Improving the Early Properties of Treated Soft Kaolin Clay with Palm Oil Fuel Ash and Gypsum

Soft soil problems and increased production of fuel waste have emerged due to world population growth. These two problems are prompting engineers to introduce new methods of using waste fuel to stabilize the soil. Previous research has shown clear sustained improvements in soil properties using palm oil fuel ash (POFA) when mixed with a calcium-based binder such as NaCl, lime or cement. The use of such a stabilizing agent can reduce the economic problems associated with reducing the cost of waste disposal and create a sustainable ecological system. It is an alternative method of replacing part of the soil to ensure a balance between economic growth and ecological privilege, leading to the achievement of green technology goals. However, this research is aimed at improving the properties of processed soft kaolin clay with a combination of POFA and gypsum. The physical and mechanical properties of all samples were tested. The results showed a decrease in the specific gravity with the addition of POFA and an increase with gypsum alone, as well as a decrease with a mixture of POFA and gypsum and a decrease in the soil plasticity index due to a better increase in the plasticity limit compared to the liquid limit. This is considered a sign of improved geotechnical properties and reduced linear shrinkage. It was also shown that the treated clay showed an increase in the optimal water content and a drop in the maximum dry density. Nevertheless, it can be concluded that the initial properties of the processed soft kaolin clay with the addition of POFA can be significantly improved.


Introduction
Global water presence in clay affects the behavior of it to either shrink or swell [1]. The swelling of expansive clay soil is a result of water content variation, which leads to significant destruction of overlying structures; thus, civil engineering researchers consider this issue [2]. Soil stabilization is the most economical way to improve the properties of the problematic soil; generally, soil stabilization is a method of modifying and enhancing soil properties by blending other materials. Improvements include increasing the compressive and shear strength and bearing capabilities to strengthen geotechnical properties and other applications [3]. When a stabilizer is mixed with soil, the action of change takes four stages. The first two stages are defined as soil modification and the last two stages are classified as stabilization. The first stage is cation exchange, followed by flocculation and agglomeration as a result of water gypsum also was added as a pozzolanic activator and stabilizer [43]. Further, soft soil issues and the increment in palm oil waste productions are in correspondence with the increment in global demand for oil because of world population growth. Malaysia is facing the two main issues that encourage engineers to adopt new methods of using palm oil waste in soil stabilization. Previous investigations have shown an apparent means of sustainable enhancement in soil properties by using POFA when mixed with a calcium-based binder such as NaCl, lime or cement. Utilizing POFA as the stabilizing agent can reduce economic issues by reducing the cost of waste landfill and help attain a sustainable ecological system. To stabilize the soft soil, the addition of POFA and gypsum is required to investigate their effect on modifying the properties of the soil. This is an alternative method to replace some of the soil so that it balances between economic growth and environmental privilege that leads to achieving green technology goals.
The novelty of this study is focused on the use of both POFA as pozzolanic waste, and gypsum as binder as well as the inclusion of pozzolan activator to achieve better properties modification and soil stabilization to sustain a higher load. POFA is abundant and cost-effective, and it is a better method to dispose of it in the engineering field than in landfills. Gypsum is more economical than cement or lime; it can be produced in soil, mined and has no negative impact on environment. However, this study aims to improve the early properties of soft kaolin clay with the combination of POFA and gypsum. This study is presents the materials and methodology used, results, discussion and conclusions to explain the investigation and the effectiveness of POFA and gypsum in the stabilization of soft kaolin clay in terms of physical properties, compaction properties of optimum moisture content and maximum dry density.

Materials and Methodology
This research was basically planned to determine the physical properties and the variation of compaction characteristics of the original soft clay and clay treated with POFA and gypsum. The soft clay utilized in this experiment was kaolin; it was substituted and replaced with different percentages of gypsum and POFA. The flow of works followed standard laboratory test procedures and data analysis methods, according to British Standards (BS) and the American Society for Testing and Materials (ASTM). The specific gravity test was conducted according to BS 1377: Part 2: 1990: 8.3 [44]. The liquid limit was conducted according to BS 1377: Part 2: 1990: 4.3 [44], whereas the plastic limit test was conducted according to BS 1377: Part 2: 1990: 5.3 [44]. The linear shrinkage test was conducted according to BS 1377: Part 2: 1990: 6.5 [44]. A standard proctor test was carried out according to BS 1377: Part 4: 1990: 3.3 to determine compaction parameters (optimum moisture content and maximum dry unit weight) [44]. Figure 1 shows the experimental study workflow.  Table 1 POFA and gypsum. Gypsum and kaolin were imported, whereas POFA was collected from LKPP Corporation Sdn.Bhd, Lepar Hilir Palm Oil Mill, Gambang, Kuantan, Pahang, Malaysia. The amount required of gypsum and POFA was added based on the dry weight of the kaolin S300. All materials were dried in the oven at 105 °C for 24 hours.   Table 1 POFA and gypsum. Gypsum and kaolin were imported, whereas POFA was collected from LKPP Corporation Sdn.Bhd, Lepar Hilir Palm Oil Mill, Gambang, Kuantan, Pahang, Malaysia. The amount required of gypsum and POFA was added based on the dry weight of the kaolin S300. All materials were dried in the oven at 105 • C for 24 h. After that, the materials were sieved based on the specific sieve size per test in accordance with British Standard (BS). The mixture prepared was carefully mixed by using a soil mixer until the mixture looked homogenous before testing was carried out. The 4% and 6% gypsum were selected based on the literature review, whereas the POFA (5%, 10%, 15%) percentages were randomly selected to test POFA usage in different amounts with gypsum to treat soil and investigate the enhancement of geotechnical physical and mechanical properties. All the tests were conducted with a minimum of three to five repeated times per sample per test to ensure the consistency and accuracy of the obtained results. Table 1 shows the twelve samples prepared and tested per test.

Test Setup and Procedures
The specific gravity (SG) is the ratio between the mass of soil to its volume. It is essential in relation to other tests, especially for calculating porosity and voids. Kaolin, POFA and gypsum were sieved through a 2 mm sieve size and a small pycnometer test was used. Pycnometer bottles maximum capacity is 100 mL [45]. The specific gravity values of kaolin and POFA were obtained with distilled water, whereas gypsum's specific gravity was obtained using kerosene as dissolving material for gypsum because gypsum hardened when utilizing water to obtain specific gravity; the same method was used by [46,47].
The liquid limits of kaolin, POFA and gypsum are determined by using the cone penetrometer method as it is more accurate and less liable to personal and experimental errors. The liquid limit is determined by obtaining a minimum of four points for plotting the curve. The consistency of the soil specimens is adjusted so the cone penetration ranges between 15 mm and 25 mm. The plastic limits of kaolin, POFA and gypsum are determined from the rolling thread method. Liquid limit and plastic limit are very critical indicators corresponding to undrained shear strengths [48]. The liquid limit was conducted after taking a weight of the sample, 300 g (kaolin or kaolin+ POFA, kaolin+ gypsum or kaolin+ POFA+ gypsum), which passed a 425 µm sieve. Then, distilled water was added to the soil specimen after it was mixed for at least 10 min to be homogenous and mixed with two spatulas. The well-mixed soil specimen was then put into the metal cup. The soil's liquid limit is obtained as moisture content at penetration at 20 mm. The plastic limit represents the moisture content at which soil changes to dry to plastic where it is considered the upper strength limit of consistency. The method of rolling a thread is suggested by Casagrande [49]. The sample was rolled on a glass plate until it crumbled at a diameter of 3 mm, weighed and then put in the oven for 24 h; then, the dry samples were weighed. The plasticity index is a workability indicator defined by the difference between the liquid and Sustainability 2021, 13, 10910 5 of 18 plastic limit of soil. The most challenging aspects in the soil are shrinkage and swelling, relying on the study of soil plasticity.
Linear shrinkage is defined as the change in the length of the soil bar when dried in the range below the liquid limit [50], which can be calculated as a percentage of the oven-dried soil length to the initial length of the soil bar according to BS 1377: Part 2: 1990: 6.4. Samples were sieved and passed a 425 µm sieve. The mold used was about 140 ± 1 mm and had an internal radius of 12.5 ± 0.5. To establish consistent results, after obtaining the liquid limit per sample, water used for mixing was equal to the percentage of water obtained at 20 mm from liquid limit results [44]. The linear shrinkage test determines the shrinkage strain of elongated specimens placed in a mold and subjected to drying in an oven for 24 h [50]. Linear shrinkage is when the contained water content reduced from a specific range reaches the maximum loss of moisture content and there is no more volume change. Linear shrinkage is a sign that indicates the soil reactivity but not necessarily the particular clay present in the soil [51].
The procedure for conducting this test is described in BS 1377: Part 4 1990: 3.3 [44] and ASTM D 698 [52]. The test is conducted to obtain relationships between compacted dry density and soil moisture content, where the measurement of the degree of compaction can be identified in these terms. The materials were sieved and passed through 4.75 mm. When water is mixed with soil, added in a sequent way of 5% until the mass of mold and mixture showed a drop in weight, water plays the role of a softening agent for the soil particles. In this situation, the soil particles near closer to each other. They are forced to move into a dense position. The sample was compacted in three layers with equal thickness into a metal of about 105 mm internal diameter and about 985 cm 3 volume by a metal rammer with a 50 mm diameter circular face, weighing 2.5 kg and dropping from a height of 300 mm with 25 blows in the mold. The obtained samples were collected during the experiment to attain a representative graph on the variation of dry density with the addition of 5% of water. The effective compaction is determined by four dependents: dry density, water content, compaction type (heavy or light) and soil type. When water content added to the soil sample is plotted against dry densities for each cycle, the highest value in the curves is the most important point, as it highlights optimum moisture content (OMC) and the maximum dry density (MMD).

Materials
The kaolin powder grade S300 was used in this project. The used kaolin in our experiments is hydrous alumina silicate with the general chemical of formula Al 2 (Si 2 O 5 ) (OH) 4 . Kaolin is considered as one of the huge natural resource deposits in Malaysia, which has around 112 million tones [53]. Kaolin is one of the tropical problematic soils that is exposed to volumetric change due to the seasonal water inconsistency [39]. It is selected to be treated due to its poor and expansive geotechnical properties, as it shows lows shear strength, high plasticity and low workability [39,40,54].
POFA is a by-product material obtained by burning shells and empty fruit bunches in palm oil mill boilers heated at the rated temperature of 800-1000 • C to generate steam. The obtained steam avails as a source of energy utilized in turbines to supply electricity during the milling operation [55,56]. Raw POFA is unusable because of its unknown moisture content. The POFA is shown in Figure 2; it is a pozzolanic waste material from the palm oil industry. A large amount of amorphous silica in POFA potentially contributes to the pozzolanic reaction during hydration, which results in cementation compounds called calcium aluminate hydrates (CAH) and calcium silicate hydrates (CSH). These compounds are responsible for improving engineering characteristics of soil that increase over time as a pozzolanic reaction. POFA is generated in huge quantities, mainly in developing countries. Indonesia and Malaysia are the primary POFA producers, manufacturing 86% of global supplies according to [10].
Gypsum, POFA and kaolin samples are shown in Figure 2, which exhibits that the color of any material is influenced by the mineralogy property of the material. Gypsum is appeared to be white; while POFA had a black color, but the finer the POFA, the more greyish than black; and kaolin has a creamy to a whitish color. In terms of texture, kaolin is very fine, smooth and soft; POFA has a gritty sandy texture, but it seems to be wet and smooth in particles smaller than 0.425 mm; and gypsum has a gritty, silty sand texture. Table 2 shows the chemical properties.  Gypsum is a soft white mineral consisting of hydrated calcium sulphate, as shown in Figure 2; the chemical formula is calcium sulphate dihydrate (CaSO 4 . 2(H 2 O)). It is a naturally occurring mineral that is comprised of calcium sulphate and water, and it is sometimes called hydrous calcium sulphate. Gypsum has better properties than organic additives because it does not cause air pollution and it is cheaper than cement, fire-resistant and resistant to deterioration by biological factors and chemicals [29,57]. It is also a byproduct of many industrial processes. It is suggested that both coagulation and cementation in soil may be achieved by the addition of gypsum and cement or lime, with significant improvements in soil structure [25]. Malaysia still considers the by-product of gypsum or red gypsum as scheduled waste [58].
Gypsum, POFA and kaolin samples are shown in Figure 2, which exhibits that the color of any material is influenced by the mineralogy property of the material. Gypsum is appeared to be white; while POFA had a black color, but the finer the POFA, the more greyish than black; and kaolin has a creamy to a whitish color. In terms of texture, kaolin is very fine, smooth and soft; POFA has a gritty sandy texture, but it seems to be wet and smooth in particles smaller than 0.425 mm; and gypsum has a gritty, silty sand texture. Table 2 shows the chemical properties. The analysis of particle size curves for representing samples are displayed in Figures 3 and 4. From the graph, approximately 76% of kaolin passed through sieve 0.075 mm; further, 43% passed sieve #270 (0.053 mm) with uniformity coefficient Cu = 1.6 and curvature coefficient CC = 0.87, which meant that the average size of the kaolin particles analysis was in the range of 0.001 mm to 0.2 mm (clay to fine silt). Therefore, with the corresponding results obtained from consistency limits tests, kaolin is classified as a Sustainability 2021, 13, 10910 7 of 18 fine-grained clayey silt (ML), according to USCS and classified under Group A-6 under AASHTO. POFA was classified as (SM) a poorly graded silty sand, corresponding to USCS, as 95% passed #4 (4.75 mm), only 19% passed through sieve #200 (0.075 mm) and 10% passed sieve #270 (0.053 mm) with uniformity coefficient Cu = 4 and curvature coefficient CC = 1. The sieve analysis of gypsum showed that only 4% passed through sieve 0.075 mm; therefore, it is also classified as poorly graded sand and (SP) according to USCS, with Cu = 2.2.

Particle Size Distribution
The analysis of particle size curves for representing samples are displayed in Figures 3 and 4. From the graph, approximately 76% of kaolin passed through sieve 0.075 mm; further, 43% passed sieve #270 (0.053 mm) with uniformity coefficient Cu = 1.6 and curvature coefficient CC = 0.87, which meant that the average size of the kaolin particles analysis was in the range of 0.001 mm to 0.2 mm (clay to fine silt). Therefore, with the corresponding results obtained from consistency limits tests, kaolin is classified as a finegrained clayey silt (ML), according to USCS and classified under Group A-6 under AASHTO. POFA was classified as (SM) a poorly graded silty sand, corresponding to USCS, as 95% passed #4 (4.75 mm), only 19% passed through sieve #200 (0.075 mm) and 10% passed sieve #270 (0.053 mm) with uniformity coefficient Cu = 4 and curvature coefficient CC = 1. The sieve analysis of gypsum showed that only 4% passed through sieve 0.075 mm; therefore, it is also classified as poorly graded sand and (SP) according to USCS, with Cu = 2.2.

Natural Water Content and Specific Gravity
The kaolin, POFA and gypsum were analyzed for natural moisture content directly after being imported and collected. Table 3 illustrates the particle density and natural water content of kaolin, POFA and gypsum. As POFA was collected from an open area, the natural moisture content was the highest, whereas it had the lowest specific gravity. Kaolin marked the highest specific gravity, followed by gypsum. The reason behind the low specific gravity of POFA is because it contains some fibers, and previous researchers found that the specific gravity of POFA can be increased with burning, grinding and treatment, and results confirmed that the finer the POFA, the higher the value of specific gravity [21,59]. Various factors affect the specific gravity, one of them being the fineness of the material, where a higher proportion of silt corresponds to higher specific gravity [60]. The POFA, as an example, is affected by the treatment processes and the burning temperature.

Natural Water Content and Specific Gravity
The kaolin, POFA and gypsum were analyzed for natural moisture content directly after being imported and collected. Table 3 illustrates the particle density and natural water content of kaolin, POFA and gypsum. As POFA was collected from an open area, the natural moisture content was the highest, whereas it had the lowest specific gravity. Kaolin marked the highest specific gravity, followed by gypsum. The reason behind the low specific gravity of POFA is because it contains some fibers, and previous researchers found that the specific gravity of POFA can be increased with burning, grinding and treatment, and results confirmed that the finer the POFA, the higher the value of specific gravity [21,59]. Various factors affect the specific gravity, one of them being the fineness of the material, where a higher proportion of silt corresponds to higher specific gravity [60]. The POFA, as an example, is affected by the treatment processes and the burning temperature. The m finer it is, the higher the specific gravity [21,59]. From Figure 5, K4GP10 showed the lowest specific gravity of 2.59 g/cm 3 , followed by K6GP15, with specific gravity of 2.60 g/cm 3 ; in contrast, the highest specific gravity is of K6G, which reached a value of 2.78 g/cm 3 , followed by K4G, with a value 2.74 g/cm 3 . The specific gravity of kaolin was decreased when POFA was applied alone-it dropped from 2.69 g/cm 3 to 2.61 g/cm 3 compared to the specific gravity of original kaolin, which is 2.70 g/cm 3 . On the other hand, it was increased with the addition of gypsum alone. The overall mixture of both POFA and gypsum demonstrated a reduction in the specific gravity lower than the original soil. One reason behind the specific gravity reduction is due to the low specific gravity of POFA compared to kaolin. Another reason might be the high amount of lightweight buoyant fibrous material in POFA, which float on water. The chemical reactions of POFA and gypsum also play a role in the reduction, where gypsum alone with clay showed an opposite result. An overall investigation of treated clay showed a clear decrement in its specific gravity with the addition of POFA and gypsum.   Figure 6 illustrates the liquid limit results of both untreated and treated kaolin. The addition of POFA and gypsum mixed with kaolin increased its liquid limit. K4GP15 showed the highest value of 43.11% water content, followed by K6GP15 with 42.60%. Gypsum alone at 4% and 6% also showed a slight increment in the liquid limit value, from 40.18% for original soil to 40.30% and 41.19%, respectively. POFA alone mixed with kaolin absorbed much water and showed higher liquid limit in samples K5P, K10P and K15P. One reason for the increment may be the high amount of fiber in POFA, which absorbs a significant amount of water, or it might be the increase in CaO content in POFA and gypsum, which usually absorb more water; these results agree with those of [10,37]. Another reason for the increment in the liquid limit might be due to an elongation of the diffuse double layer as a result of the increment in the specific surface area, which assisted in the  Figure 6 illustrates the liquid limit results of both untreated and treated kaolin. The addition of POFA and gypsum mixed with kaolin increased its liquid limit. K4GP15 showed the highest value of 43.11% water content, followed by K6GP15 with 42.60%. Gypsum alone at 4% and 6% also showed a slight increment in the liquid limit value, from 40.18% for original soil to 40.30% and 41.19%, respectively. POFA alone mixed with kaolin absorbed much water and showed higher liquid limit in samples K5P, K10P and K15P. One reason for the increment may be the high amount of fiber in POFA, which absorbs a significant amount of water, or it might be the increase in CaO content in POFA and gypsum, which usually absorb more water; these results agree with those of [10,37]. Another reason for the increment in the liquid limit might be due to an elongation of the diffuse double layer as a result of the increment in the specific surface area, which assisted in the soft soil's capacity to hold more water; the results showed agreement with soft clay treated with high calcium fly ash investigated by [61]. It was also found that by adding a small amount of ultrafine POFA, the liquid limit increased [24], whereas by adding more than 10%, the liquid limit decreased as along with the plastic limit, which agrees with the results found in this work. In general, the drop in soil plasticity due to using of POFA and POFA-blended gypsum is a sign of enhancement. It was also reported that the high porosity of binder materials with an agglomerated morphology could lead to a reduction in the workability due to the increase in the water absorbed by the large open areas of high porosity POFA [61]. Table 4 shows the standard deviation and variance per sample for the liquid limit tests.

Plastic Limit
The plastic limit of the control sample is 30.72%, as shown in Figure 7. It is appar that the plastic limit was increased, save for a reduction with the addition of sole gyps to kaolin in KG4 and KG6. It then showed a drop to 29.85% and 29.24% in K5P and K6G respectively. On the other hand, overall increments were detected in all other samp especially with the addition of POFA. K4GP15 marked the highest plastic limit of 35.2 followed by K6GP15 with a plastic limit of 35.16%. The previous reasons described affecting the increment in the liquid limit also affected the plastic limit. It is also in preted that a constant increment might be attributed to the agglomeration and floccula that occurred in the particles of the POFA-treated soil, similar to a study on clay trea with POFA and calcium carbide conducted by [23,61]. The enhancement of the pla limit referred to the change in the index of plasticity from silt with medium plasticit low plasticity because of the addition of POFA and gypsum, which have a higher part size than kaolin [62]. Table 5 shows the standard deviation and variance per sample plastic limit tests.

Plastic Limit
The plastic limit of the control sample is 30.72%, as shown in Figure 7. It is apparent that the plastic limit was increased, save for a reduction with the addition of sole gypsum to kaolin in KG4 and KG6. It then showed a drop to 29.85% and 29.24% in K5P and K6GP5, respectively. On the other hand, overall increments were detected in all other samples, especially with the addition of POFA. K4GP15 marked the highest plastic limit of 35.27% followed by K6GP15 with a plastic limit of 35.16%. The previous reasons described as affecting the increment in the liquid limit also affected the plastic limit. It is also interpreted that a constant increment might be attributed to the agglomeration and flocculation that occurred in the particles of the POFA-treated soil, similar to a study on clay treated with POFA and calcium carbide conducted by [23,61]. The enhancement of the plastic limit referred to the change in the index of plasticity from silt with medium plasticity to low plasticity because of the addition of POFA and gypsum, which have a higher particle size than kaolin [62]. Table 5 shows the standard deviation and variance per sample for plastic limit tests.

Plasticity Index
From Figure 8, it can be seen that even though the plasticity index is reduced by adding POFA, more POFA corresponds to more deduction in plasticity index value. The plasticity index was slightly increased with the addition of gypsum because of the clear raise in soil plastic limit to a tendency higher than the increment in the liquid limit with the addition of both POFA and gypsum. K4G, K6G and K6GP5 showed a small increment due to the chemical reaction of soil with gypsum. K4GP15 was marked by the lowest plasticity index, with 7.84%. Consistency limits of treated soil with gypsum found that the decrement of the PI is controlled by shearing resistance and the double-layered diffusion of clay [25]. Table 6 demonstrates the standard deviation and variance per sample for plasticity index.    Figure 8, it can be seen that even though the plasticity index is reduced by adding POFA, more POFA corresponds to more deduction in plasticity index value. The plasticity index was slightly increased with the addition of gypsum because of the clear raise in soil plastic limit to a tendency higher than the increment in the liquid limit with the addition of both POFA and gypsum. K4G, K6G and K6GP5 showed a small increment due to the chemical reaction of soil with gypsum. K4GP15 was marked by the lowest plasticity index, with 7.84%. Consistency limits of treated soil with gypsum found that the decrement of the PI is controlled by shearing resistance and the double-layered diffusion of clay [25]. Table 6 demonstrates the standard deviation and variance per sample for plasticity index. addition of both POFA and gypsum. K4G, K6G and K6GP5 showed a small increment due to the chemical reaction of soil with gypsum. K4GP15 was marked by the lowest plasticity index, with 7.84%. Consistency limits of treated soil with gypsum found that the decrement of the PI is controlled by shearing resistance and the double-layered diffusion of clay [25]. Table 6 demonstrates the standard deviation and variance per sample for plasticity index.

Linear Shrinkage
The results of linear shrinkage in Figure 9 show that kaolin does not have a high tendency to shrink, as only a 1.55% reduction in the mold length occurred. The results show that linear shrinkage was generally reduced with the addition of POFA or gypsum, where KG6 illustrated the lowest value of linear shrinkage with only 0.5%, followed by KP15 with 0.77%. The linear shrinkage showed a different reaction with the mixture of both POFA and gypsum. The value increased to 2.13% for K6GP10, which marked the highest overall value among all samples. The slight reduction that occurred is predictable due to the chemical reactions between gypsum with POFA and clay, which lead to the reduction in plasticity properties and related linear shrinkage [50].  The results of linear shrinkage in Figure 9 show that kaolin does not have a high tendency to shrink, as only a 1.55% reduction in the mold length occurred. The results show that linear shrinkage was generally reduced with the addition of POFA or gypsum, where KG6 illustrated the lowest value of linear shrinkage with only 0.5%, followed by KP15 with 0.77%. The linear shrinkage showed a different reaction with the mixture of both POFA and gypsum. The value increased to 2.13% for K6GP10, which marked the highest overall value among all samples. The slight reduction that occurred is predictable due to the chemical reactions between gypsum with POFA and clay, which lead to the reduction in plasticity properties and related linear shrinkage [50].

Variation of Dry Density with Addition of Water
The compaction curves were plotted, as shown in Figures 10-13. From Figure 10, water absorption was increased by adding POFA dosage; it is apparent that with more POFA, higher water content is needed to obtain the optimum moisture content and maximum dry density. With the addition of more POFA content to kaolin, the maximum dry density achieved was decreasing accordingly. It can be interpreted as the POFA having

Variation of Dry Density with Addition of Water
The compaction curves were plotted, as shown in Figures 10-13. From Figure 10, water absorption was increased by adding POFA dosage; it is apparent that with more POFA, higher water content is needed to obtain the optimum moisture content and maximum dry density. With the addition of more POFA content to kaolin, the maximum dry density achieved was decreasing accordingly. It can be interpreted as the POFA having lower specific gravity compared to kaolin and gypsum. The results showed an increase in OMC of the treated kaolin with POFA, results which are in line with previous work [63][64][65][66]. The referred increment can be explained as a result of crowding out of calcium ions released from POFA during ionic dissociation of hydrolyzed calcium oxide of POFA and pozzolanic reaction between calcium ions of POFA and SiO 2 in kaolin, where both are considered as a factor of soil chemical stabilization.
Sustainability 2021, 13, x FOR PEER REVIEW 13 pozzolanic reaction between calcium ions of POFA and SiO2 in kaolin, where both considered as a factor of soil chemical stabilization. As seen in Figure 11, there was a slight reduction in maximum dry density an crement in the optimum water content of the treated samples with the addition of sum. The change in standard compaction test results was not high with the additio gypsum only, compared to the control sample. These results agree with previous rese works that reported that the addition of gypsum to clay soil can increase the absorp of water content needed for the chemical reaction [34]. The addition of different percentages of POFA with 4% and 6% of gypsum conte shown in Figure 12. The maximum dry density of KG4 and KG6, 1.63 gm/cm 3 and  As seen in Figure 11, there was a slight reduction in maximum dry density crement in the optimum water content of the treated samples with the addition sum. The change in standard compaction test results was not high with the add gypsum only, compared to the control sample. These results agree with previous r works that reported that the addition of gypsum to clay soil can increase the abs of water content needed for the chemical reaction [34]. The addition of different percentages of POFA with 4% and 6% of gypsum co shown in Figure 12. The maximum dry density of KG4 and KG6, 1.63 gm/cm 3 a gm/cm 3 , respectively, was higher than the control sample, but it slightly dropped t the lowest at K4GP15 with 1.46 gm/cm 3 compared to 1.58 gm/cm 3 as MDD of kaol can be explained as a result of higher water absorption required by porous POFA conquers the pore space of POFA soil and leads to particle buoyancy [65]. The amount of fiber in the added POFA also affected the reduction in maximum dry  and the specific gravity of treated clay [19,67]. The reduction in maximum dry dens might be due to the resistance of compaction as a result of the flocculation of soil partic during soil stabilization [5].

Maximum Dry Density (MDD)
The maximum dry density increased from 1.59 gm/cm 3 for kaolin alone to 1.63 g/c and 1.62 g/cm 3 for K4G and K6G, respectively, as the gypsum absorbs more water dur compaction, as shown in Figure 15; these results are in line with those presented [32,35,68,69]. However, when adding up 5%, 10% and 15% of POFA with gypsum, MDD continued to drop until reaching 1.465 g/cm 3 for K6GP15 as the lowest. On the ot hand, POFA alone mixed with kaolin of 15% started to increase, as the maximum density of K10P is 1.475 g/cm 3 whereas K15P reached 1.482 g/cm 3 . A drop in the maxim dry density value by adding POFA is predicted due to the too low specific gravity POFA (2.25) [67]. Previous results on POFA used as a stabilizer with lime conducted [19] found that the density of soil is affected by lime content-predicted to be presen our gypsum-where the lime tends to absorb the water and hence, decrease the maxim and the specific gravity of treated clay [19,67]. The reduction in maximum dry dens might be due to the resistance of compaction as a result of the flocculation of soil partic during soil stabilization [5].

Maximum Dry Density (MDD)
The maximum dry density increased from 1.59 gm/cm 3 for kaolin alone to 1.63 g/c and 1.62 g/cm 3 for K4G and K6G, respectively, as the gypsum absorbs more water duri compaction, as shown in Figure 15; these results are in line with those presented [32,35,68,69]. However, when adding up 5%, 10% and 15% of POFA with gypsum, t MDD continued to drop until reaching 1.465 g/cm 3 for K6GP15 as the lowest. On the oth hand, POFA alone mixed with kaolin of 15% started to increase, as the maximum d density of K10P is 1.475 g/cm 3 whereas K15P reached 1.482 g/cm 3 . A drop in the maximu dry density value by adding POFA is predicted due to the too low specific gravity POFA (2.25) [67]. Previous results on POFA used as a stabilizer with lime conducted [19] found that the density of soil is affected by lime content-predicted to be present our gypsum-where the lime tends to absorb the water and hence, decrease the maximu  Figure 13. Variation of MDD and OMC of treated kaolin with 4% gypsum and various POFA content.
As seen in Figure 11, there was a slight reduction in maximum dry density and increment in the optimum water content of the treated samples with the addition of gypsum. The change in standard compaction test results was not high with the addition of gypsum only, compared to the control sample. These results agree with previous research works that reported that the addition of gypsum to clay soil can increase the absorption of water content needed for the chemical reaction [34].
The addition of different percentages of POFA with 4% and 6% of gypsum content is shown in Figure 12. The maximum dry density of KG4 and KG6, 1.63 gm/cm 3 and 1.62 gm/cm 3 , respectively, was higher than the control sample, but it slightly dropped to reach the lowest at K4GP15 with 1.46 gm/cm 3 compared to 1.58 gm/cm 3 as MDD of kaolin. This can be explained as a result of higher water absorption required by porous POFA, which conquers the pore space of POFA soil and leads to particle buoyancy [65]. The higher amount of fiber in the added POFA also affected the reduction in maximum dry density and the specific gravity of treated clay [19,67]. The reduction in maximum dry density might be due to the resistance of compaction as a result of the flocculation of soil particles during soil stabilization [5].

Maximum Dry Density (MDD)
The maximum dry density increased from 1.59 gm/cm 3 for kaolin alone to 1.63 g/cm 3 and 1.62 g/cm 3 for K4G and K6G, respectively, as the gypsum absorbs more water during compaction, as shown in Figure 15; these results are in line with those presented by [32,35,68,69]. However, when adding up 5%, 10% and 15% of POFA with gypsum, the MDD continued to drop until reaching 1.465 g/cm 3 for K6GP15 as the lowest. On the other hand, POFA alone mixed with kaolin of 15% started to increase, as the maximum dry density of K10P is 1.475 g/cm 3 whereas K15P reached 1.482 g/cm 3 . A drop in the maximum dry density value by adding POFA is predicted due to the too low specific gravity of POFA (2.25) [67]. Previous results on POFA used as a stabilizer with lime conducted by [19] found that the density of soil is affected by lime content-predicted to be present in our gypsum-where the lime tends to absorb the water and hence, decrease the maximum dry density of treated soil. The maximum dry densities of kaolin and treated kaolin are shown in Figure 14. Table 7 displays the standard deviation and variance per sample for the liquid limit tests. dry density of treated soil. The maximum dry densities of kaolin and treated kaolin are shown in Figure 14. Table 7 displays the standard deviation and variance per sample for the liquid limit tests.

Optimum Moisture Content (OMC)
The optimum water content of kaolin and POFA-gypsum-treated kaolin is shown in Figure 15. The optimum water content increased because the tendency of POFA for water absorption is high. This tendency is clearly explained by OMC, which increased from 18% for the control sample to 24% for KP10, but when adding more POFA, the water content drops until reaching 19.2% for KP15. The water content is slowly increased when gypsum is added compared to the values of POFA, therefore, with kaolin soil, the more POFA added, the higher the optimum moisture content achieved, similar to results published by [30,65,67,70]. The high amount of CaO in gypsum can be one of the reasons behind the increment in water absorption by treated soil; [3] mentioned that lime is known to decrease the plasticity index and maximum dry density of the soil and increase its optimum water content. Table 8 summaries the standard deviation and variance of OMC values.   The optimum water content of kaolin and POFA-gypsum-treated kaolin is shown in Figure 15. The optimum water content increased because the tendency of POFA for water absorption is high. This tendency is clearly explained by OMC, which increased from 18% for the control sample to 24% for KP10, but when adding more POFA, the water content drops until reaching 19.2% for KP15. The water content is slowly increased when gypsum is added compared to the values of POFA, therefore, with kaolin soil, the more POFA added, the higher the optimum moisture content achieved, similar to results published by [30,65,67,70]. The high amount of CaO in gypsum can be one of the reasons behind the increment in water absorption by treated soil; [3] mentioned that lime is known to decrease the plasticity index and maximum dry density of the soil and increase its optimum water content. Table 8

Conclusions
This research is aimed at improving the properties of processed soft kaolin clay with a combination of POFA and gypsum. POFA and gypsum were separately utilized to treat kaolin to evaluate the singular effect per material. POFA alone exhibited low effectiveness in enhancing soil properties; it is better to be mixed with gypsum. The changes in compressive and shear strength capabilities provided with SEM and EDX can highlight a greater effect of pozzolanic reaction with 0 days, (24 hours) 1 day, 7 days and 28 days of curing. Over the course of the studies carried out, the following main results were obtained:

•
In general, kaolin had the highest specific gravity of 2.70 compared to gypsum (2.66) and POFA (2.25). The result also showed a reduction in specific gravity of kaolin when mixed with POFA; however, specific gravity of kaolin is increased when mixed with gypsum, likely due to the type of chemical composition generated and the degree of decomposition or dissolution with water.

•
The liquid limit increased with the addition of gypsum and POFA, as they absorb more water. The plastic limit increased due to the sand properties of POFA and gypsum, and thus, the plasticity index marked a clear decrement that can be a sign of the improvement of soil properties when treated with POFA and gypsum. The effect of sandy POFA and gypsum also decreased the linear shrinkage of kaolin.

•
The compaction test is considered an important geotechnical indicator in this research, as other tests such as unconfined compression tests and falling head permeability tests depend on the results obtained from this test. The result of this test demonstrated an overall increment in optimum water content by adding more POFA and a decrement in maximum dry density, related to the porosity and low specific gravity of POFA as well as the tendency of water absorption by both POFA and gypsum. Meanwhile, the results did not show a remarkable increment in water content when kaolin was mixed with gypsum alone.

Conclusions
This research is aimed at improving the properties of processed soft kaolin clay with a combination of POFA and gypsum. POFA and gypsum were separately utilized to treat kaolin to evaluate the singular effect per material. POFA alone exhibited low effectiveness in enhancing soil properties; it is better to be mixed with gypsum. The changes in compressive and shear strength capabilities provided with SEM and EDX can highlight a greater effect of pozzolanic reaction with 0 days, (24 h) 1 day, 7 days and 28 days of curing. Over the course of the studies carried out, the following main results were obtained:

•
In general, kaolin had the highest specific gravity of 2.70 compared to gypsum (2.66) and POFA (2.25). The result also showed a reduction in specific gravity of kaolin when mixed with POFA; however, specific gravity of kaolin is increased when mixed with gypsum, likely due to the type of chemical composition generated and the degree of decomposition or dissolution with water.

•
The liquid limit increased with the addition of gypsum and POFA, as they absorb more water. The plastic limit increased due to the sand properties of POFA and gypsum, and thus, the plasticity index marked a clear decrement that can be a sign of the improvement of soil properties when treated with POFA and gypsum. The effect of sandy POFA and gypsum also decreased the linear shrinkage of kaolin.

•
The compaction test is considered an important geotechnical indicator in this research, as other tests such as unconfined compression tests and falling head permeability tests depend on the results obtained from this test. The result of this test demonstrated an overall increment in optimum water content by adding more POFA and a decrement in maximum dry density, related to the porosity and low specific gravity of POFA as well as the tendency of water absorption by both POFA and gypsum. Meanwhile, the results did not show a remarkable increment in water content when kaolin was mixed with gypsum alone.