Eco-Friendly Improvement of Comprehensive Engineering Properties of Collapsible Loess Using Guar Gum Biopolymer

Abstract

Collapsible loess is characterized by its unique soil-forming environment, mineral composition, and microstructure, resulting in poor engineering properties such as high water sensitivity, high collapsibility, high compressibility, and low strength. To improve the poor engineering properties of collapsible loess, we selected a suitable eco-friendly material—guar gum (GG)—for its improvement and reinforcement, and investigated the improvement effect of different GG dosages (0.5~1.5%) and curing ages (0~28 days) on collapsible loess. The mechanical properties of soil samples were determined by direct shear tests, unconfined compressive strength tests, and splitting tests. The water stability of soil samples was evaluated by both cube and sphere crumb tests. SEM and EDS analyses were also conducted to determine the microstructural and mineral changes in soil. The results indicate that the incorporation of GG is beneficial to inhibit the collapsibility of the soil and improves the water stability and strength of the soil. The collapsibility coefficient of loess is reduced to below 0.015 when 0.75% and above of GG is admixed, which is considered a complete loss of its collapsibility. When the GG dosage increases from 0% to 1.25%, the compressive strength and tensile strength of the soil samples increase by 43.5% and 34.9%, respectively. However, by further increasing the GG dosage to 1.5%, the compressive strength and tensile strength decrease by 3.8% and 6% compared to those with 1.25% GG. This indicates that the strength of the specimens shows an increasing trend and then a decreasing trend with the increase in GG dosage, and 1.25% GG was found to be the best modified dosage. Microstructural and mineral analyses indicate that the addition of GG does not change the mineral composition of loess, but, rather, it significantly promotes the agglomeration and bonding of soil particles through cross-linking with Ca²⁺ ions in the soil to form a biopolymer network, thus achieving a reliable reinforcement effect. Compared with the existing traditional stabilizers, GG is a sustainable and eco-friendly modified material with a higher low-carbon value. Therefore, it is very necessary to mix GG into collapsible loess to eliminate some of the poor engineering properties of loess to meet engineering needs. This study can provide test support for the application and promotion of GG-modified loess in water agriculture and road engineering.

1. Introduction

As a special aeolian soil, loess is widely distributed in many countries, such as the United States, the United Kingdom, and China [1]. Among these countries, about 6.6% of Chinese land area is covered by loess, and the total area of loess is about 6.31 × 10⁵ km² [2]. The unique soil-forming environment, mineral composition, and microstructure of loess results in a loose structure, large pores, weak cementation ability, high soluble salt content, and low soil strength [3]. These shortcomings further lead to the poor engineering properties of loess, such as high compressibility, strong water sensitivity, and strong collapsibility [4]. Consequently, if the loess is wetted by water, a large amount of water enters the soil through the pores, causing rapid damage to the soil structure and a sharp decline in strength. Under the action of external load and its own gravity, it is easy to produce a large additional settlement. In the process of engineering construction, problems such as uneven foundation settlement, road collapse, and tunnel collapse are often encountered. Therefore, in practical engineering, various physical or chemical methods are usually used to improve and reinforce collapsible loess in order to achieve the purpose of eliminating its poor engineering properties [5]. The main purpose of physical reinforcement is to reinforce soil by changing the internal structure of soil and soil particle gradation. The most common methods include the dynamic compaction method, vibration compaction method, and displacement filling method [6]. However, the physical reinforcement process may produce noise and even cause a series of vibration problems, and the construction cost is high, requiring a large investment of consumables and economic resources, which has certain drawbacks. The principle of chemical reinforcement refers to improving the cohesion and cementation ability of soil by adding some kind of reinforcement materials to interact or chemically react with soil, so as to achieve the purpose of improving the mechanical properties of soil [7]. Although some traditional reinforcement materials, such as lime, cement, and fly ash, have played a role in improving loess, the widespread use of these materials has significant limitations due to their disadvantages, such as environmental pollution and high carbon emissions, which have brought about serious environmental problems [8]. In view of the shortcomings of traditional reinforcement materials, it is important to find eco-friendly reinforcement materials for collapsible loess.

Biopolymers are natural polymers produced by plants, microorganisms, and other living organisms [9] which have the advantages of ecological and environmental protection, rich variety, and sustainable regeneration, and are mostly used as thickeners and stabilizers in the food and medical fields [10,11]. In recent years, biopolymers have been widely studied as a new type of soil stabilizer in some geotechnical engineering applications. For example, Latifi et al. [12] investigated the microscopic behavior and characteristics of bentonite and kaolin reinforced by xanthan gum (XG) biopolymer, and the results showed that 1~1.5% XG can effectively improve the strength and stiffness of bentonite and kaolin. Chang et al. [13] discussed the influencing factors on the behavior of soil strengthening by gellan gum and agar gum biopolymers and found that water content was a key parameter for soil strengthening by thermo-gelation biopolymers. Cabalar et al. [9] investigated the mechanical properties of XG-modified low-plasticity clay. The test results showed that the strength of the clay specimens increased significantly with the increase in XG content and the prolongation of the curing time. Reddy et al. [14] utilized GG and XG to improve the high dispersivity of red mud waste, and the results showed that GG had a slightly better improvement effect than XG. Bao et al. [15] conducted protection tests using GG and polypropylene fibers on a typical bare slope of the Loess Plateau, and the results showed that both materials had good protection effects. GG proved to be more effective than polypropylene fibers in slope initial protection effects. Reddy et al. [16] treated dispersive soil with XG biopolymer and found that the addition of XG could significantly improve the long-term strength and durability of dispersive soil. Zhao et al. [17] investigated the improvement effect of different dosages of sodium alginate on collapsible loess, and the results showed that the strength and water stability of collapsible loess were significantly improved by the addition of sodium alginate. Chang et al. [18] found that the expansion index, plasticity index, and shear modulus of residual soil gradually increased with the increase in Beta 1,3/1,6 glucan biopolymer content in the soil. However, the addition of Beta 1,3/1,6 glucan had no effect on the compressive stiffness of the residual soil.

Among these biopolymers, GG and XG show great potential and broad prospects in soil treatment. Previous studies have revealed that GG is more effective in improving soil strength, collapsibility, and water stability compared to XG [19,20]. Although GG has achieved some success in improving the engineering properties of some special soil, GG is still underutilized in practical geotechnical engineering. Based on this, the feasibility of applying GG to improve collapsible loess was investigated in this paper. The effects of GG dosage and curing time on the mechanical properties of collapsible loess were analyzed by direct shear tests, unconfined compressive strength tests, and splitting tests. The water stability of collapsible loess was also evaluated by both cube and sphere crumb tests, and the fractal dimension of the specimens after disintegration was measured by the box dimension method. Finally, by analyzing the microstructure and mineral composition of collapsible loess, the mechanism of GG improvement of collapsible loess was elucidated. This study provides certain scientific bases for the safe and scientific application of GG in the reinforcement of collapsible loess areas.

2. Materials and Test Methods

2.1. Materials

2.1.1. Soil Samples

The collapsible loess used in this test was taken from Sanmenxia, Henan Province. The soil samples taken were crushed and dried, and then passed through a 2 mm sieve for preparation. According to the Standard for Geotechnical Test Methods (GB/T 50123-2019) [21], the basic physical property indicators of the soil samples used in the test were measured and are shown in

Table 1. Basic physical properties of collapsible loess.

ρ/(g/cm3)	w/%	wL/%	wP/%	IP/%	e0	GS
1.46	10.06	23.71	14.08	9.63	1.04	2.70

. Referring to the Highway Geotechnical Test Regulations (JTG 3430-2020) [22], the compaction test was carried out by a heavy-duty compaction instrument. The maximum dry density of the specimen was measured to be 1.81 g/cm³ and the optimum moisture content was 15.4%. The test results are shown in Figure 1. The particle size distribution curve of soil samples measured by a laser particle size analyzer is shown in Figure 2. According to the calculation of the soil samples’ particle size distribution curve results, the uniformity coefficient and gradation coefficient of the soil samples are calculated as C_u = D₆₀/D₁₀ = 22.85 > 5, 1 < C_c = D₃₀²/(D₆₀ × D₁₀) = 2.31 < 3, so the soil sample particle is well graded.

2.1.2. Guar Gum

GG is a kind of high-molecular-weight polysaccharide material extracted from the guar gum plant and consists of α-D-galactose and β-D-mannose [23], as shown in Figure 3. GG has higher water solubility compared to other biopolymers and it is also naturally charged and exists in cationic and anionic states [20]. Due to the presence of a large number of hydroxyl ions in GG, the hydroxyl ions form hydrogen bonds with the hydrated minerals, and through the hydrogen bonds, a hydrogel network is formed between the soil particles and H⁺, which enhances the degree of bonding between the soil particles [24]. Therefore, GG can be a good soil stabilizer.

2.1.3. Sample Preparation

The collapsible loess taken from the field was dried, crushed, and screened by a 2 mm sieve for preparation. Using the dry mixing method, GG and collapsible loess were mixed and stirred well according to a GG-to-collapsible-loess mass ratio of 0%, 0.5%, 0.75%, 1%, 1.25%, and 1.5%. An appropriate amount of distilled water was added to the mixture according to the optimum moisture content of the soil samples and the treated soil samples were placed in a Ziploc bag and left to stand for 24 h to ensure that the water in the soil samples was evenly dispersed. Subsequently, the GG–loess mixture was compacted into a cylindrical specimen of Φ39.1 mm × 80 mm (diameter × height) and a ring blade specimen of Φ61.8 mm × 20 mm (diameter × height) using a static compaction method. According to the requirements of each test, the compaction degree of the collapsibility test specimen was 0.85, and the compaction degree of the rest of the test specimens was 0.94. Among them, each group of direct shear tests prepared 4 specimens and we prepared 3 groups of the same parallel specimens, for a total of 288 specimens. Three parallel specimens were prepared for an unconfined compressive strength test and a splitting test, respectively, each with a total of 72 specimens for each test. As the collapsibility test adopted the double-line method, 2 specimens were prepared in each test group, and 3 groups of identical parallel specimens were prepared, totaling 144 specimens. The prepared specimens were sealed with cling film and placed in a standard curing box with a temperature of 20 ± 2 °C and a relative humidity of 95 ± 3%, and were cured for 0 d, 7 d, 14 d, and 28 d. This study mainly used GG to improve collapsible loess from the perspective of construction. If the curing time of GG-modified collapsible loess is too long in the construction process, the curing cost is too high, so the curing age selected in this study was 28 d.

2.2. Test Methods

2.2.1. Mechanical Tests

Shear strength (SS), unconfined compressive strength (UCS), and tensile strength (TS) are important mechanical indexes to evaluate the strength of soil. In this paper, the improvement behavior of GG on the mechanical properties of collapsible loess specimens was investigated through direct shear tests, unconfined compressive strength tests, and splitting tests. The instrument used for the direct shear test was a ZJ-type strain-controlled direct shear instrument. According to the Highway Geotechnical Test Regulations (JTG 3430-2020) [22], the shear rate was set at 0.08 mm/min. The vertical pressures applied during the test were 100 kPa, 200 kPa, 300 kPa, and 400 kPa. The unconfined compressive strength test and splitting test were conducted by a YYW-II strain unconfined pressure instrument produced by Nanjing Soil Instrument Factory. During the test, the strain control method was mainly adopted to maintain a rate of l mm/min for loading. The sizes of the specimens were Φ39.1 mm × 80 mm and Φ61.8 mm × 20 mm, respectively.

2.2.2. Water Stability Tests

The water stability of soil was evaluated by a cube sample crumb test (CSC test) and a remolded sphere sample crumb test (RSSC test) from ASTM D2487-17el [25] and D6572-21 [26]. The specimens for the CSC test and RSSC test were a cube with a side length of 10 mm and a sphere with a diameter of 10 mm, respectively, as shown in Figure 4. In this case, the cube specimen was cut from the cylindrical specimen, and the spherical specimen was obtained by remolding the cylindrical specimen. At the beginning of the test, the cube specimens and the remolded spherical specimens were placed into a beaker containing 200 mL of distilled water. The disintegration phenomenon of each specimen was observed and recorded. The area and fractal dimension of each specimen after disintegration were calculated using the box dimension method.

2.2.3. Collapsibility and Consolidation Tests

Both the collapsibility test and the consolidation test were carried out using a WG-type single lever consolidation instrument produced by Nanjing Soil Instrument Factory, and with reference to the Standard for Geotechnical Test Methods (GB/T 50123-2019) [21]. The load was applied in stages during the test, and the stability standard under each stage of load was that the deformation did not exceed 0.01 mm per hour. Among them, the collapsibility test was conducted with compressive loads of 50 kPa, 100 kPa, 150 kPa, and 200 kPa. The coefficient of collapsibility of the soil samples was determined by the double line method. The compressive load of the consolidation test was divided into 6 grades: 50 kPa, 100 kPa, 200 kPa, 300 kPa, 400 kPa, and 600 kPa. The average value of compression coefficients measured by 3 parallel specimens was taken as the measurement result for each group of tests.

2.2.4. pH Test

An INESA DOSJ-308A pH indicator supplied by Shanghai Yidian Scientific Instrument Company was used to measure the pH of the specimen. The 10 g soil sample was taken from the center of specimens and put into a test tube containing 50 mL of distilled water, and the solution was shaken well by a VM-300S vortex mixer. The solution was allowed to stand for 48 h and then the supernatant was measured to determine the pH.

2.2.5. Microstructure and Mineral Analysis

In this study, SEM and EDS tests were conducted by a HITACHI0-S-4800 scanning electron microscope to further analyze the reinforcement mechanism of collapsible loess after GG improvement from the microstructure level. Firstly, representative samples were dried and then cut to the size required to meet the test requirements. The microstructure of the cut samples was analyzed by SEM after a gold spraying treatment. Afterwards, a part of the sample was ground and passed through a 0.075 mm sieve, and XRD analysis was performed using a RINT-2000 analyzer in the range of 2θ = 5~75° with a controlled scanning rate of 5°/min [27]. At the same time, a Netzsch STA449F3 instrument was used to conduct TG analysis in the temperature range of 30~1000 °C and the temperature rise rate was controlled at 10 °C/min [28].

3. Results

3.1. Mechanical Properties

3.1.1. Shear Strength

SS is one of the important parameters for evaluating the improvement effect of soil strength, and its magnitude mainly depends on the cohesion and internal friction angle of the soil. In order to investigate the effect of GG on the SS of collapsible loess, a direct shear test was conducted and the results are shown in Figure 5 and Figure 6. It is noteworthy that the SS, cohesion, and internal friction angle of GG-modified loess are significantly improved compared to natural collapsible loess. This enhancement can be attributed to the reaction of GG with water and the formation of GG hydrogels [24]. These gelling products weld the soil particles together and fill the pores of the GG–loess matrix, making the soil samples denser and increasing the cohesion and internal friction angle between the soil particles, thus improving the SS of the soil [12]. Meanwhile, the SS of GG-modified loess shows a tendency of increasing and then decreasing with the increase in GG dosage. As an example, the SS of GG-modified loess increases by 16.2% and 34.1% compared to natural collapsible loess when the GG dosage is 0.5% and 1.25%, respectively, for 14 d of curing time and σ_n = 200 kPa. However, by further increasing the GG dosage to 1.5%, the SS decreases by 4.5% compared to that with 1.25% GG. The results show that GG does not increase the SS of the soil in a linear way, but that there is an optimum dosage. From the test data, the optimum dosage of GG-modified loess is 1.25%. In addition, the SS of the GG-modified loess increases with the prolongation of the curing time. However, the improvement in the SS is more significant in the first 7 d of curing, while the improvement in the SS levels off with a longer curing time.

Figure 6 shows the test results of cohesion and the internal friction angle of different specimens. In this case, the bar graph and the dashed line indicate the cohesion and the internal friction angle, respectively. From Figure 6, it can be seen that increasing the dosage of GG and the curing time has a positive effect on improving the cohesion and the internal friction angle of GG-modified loess. The trends of cohesion and internal friction angle are basically consistent with the changes in SS, which further indicates that GG can increase the SS of collapsible loess by increasing the cohesion and internal friction angle between soil particles. The effect of GG on the internal friction angle is relatively limited compared to the increase in the cohesion of collapsible loess with GG.

3.1.2. Compressive and Tensile Strengths

Figure 7 shows the effects of different GG dosages and different curing ages on the UCS and TS of collapsible loess. Among them, the bar graph indicates the UCS and the dashed line indicates the TS. As can be seen from Figure 7, the addition of GG can significantly improve the UCS and TS of collapsible loess, and the UCS and TS increase with the increase in GG dosage. This trend is consistent with the results of other studies [12,20,29,30]. After 14 d of curing, the UCS and TS of collapsible loess increase from 379.17 kPa and 38.93 kPa to 544.18 kPa and 52.5 kPa when the dosage of GG is increased from 0% to 1.25%, which is an increase of 43.5% and 34.9%, respectively. The increase in UCS and TS with the increase in GG dosage can be attributed to a number of factors. On the one hand, with the increase in GG dosage, more and more GG hydrogels cross-link with Ca²⁺ ions in the soil to form a network structure of mutual penetration, which increases the contact area between soil particles and soil and improves the bonding and friction between soil particles [31]. On the other hand, the hydroxyl groups present in GG interact with hydrated minerals to form hydrogen bonds. Under the action of hydrogen bonds and van der Waals forces, the degree of connection between the soil particles increases, and the UCS and TS increase correspondingly [24].

The UCS and TS of collapsible loess show an increasing trend with the prolongation of curing time at the same GG dosage. The UCS and TS of the specimens increase significantly after 28 d of curing. When the GG dosage is 1.25%, the UCS and TS of the specimens after 28 d of curing are increased by 15.56% and 20.52%, respectively, compared with the uncured specimens. This enhancement can be explained by the growth of gelling products. With the prolongation of curing time, the hydration proceeds further, which promotes the growth of gelling products and enhances the bonding and agglomeration of soil particles [12]. Therefore, it can be concluded that GG needs a certain amount of curing time to realize its potential.

3.2. Water Stability

3.2.1. Disintegration Area

In order to test the effect of GG on the water stability of collapsible loess, RSSC and CSC were carried out on GG-modified loess, and the test results are shown in Figure 8, Figure 9 and Figure 10. As can be seen from Figure 8, for the RSSC test, the specimen begins to be in the state of water absorption, and small cracks appear in the center of the specimen. At this time, the cohesion between soil particles is relatively large, and the specimen does not collapse. With the continuous entry of water into the pore space of the specimen, on the one hand, it causes the dissolution of salts inside the soil and the structure of the soil is rapidly damaged. On the other hand, it leads to the thickness of the water film on the surface of the soil particles becoming thicker, the cohesion between the soil particles being reduced, the cracks constantly becoming bigger, and the specimen disintegrating. In the CSC test, cracks appear at the edge of the specimen first. With the passage of time, the pores of the specimen continue to be filled with water and the cracks spread to the middle of the specimen. The development of cracks intensifies the disintegration of the specimen, resulting in a large amount of disintegration material at the bottom of the beaker.

As can be seen from Figure 9, the disintegration material of natural collapsible loess before modification is loose fine particles and easy-to-produce turbid liquid. After the addition of GG, the disintegration material of collapsible loess turns into clear large particles or agglomerates that cannot form turbid liquid. It is noteworthy that the mud ball specimens are basically not disintegrated after 28 d of curing. The above results indicate that the disintegration resistance of the specimens is significantly improved with the prolongation of the curing time and the increase in GG dosage. This enhancement is mainly due to the fact that GG reacts with water to form hydrogels. The hydrogel reduces the seepage channels of water in the soil by attaching to the surface of the soil particles or filling the pores between the soil particles, thereby preventing the structural damage of the soil caused by water entering the soil through the pores and causing the dissolution of salts. [12]. At the same time, hydrogel aggregates individual soil particles into larger agglomerates through cementation, which increases the cohesion between soil particles and limits the movement of soil particles, thus preventing the generation and further development of disintegration and improving the anti-disintegration and water stability of collapsible loess [14].

In order to increase the accuracy of the test results, ImageJ software (https://imagej.net/software/fiji/downloads, accessed on 22 November 2024) is utilized in this paper to calculate the area size of different specimens after disintegration, and the results are shown in Figure 10. In the RSSC test, the disintegration area of the specimens decreases and tends to be stable with the increase in curing time and GG dosage, which indicates that when the GG dosage reaches a certain level, the mud ball specimens are in a stable state and basically no longer disintegrate. In the CSC test, the disintegration area of the specimens decreases with the increase in curing time and GG dosage. After 28 d of curing, when the GG dosage reaches 1%, the disintegration area of the mud ball specimen is reduced by 71.8% compared with the natural collapsible loess specimen, and the fragment specimen is reduced by 41.2%. This indicates that GG can effectively improve the anti-disintegration ability and water stability of collapsible loess. Therefore, the addition of GG in actual projects is conducive to reducing soil loss due to soil disintegration, which in turn prevents the occurrence of natural disasters such as landslides and foundation settlement.

It should also be noted that when the GG content reaches 1% and above, the mud ball specimen is in a stable state and does not disintegrate, while the fragment specimen needs to be mixed with 1.5% GG to maintain basic stability. This further indicates that the RSSC test results are relatively more conservative than those of the CSC test, which is consistent with previous studies [32].

3.2.2. Fractal Dimension

Fractal theory is mainly used to explain self-similarity, and its computational methods mainly include the box dimension method, the yardstick method, and the self-affine fractal method [33]. Since the particle distribution of collapsible loess after disintegration in water has statistical self-similarity, the fractal characteristics of collapsible loess disintegration shape can be described by the box dimension method based on the fractal theory as shown in Equations (1) and (2) [34]:

$${\mathrm{N}}_{\mathsf{ϵ}}\left(\mathrm{A}\right)\approx \mathrm{C}{\epsilon }^{-D}$$

(1)

$$D=\underset{\epsilon \to 0}{\lim }{\displaystyle \frac{\log \left({\mathrm{N}}_{\mathsf{ϵ}}\right(\mathrm{A}\left)\right)}{\log (1/\epsilon )}}$$

(2)

where A is a non-empty bounded subset of R_n, $\epsilon$ is the length of the box side ($\epsilon$ > 0), C is a constant, and D is the fractal dimension.

The images for determining the fractal dimension were taken from the photographs of the disintegrated specimens and after optimizing the photographs. The fractal dimension of the disintegrated shape of the soil specimens was determined using ImageJ. The fractal dimensions of different specimen disintegration shapes are shown in Figure 11. The fractal dimension is positively correlated with the irregularity and complexity of the disintegration shape of the soil specimens [34]. As can be seen in Figure 10, the fractal dimension of collapsible loess after disintegration shows a decreasing trend with the prolongation of curing time and the increase in GG dosage. The reduction in the fractal dimension further indicates that GG can effectively promote the agglomeration of soil particles, limit the movement of soil particles, inhibit the disintegration of collapsible loess, and positively strengthen the water stability of soil.

3.3. Collapsibility and Consolidation Properties

Due to the poor water stability, easy collapsibility, and other undesirable characteristics of collapsible loess, it is easy to cause adverse effects on the safety of engineering structures when water is encountered. In order to analyze the effects of different GG dosages on the collapsibility characteristics of collapsible loess, indoor collapsibility tests were conducted and the results are shown in Figure 12. As can be seen from Figure 12, the collapsibility coefficients under different pressures exhibit the same variation trend. With the increase in curing time and GG dosage, the collapsibility coefficient first decreases rapidly and then tends to be stable. The collapsibility of collapsible loess changes from slight collapsibility to non-collapsibility when the GG dosage reaches 0.75% under 50 kPa pressure. Meanwhile, with the increasing upper pressure, the collapsibility coefficient of collapsible loess decreases significantly. When the upper pressure reaches 100 kPa and above, the collapsibility coefficient of GG-modified loess is reduced to below 0.015, which means that it is considered to completely lose collapsibility. It can be seen that the addition of GG can effectively improve the collapsibility of collapsible loess. This is due to the fact that GG can form a thin coating on the surface of soil particles through hydrogen bonds, which expands the contact area between soil particles and enhances the bonding and agglomeration ability between soil particles [24]. The soil particles form denser agglomerations under the cementation effect of GG, which reduces the permeability of soil and thus eliminates the collapsibility properties of collapsible loess.

Compressibility is one of the important mechanical property indexes in foundation, road, and other backfill engineering projects. In practical engineering, a_1–2 and E_s are generally used to evaluate the compressibility of soil. The a_1–2 is the compression coefficient when the pressure is increased from 100 kPa to 200 kPa, and E_s is the corresponding compression modulus. As can be seen in Figure 13, the compression coefficient of collapsible loess decreases significantly under the action of GG with lower dosages (less than 1%). On the contrary, when the GG dosage exceeds 1.25%, the compression coefficient increases slightly. This indicates that the GG dosage is not as high as possible, but that there is an optimal dosage (1.25%). This is because with the addition of GG, GG reacts with water to form hydrogel, which fills the pores between soil particles and enhances the cementation between soil particles, thus significantly reducing the compressibility of GG-modified loess. However, excessive GG absorbs more water, thus reducing the interaction between gel and soil particles. On the other hand, excess GG acts as a lubricant in the soil particles and reduces the friction between the soil particles, thus increasing the compressibility of the soil [9]. If too much GG is used in the practical process, such as road or foundation treatment, on the one hand, the improvement effect of GG will be weakened, and on the other hand, the corresponding economic cost will be increased. Therefore, understanding the optimal GG dosage is critical for engineering projects involving collapsible loess.

3.4. pH

Figure 14 shows the pH of the specimens with different GG dosages and curing times. According to the pH test data, the pH of GG-modified loess is between 8.6 and 9, which slightly lower than that of natural collapsible loess. The pH of GG-modified loess shows a decreasing trend with the increase in GG dosage and curing time. This trend can be attributed to the interaction of GG molecules with OH^- ions to form hydrogen bonds in alkaline environments, which leads to a decrease in the pH of GG-modified loess [31]. At the same time, the reduction of OH^- ions can further reduce the zeta potential of soil particles, resulting in a corresponding reduction in the number of negative charges adsorbed on the surface of the soil particles [31]. According to the charge balance theory and the double electric layer theory, it is known that the number of cations adsorbed on the surface of soil particles is reduced accordingly, which in turn leads to the thickness of the double electric layer and the thickness of the water film on the surface of soil particles becoming thinner [35]. This process can improve the mechanical properties of collapsible loess by increasing the cohesion and friction between soil particles.

3.5. Microstructure and Mineral Analysis

The microscopic mechanism of GG-modified loess was further investigated using a HITACHI0-S-4800 scanning electron microscope, and the results are shown in Figure 15. As can be seen from Figure 15, the arrangement of soil particles in plain soil is relatively loose. The point-to-point contact and point-to-surface contact are dominant between soil particles, resulting in more small and medium pores, which weakens the connection between soil particles. With the incorporation of GG, GG fills the pores between soil particles and effectively reduces the porosity of the soil, and the contact mode is surface-to-surface contact. At the same time, it can be clearly observed from the figure that the network structures penetrating each other are formed between soil particles and GG, which promotes the agglomeration of individual soil particles into a dense agglomerate and increases the cementation ability of soil particles. The main reason for the above phenomenon is that GG forms a hydrogel with high cohesion when it encounters water. The hydrogel adheres to the surface of soil particles and enhances the cementation ability of soil particles by utilizing its own cohesion. With the increase in GG, the coverage of the hydrogel is wider, and a ‘bridge’ is built between non-adjacent soil particles, so that the non-adjacent soil particles are connected together and the integrity of the soil is improved. In conclusion, the addition of GG largely changes the microstructure of the soil samples, which shows that the strength, disintegration resistance, and water stability of the soil are improved macroscopically.

The SEM images of the plain soil and GG-modified loess were selected as typical representative areas 1, 2 for EDS analysis, as shown in Figure 16. The results of elemental determinations of the specimens are shown in

Table 2. Results of elemental determination in area 1.

Type of Element	C	O	Na	Mg	Al	Si	S	K	Ca	Fe
Mass ratio (%)	9.27	40.41	0.34	1.75	6.22	22.96	3.07	1.49	12.32	2.17
Atomic ratio (%)	15.71	51.41	0.30	1.46	4.69	16.64	1.95	0.78	6.26	0.79

and

Table 3. Results of elemental determination in area 2.

Type of Element	C	O	Na	Mg	Al	Si	S	K	Ca	Fe
Mass ratio (%)	12.28	39.57	0.29	1.63	6.46	22.77	1.57	2.34	9.99	3.08
Atomic ratio (%)	20.29	49.08	0.25	1.33	4.75	16.09	0.97	1.19	4.95	1.09

. It can be seen from Table 2 and Table 3 that C, O, Si, Ca, and Al elements are the main elements in the soil. Through comparative analysis, it was found that after the incorporation of GG, the proportion of C element in the soil samples increased by 3.01%, the proportion of Ca element decreased by 2.33%, and the other major elements did not change significantly.

XRD analysis was conducted on the specimens of plain soil and GG-modified loess, and the results are shown in Figure 17. As can be seen from Figure 17, the main components of each specimen are mainly Gypsum, SiO₂, CaCO₃, and C-A-H. By analyzing and comparing the positions and peaks of each wave peak in the graphs, it can be found that the incorporation of GG does not change the map morphology of the soil. The mineral composition of loess does not change significantly due to the incorporation of GG, which indicates that the change in soil microstructure is not caused by the production of new substances.

4. Discussion

4.1. Improvement Mechanism of Mechanical Properties

In this paper, based on the chemical changes and microstructure differences in collapsible loess before and after improvement, and combined with macroscopic manifestations such as strength changes and hydrophysical properties of the improved collapsible loess, the reinforcement mechanism of GG-modified loess is further investigated, as shown in Figure 18. The SS, UCS, and TS of the modified collapsible loess are significantly higher than those of the natural collapsible loess, which can be attributed to the following reasons: 1. GG reacts with water to form a hydrogel, which is attached to the surface of soil particles by hydrogen bonds and van der Waals forces, resulting in the formation of a denser GG–loess matrix. This process enhances the cohesion and friction between soil particles [12]. 2. The addition of GG can not only fill the pores between soil particles and change the size and number of pores, but also increase the compactness of the soil, thus enhancing the strength of the soil. At the same time, GG has a high specific surface area, which can increase the contact area between soil particles and thus improve the internal friction between soil particles [13]. 3. Due to the high adsorption capacity of GG for Ca²⁺, Mg²⁺, and other metal ions, GG can cross-link with Ca²⁺ in collapsible loess to form a mutually penetrating polymer network [24]. In this process, individual fine soil particles form larger agglomerates through cementation, which enhances the cohesion between soil particles, thus improving the cementation strength of the modified collapsible loess.

4.2. Improvement Mechanism of Hydrophysical Properties

The improvement mechanism of GG on the hydrophysical properties of collapsible loess is actually a complex process of multiple effects collaborating and jointly promoting, and the specific reinforcement mechanism can be summarized as the following two main processes: the first process is to stop the flow of water inside the soil. The addition of GG fills the pores between the soil particles, effectively reduces the size and number of pores, inhibits the flow rate of water inside the soil, and then improves the water stability of collapsible loess. The second process is to prevent the infiltration of water from the outside of the soil. GG reacts with water to form hydrogels. The hydrogels form a thin coating on the surface of soil particles under the action of hydrogen bonds and van der Waals forces [13]. To a certain extent, the coating can prevent the infiltration of water and reduce the direct contact between water and collapsible loess particles, thus playing a certain protective role and helping to improve the water stability of collapsible loess.

4.3. Practicality and Feasibility of GG-Modified Loess

The disintegration characteristics and collapsibility of loess have a crucial influence and effect on soil erosion, slope damage, foundation failure, the formation of gullies and caves, and the occurrence and evolution process of landslides, debris flow, and other geological disasters. At present, the commonly used traditional reinforcement materials include lime, cement and fly ash. These materials produce a large amount of CO₂ in the process of production and use, which seriously pollutes the environment and contradicts the green concept of environmental protection. At the same time, the subsequent treatment of improving soil with lime and cement is relatively difficult. Compared with lime and cement, the use of GG to improve soil can avoid these problems well. GG can improve the anti-disintegration of collapsible loess and eliminate its collapsibility, so as to prevent the occurrence of the above geological disasters. At the same time, GG is an environmentally friendly biopolymer that does not emit greenhouse gases during the modification process and avoids adverse effects on the environment. Meanwhile, GG in soil is degraded with the passage of time under the action of microorganisms [14]. This indicates that GG has the advantages of sustainable regeneration, low carbon emissions, and environmental friendliness, which makes it a good renewable soil modification material. However, due to the relatively limited application of GG in geotechnical engineering and low production, the improvement cost of GG is slightly higher than that of traditional modification materials such as cement and lime. Fortunately, with continuous and in-depth research on biopolymers, the application market of GG continues to expand, and the economic feasibility of GG improvement of collapsible loess is gradually increasing.

5. Conclusions

In this paper, through a series of mechanical tests and water stability tests, the strengthening behavior of different dosages of GG on the mechanical properties and hydrophysical properties of collapsible loess were investigated, the improvement mechanism of GG on collapsible loess was analyzed from the microscopic point of view, and the following conclusions were drawn:

1. Compared with natural collapsible loess, the shear strength, compressive strength, and tensile strength of GG-modified loess are significantly increased and show a trend of increasing and then decreasing with the increase in GG dosage; the optimal dosage of GG is 1.25%.

2. The addition of GG makes the disintegration material of collapsible loess change from loose fine particles to clear large particles or agglomerates. With the prolongation of the curing time and the increase in GG dosage, the disintegration area and fractal dimension of the specimens gradually decreased. This further indicates that GG plays a positive role in improving the disintegration resistance and water stability of collapsible loess.

3. The addition of a lower dosage of GG (less than 1.25%) not only reduces the compressibility of collapsible loess, but also eliminates the collapsibility of collapsible loess. However, the improvement effect is slightly decreased by further increasing the GG dosage.

4. Microstructural and mineral analyses show that the addition of GG does not change the mineral composition of collapsible loess. The addition of CG only forms a coating on the surface of soil particles through hydrogen bonds and van der Waals forces, which enhances the cementation ability of soil particles and makes individual soil particles cement to form denser agglomerates, thus improving the strength and water stability of collapsible loess.