Conclusions
Three cereal husks, available in large quantities as agro-industrial wastes, were used for making composites as alternatives to current wood-made products. The matrix used was a geopolymer using metakaolin and sol-gel silica solutions. The mechanical resistances were optimized by adjusting the husk and sodium hydroxide ratio to prepare the materials. The moisture content and the procedure for preparing the samples strongly determined the mechanical characteristics obtained. Different composites were obtained which varied in terms of aesthetic and mechanical properties. Some were suitable for decorative indoor finishing. The Al/Si ratio changed between the different brands of metakaolin, which was a key factor in the structure of the geopolymer and defined its mechanical properties. The size and shape of the husks also determined the characteristics and performance of each kind of composite.
According to the results obtained, it can be concluded that it is possible to form a geopolymer composition based on metakaolin as a matrix with agribusiness waste aggregates. This composite has aesthetic and mechanical properties suitable for decorative indoor finishing. For mechanical properties, the achieved hardness was between 50 and 60 HSD, comparable to the hardness of the tire of a shopping cart. With these properties, the composite can be considered an alternative for some uses of wood in interior finishes. However, this can be adjusted to achieve enhanced performance and fulfill the requirements for such alternatives and the enhance the scope for its use.
The homogeneous structure in the composite made with geopolymer, rice husks, and

Composite Characteristics Using Sol-Gel Silica Binder
A difference of the present study from others undertaken is that sol-gel silica was used as the matrix. This solution has a colloidal solution of nanoparticles 30 wt/% in solid and sodium silicate. The pH was >11.
The key factor for the conformation of these composites was the proportion of NaOH. As can be seen in Table 1, there was a unique sol-gel solution with a unique quantity of NaOH. Nonetheless, each type of husk and fly ash (FA) required different quantities for mercerization or FA dissolution.
We did not change the NaOH content to compare the composites directly. So, the geopolymer-rice husk did not produce favorable results. However, this composite was applied in a building because of its aesthetic finish.
This investigation involved a comparison, under similar conditions, of the tested composites. It was not about determining the ideal quantities of components to achieve optimal results in terms of mechanical test performance or desirable external aesthetic finishing.
It is well known that fly ash requires much water for preparing concrete, which adversely affects its mechanical properties. Fly ash requires higher amounts of NaOH to dissolve the cenospheres to enable the material to be used in a new composite structure.

Conclusions
Three cereal husks, available in large quantities as agro-industrial wastes, were used for making composites as alternatives to current wood-made products. The matrix used was a geopolymer using metakaolin and sol-gel silica solutions. The mechanical resistances were optimized by adjusting the husk and sodium hydroxide ratio to prepare the materials. The moisture content and the procedure for preparing the samples strongly determined the mechanical characteristics obtained. Different composites were obtained which varied in terms of aesthetic and mechanical properties. Some were suitable for decorative indoor finishing. The Al/Si ratio changed between the different brands of metakaolin, which was a key factor in the structure of the geopolymer and defined its mechanical properties. The size and shape of the husks also determined the characteristics and performance of each kind of composite.
According to the results obtained, it can be concluded that it is possible to form a geopolymer composition based on metakaolin as a matrix with agribusiness waste aggregates. This composite has aesthetic and mechanical properties suitable for decorative indoor finishing. For mechanical properties, the achieved hardness was between 50 and 60 HSD, comparable to the hardness of the tire of a shopping cart. With these properties, the composite can be considered an alternative for some uses of wood in interior finishes.
However, this can be adjusted to achieve enhanced performance and fulfill the requirements for such alternatives and the enhance the scope for its use.
The homogeneous structure in the composite made with geopolymer, rice husks, and coffee provided the material with superior compression resistance compared to barley husks. This resistance can be optimized by adjusting the ratio of husks and sodium hydroxide used in the preparation of the material.
It was shown that the time and mode of drying of the composites made with a geopolymer were the most important factors in obtaining the required strengths of the material. This was also associated with the molding used for the materials since the composites obtain their final form as a result of the heat treatment and the drying conditions. Adjusting temperature and pressure when making the mixture enables the generation of materials with better mechanical properties since their resistance is associated with higher density resulting from the compaction of the samples and the bond between particles.
The moisture present in the material decisively influences its resistance. The material was observed to be fragile when it had high moisture content, breaking when unmolded. However, highly dried samples or those with low moisture content did not present unmolding problems. After heat treatment, the reaction of the matrix and the final resistance of the compound were ensured.
The Al/Si ratio and the cation content of Na, Ca, K, and Mg varied between the different brands of metakaolin. Since these relationships are key factors in the structure of the geopolymer and define its mechanical properties, the metakaolin content should be determined before use to guarantee achievement of the proposed objectives.
The husks are a waste product with added value that can provide an alternative in the manufacture of composites, enabling obtaining materials that are less costly and supporting the use of materials that currently can create contamination. Materials with different properties were obtained depending on the origin, the geometric shape, and the husk quantities used to produce the compounds. This allows for potential combination of materials for a wider range of applications. In addition, the husks can be covered with heat retardant chemicals, water repellents, and wood preservatives, among other treatments, to prolong their useful time and reduce deterioration under environmental conditions. The sol-gel/fly ash and barley husks with sol-gel achieved the best behaviors in flexural and compression tests.