IntroductionUndrained shear strength (su) is an important geotechnical engineering design parameter for clayey soils in undrained conditions (e.g., submarine, offshore, glacial, and deep sediments), which is generally assessed by high shear rates and poor drainage conditions (Gutierrez et al. 2008; Kayabali and Tufenkci 2010). Despite several factors affecting the su of clays, the net derjaguin-landau-verwey-overbeek (DLVO) forces between clay particles (Warkentin and Yong 1962) and pore (or double-layer) fluid chemistry (Sridharan and Prakash 1999) are important factors affecting the su of clays.Factors influencing the net DLVO forces among clays include clay mineralogy, plasticity, fabrics, pore–fluid chemistry, water content, stress history, anisotropy, and organics (Mayne 1985; Hyde and Ward 1986; Dolinar and Trauner 2007; Kayabali and Tufenkci 2010; Spagnoli et al. 2010; Hong et al. 2013). The net DLVO forces change the particle associations, thereby affecting the su of clays; the edge-to-face associations show higher su compared to other types of associations (i.e., edge-to-edge and face-to-face) because the Coulombic attraction between the oppositely charged edge and the face surface is significantly stronger than their van der Waals attraction (Dolinar and Trauner 2007). Furthermore, the collapse and reorientation of the edge-to-face contact result in a denser and more stable local geometry (Sachan and Penumadu 2007).Pore–fluid chemistry can alter particle arrangements (e.g., flocculation and deflocculation); thus, it plays an important role in the su of clays. For example, salinity conditions affect the shear strength of soils, which experiences frequent changes in salinity due to freshwater or saltwater mixing (Schlue et al. 2010). Furthermore, the su of clays increases as the particle grain size decreases (i.e., the specific surface area increases) due to the larger water absorption capacity (Trauner et al. 2005).Microbial biopolymers are polymeric biomolecules that are produced through the metabolism of microorganisms. Biopolymers contain side chains (e.g., trisaccharide) that can interact with charged clay particles, counterions, and pore water through hydrogen bonding (Chenu and Guérif 1991; Chang and Cho 2019; Chang et al. 2019). Thus, biopolymers have been reported as effective and ecofriendly soil modification materials owing to their ability to clog pore spaces (Khachatoorian et al. 2003; Bouazza et al. 2009; Chang et al. 2016a; Lee et al. 2019; Zhou et al. 2020) and enhance interparticle bonding between soil particles (Malik and Letey 1991; Kwon et al. 2017; Latifi et al. 2017; Chang et al. 2020; Soldo et al. 2020; Sujatha et al. 2020). Moreover, biopolymers have been shown to form viscous hydrogels and interact with the charged surfaces of clays (Chang et al. 2015, 2020). For instance, Sujatha et al. (2020) showed that xanthan gum (XG) biopolymer forms an interconnected network between the biopolymer molecules and soil surface, which results in a decrease in permeability and increase in strength. The interconnected network also resists compaction, resulting in a lower maximum dry density and higher optimal moisture content (Dehghan et al. 2019).In addition, biopolymer-based soil treatments have been shown to significantly enhance the hydraulic erosion resistance of soils (Ham et al. 2018; Kwon et al. 2020), which is expected to effectively mitigate erosion or scouring problems of waterfront geotechnical engineering structures. Additional studies showed that biopolymer-induced soil flocculation accelerates the settling velocity and reduces the turbidity of soil suspensions (Kwon et al. 2017; Kang and McLaughlin 2020), which is useful for sludge dewatering (Guo and Wen 2020). Meanwhile, the redistribution of XG (i.e., from strong intergranular bridges to a uniform distribution around the grains) occurs during the wetting–drying cycles, which gradually reduces the overall shear strengthening effect of the biopolymer (Chen et al. 2019a). Singh and Das (2020) showed that biopolymer treatment reduces mass loss during repeated freeze–thaw cycles.Biopolymer interactions with pore fluids and clay particles can improve the shear strength of soils, accompanied by an increase in the friction angle or cohesion (Khatami and O’Kelly 2012; Chang and Cho 2019; Chen et al. 2019b) via particle aggregation and the following conglomeration effect (Chang and Cho 2019); this would improve the erosion resistance of fine-grained soils (Ham et al. 2018; Kwon et al. 2020). However, the effect of the pore–fluid chemistry on the su of biopolymer-treated clays has not yet been investigated.In this study, we explored the effect of the XG content, clay mineralogy, and pore–fluid type (deionized [DI] water, 2-M-NaCl brine, and kerosene) on the su of kaolinite and montmorillonite clays. The su of the XG-treated soils with different pore–fluid types and contents was evaluated through fall cone tests. The morphology of the clay particles in the presence of different pore fluids, ions, and XG was analyzed using scanning electron microscopy (SEM).Materials and MethodsCharacteristics of the Clay MaterialsKaolinite (1:1 type; Bintang, Indonesia) and montmorillonite (2:1 type, CAS: 1302-78-9; Sigma-Aldrich, Bellefonte, Pennsylvania) were used to examine and compare the effects of clay mineralogy on the su. Kaolinite has low swelling and reactive areas on its external surfaces because the basal oxygen atoms of the tetrahedral silica sheet and the hydroxyls of the octahedral alumina sheet are held together by hydrogen bonding (Giese 2003). Meanwhile, montmorillonite shows a high swelling potential upon water addition because its layers are loosely held together by van der Waals forces (Chen and Peng 2018).Although kaolinite presents a larger particle size than montmorillonite, the latter exhibits a higher specific surface area (SSA), cation exchange capacity (CEC), and a thicker double layer, as shown in Table 1. Kaolinite presents a lower aspect ratio and a higher positive edge charge portion (Tombácz and Szekeres 2006; Villanueva et al. 2010), whereas montmorillonite contains approximately 90%–95% of pH-insensitive negative charges contributing to the total charge (Au and Leong 2016). Thus, the attraction between the edge and face charges is more critical to the su of kaolinite than in montmorillonite (Penner and Lagaly 2001).Table 1. Index properties of the clays that were used in this studyTable 1. Index properties of the clays that were used in this studyType of clayD50 (μm)SSA (m2/g)aCEC (meq/100 g)bpHcXG (%)PL (%)LL (%)PL (%)LL (%)PL (%)LL (%)Kaolinite3226.14.2031703451—930.135743652930.532973755911408033618423679467192Montmorillonited0.0722164.16.50623894286—610.1543864696600.555373441006017137043106582693384111459Montmorillonite clays have more negative zeta potential (i.e., the potential difference between the media and clay layers) than kaolinite minerals because they have more SiO− groups on their surface, owing to their 2:1 structure (Doi et al. 2019). When salt (e.g., Na- or Cs-halide) is added, the negative zeta potential of kaolinite tends to be more negative owing to the formation of pH-dependent groups (Liu et al. 2012), whereas that of montmorillonite becomes neutralized by cation adsorption owing to the enhanced interlayer activity (Doi et al. 2019).Prior to the experiments, kaolinite was washed six times with DI water to remove the supernatants after solid and fluid separation (Wang and Siu 2006), whereas montmorillonite was used without washing. The clays were dried in an oven (110°C) according to ASTM D2216-19 (ASTM 2019). The index properties of the clays used are listed in Table 1.Characteristics of Xanthan GumXG is a biopolymer produced by Xanthomonas campestris; its structure consists of repeating glucose units and side chains with three sugar units (García-Ochoa et al. 2000). XG is water soluble, and once dissolved, its structure expands, and the side chains that consist of glucuronic and pyruvic acid groups become negatively charged (Hassler and Doherty 1990; Nugent et al. 2009; Petri 2015). Therefore, XG is widely used as a thickener and stabilizer for suspensions, solid particles, and foams in aqueous formulations (Katzbauer 1998; Chang et al. 2016b). XG has shown high potential for soil stabilization through the enhancement of its shear strength (Lee et al. 2017; Chang and Cho 2019) and compressive strength (Kwon et al. 2019) via clay–XG aggregation (Chang et al. 2019). Moreover, XG reduces the hydraulic conductivity of soil owing to pore clogging induced by XG hydrogels (Cabalar et al. 2017). A research-grade XG (CAS: 11138-66-2; Sigma-Aldrich, Saint Louis) was used in this study for the experiments.Preparation of the XG–Clay MixtureWashed and oven-dried clays were mixed with XG powder at mass ratios (mb/ms) of 0% (untreated), 0.1%, 0.5%, 1.0%, and 2.0%. Three chemically distinct pore fluids, namely DI water, 2-M-NaCl brine (high ionic concentration), and kerosene (low permittivity), were used, as in previous studies (Jang and Santamarina 2016; Cao et al. 2019). XG solutions were prepared by mixing these pore fluids with dry XG powder. Oven-dried clays and XG solutions were mixed until the formation of uniform XG–clay-pore fluid mixtures. We prepared 2-M-NaCl brine (i.e., two moles of NaCl per liter of DI water) by dissolving 58.44 g (1 mol) of NaCl (CAS No. 1310-58-3; Junsei Chemical Co., Tokyo) in 500 mL of DI water in a beaker using a magnetic stirrer to form 500 mL of a uniform 2-M-NaCl solution. During the experimental procedures, the fluid-to-dry soil ratio (w=mw/ms) was adjusted by adding the prepared pore fluids into the soils at gradual increments and thoroughly mixed using a laboratory spatula to obtain uniform soil samples. The liquid limit (LL) and plastic limit (PL) of the XG–clay mixtures in Table 1 were measured via laboratory fall cone tests (British Standards Institution 2017) and the thread rolling method (ASTM 2017), respectively.Experimental Details and Testing ProcedureThe su of clay was evaluated via laboratory fall cone tests (Wood 1990; Koumoto and Houlsby 2001; O’Kelly et al. 2018). The effect of the shearing rate on the su values obtained from the fall cone tests was not investigated in this study; however, differences in the shearing rates may have existed between the fall cone tests and could have nonnegligible influences on the obtained su values. Koumoto and Houlsby (2001) analyzed the strain effect on su by comparing laboratory triaxial test and fall cone test measurements; su was reported to have increased by 60% at high strain rates. Kulhawy and Mayne (1990) analyzed the triaxial compression data and observed a 10% increase in su per log cycle of the strain rate. In contrast, Chow et al. (2012) argued that the su measured with parallel plate rotational viscometer tests was smaller than that assessed using conventional triaxial tests owing to the viscous rate effect.In this study, XG–clay mixtures were tamped into a container with a 55-mm diameter and 40-mm height. Soils with a degree of saturation higher than 90% were used. After surface trimming, the wet weight was measured to obtain the density information (e.g., dry density, total density, and void ratio) for each specimen. Simultaneously, a cone penetration test was conducted to prevent alterations in the specimen moisture contents. The 80-g and 30°-apex cone could penetrate the soil pastes for 5 sec. The su of the clay obtained from the fall cone test (suFC) was calculated as follows: (1) where K = cone factor that reflects the friction effect between the cone and soil (Stone and Phan 1995); W = weight of the cone (80 g); and d = cone penetration depth (mm). A K value of 0.867 was adopted for the 30°-apex cone, which corresponds to suLL=1.7  kPa (suLL is the su when w=LL) (Sharma and Bora Padma 2003; Kayabali and Tufenkci 2010). The su values of the soils with liquidity index [LI=(w−PL)/(LL−PL)] values greater than 0.2 were measured to avoid potential overestimations (Vardanega and Haigh 2014).A laboratory vane shear test was performed simultaneously on untreated clays under the same conditions to obtain the vane shear strength (suV) of untreated clays for comparison and verification. This was achieved using a rectangular vane (width=12.7  mm, height=12.7  mm, and thickness = 0.05 mm) with a rotation velocity of 60°/min according to ASTM D4648/D4648M-16 (ASTM 2016). For each XG–clay–water content (w) condition, fall cone and vane shear tests were conducted with three different specimens to obtain reliable average values for suFC and suV. After the measurements, the XG–clay-pore fluid samples around the cone and vane intrusion were collected and dried in an oven at 110°C for 24 h to obtain the w following ASTM D2216-19 (ASTM 2019).The microstructures of the clays (untreated and XG-treated) were observed using SEM (S-4800 and SU-8230, Hitachi High Technologies, Tokyo). The clays were sampled at w=LL. The samples were air-dried at room temperature and attached to a 25-mm-diameter SEM mount with carbon conductive tabs (Pelco Tabs; Ted Pella, Redding, California). The specimens were coated with osmium (OsO4) for 10 s under vacuum using a plasma coater (OPC-60A). Conventional SEM was conducted to observe the morphology in terms of the effect of XG and pore fluids on the clays. However, advanced microscopy techniques, such as environmental-SEM (Carrier et al. 2013) or cryo-SEM (Du et al. 2009), are needed to analyze the XG–clay–salt water interactions in moist conditions.Correlations between su and Index PropertiesKoumoto and Houlsby (2001) suggested the following linear regression equation showing the relationship between log su and log w: (2) where α = water content when su is 1 kPa; and β = slope of log w and log su. In addition, α and β depend on the soil plasticity, compressibility, SSA, particle structure, pore–fluid chemistry, and organic matter content (Koumoto and Houlsby 2001; Trauner et al. 2005; Dolinar and Trauner 2007).From Eqs. (1) and (2), su and LI can be correlated as follows (Wroth and Wood 1978; Wood 1990): (3) where suLL = undrained shear strength at LL; and R = ratio between the su values at PL and LL. The R factor value depends on the clay mineral type and soil activity (Wood 1990). The correlation factors (α, β, and R) can be used to estimate the su of the clays, with w and the index properties (e.g., LL, LI) assessed in the laboratory.Results and AnalysesEffect of Pore–Fluid Chemistry on the su−w RelationshipThis study investigated the differences in clay behaviors between water-based (DI water and 2-M-NaCl brine) and kerosene-based mixtures. In clay, the relative permittivity of the pore fluids affects the double-layer forces and accompanying clay fabrics (Espinoza and Santamarina 2012); moreover, the double-layer thickness increases with the relative permittivity of the pore fluids (Mitchell and Soga 2005). However, the van der Waals attraction forces related to the Hamaker constant decrease with an increasing pore–fluid relative permittivity (Sridharan and Rao 1979).Fig. 1 shows the su−w relationship for the different clay types and pore fluids, and Fig. 2 presents the α and β variations of the XG-treated clays in terms of the pore–fluid relative permittivity variation based on the data from Fig. 1. The similar trends shown by suV and suFC for untreated clays with DI water [Figs. 1(a and d)] validate the fall cone tests conducted to assess the su of the clays.The su of kaolinite is affected by the net attractive forces between the clay particles, whereas that of montmorillonite is affected by the viscous resistance of the double layer (Sridharan and Prakash 1999). Among the different pore fluids, kaolinite shows the highest su [Figs. 1(a–c)] and highest α and β values [Figs. 2(a and b)] with kerosene, followed by DI water and brine, whereas montmorillonite shows the highest su with DI water, followed by brine and kerosene [Figs. 1(d–f) and 2(c and d)]. Generally, the attraction forces and particle fabrics govern the su of kaolinite (Warkentin and Yong 1962; Sridharan and Prakash 1999). Thus, kaolinite exhibited the highest su and shear parameters in kerosene, which has the lowest relative permittivity among the three pore fluids. This was due to the presence of edge charges and van der Waals forces, which induced interparticle interactions and consequently increased the su of kaolinite (Wang and Siu 2006). Interestingly, various studies have also observed that kaolinite shows an LL increase in nonpolar (low relative permittivity) pore fluids (Jang and Santamarina 2016; Chang et al. 2019).In contrast, the thick viscous double layers mainly govern the shear resistance of montmorillonite (Warkentin and Yong 1962; Sridharan and Prakash 1999) owing to the high aspect ratio and 2:1 structure of montmorillonite, resulting in low net attraction forces (Secor and Radke 1985; Santamarina et al. 2001; Tombácz and Szekeres 2006). Thus, montmorillonite could have higher su [Figs. 1(d–f)], α, and β [Figs. 2(c and d)] values in fluids with higher permittivity.Effect of Clay Mineralogy on the su−w RelationshipWhile considering the effect of clay mineral type on the su, montmorillonite [Fig. 1(d)] required a much higher w than kaolinite [Fig. 1(a)] to achieve the same su in DI water owing to the high SSA of montmorillonite; this is in accordance with the findings by Sridharan and Prakash (1999). Because the su generally decreases as w increases, montmorillonite shows a higher su than kaolinite at the same w.However, in 2-M-NaCl brine, the salt-induced double-layer suppression and particle rearrangement reduced the shear resistance of both clays (Sridharan and Prakash 1999). Kaolinite [Fig. 1(b)] and montmorillonite [Fig. 1(e)] showed a drastic su reduction compared to that in DI water [Figs. 1(a and d) and (2)], whereas montmorillonite had a higher su than kaolinite at a similar fluid (2-M NaCl) content. The double-layer compression caused by the cations appeared more prominent for montmorillonite because it presents a higher CEC and interlayer activity than in kaolinite (Doi et al. 2019).In the absence of double-layer formation with nonpolar kerosene, kaolinite with a larger edge portion (Wan and Tokunaga 2002) shows higher su values than montmorillonite [Figs. 1(c and f) and (2)]. In kerosene, the su of the clay is mainly governed by the net attraction forces and fabrics instead of the viscous resistance of the double layers (Jang and Santamarina 2016) because the double-layer thickness of the clay decreases in kerosene (Dukhin and Goetz 2010).Effect of XG on the su−w RelationshipThe XG treatment changes the shear resistance of the clays because the net negative charges on the XG branches can interact with the clay particles, salt ions, and water molecules (Chang et al. 2015; Turkoz et al. 2018). Water and salt adsorption by XG increases the su of the clays, whereas particle aggregation reduces the su owing to the decrease in clay surface area. Thus, kaolinite has more prominent variations in its shear resistance as compared to montmorillonite.For kaolinite in DI water [Fig. 1(a)], the su peaked at mb/ms=0.5% and decreased at 0.5%

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *