Effect of EcSS 3000 on Expansive Clays

Effect of EcSS 3000 on Expansive Clays R. Malek Director, Particle Characterization Laboratory, Materials Research Institute, Pennsylvania State University, University Park, PA 16802. RQM@PSU.EDU (814) 865-7341 November 9, 2004

Table of Content Abstract... 3 Background information... 4 Introduction... 4 Methods... 5 a) X-Ray Diffraction... 5 b) Thermal Analysis... 5 c) Magic Angle Nuclear Magnetic Resonance Spectroscopy... 5 d) Zeta potential:... 6 e) Particle size analysis... 6 f) Uniaxial compression:... 6 Results and Discussion... 6 Structure of expansive clay... 6 a) X-Ray Diffraction... 7 b) Thermal analysis... 7 c) Particle size/size distribution... 9 d) 29 Si Magic angle NMR... 14 e) Zeta potential... 14 f) Unidirectional compaction... 14 Conclusions... 15 2

Abstract Two soil samples were provided by Environmental Soil Stabilization LLC (ESSL), Arlington Texas, which were described as an EcSS 3000 treated soil (CB-7A) and an untreated soil (CB-7B). These two undisturbed samples were collected on September 10, 2004 from an ESSL project site injected with EcSS 3000 electrochemical in August, 2004. The samples were examined by X-ray diffraction, thermal analysis, particle size distribution, magic angle NMR, zeta potential and unidirectional compaction. The X-ray diffraction of the treated and untreated samples indicates the presence of high proportion of swelling clay, montmorillonite with a basal spacing of 14A. However, it has lower intensity in the treated sample. Thermal analysis indicated that the interlayer water which causes swelling is tending to become free water. The particle size distribution of the treated sample indicates a shift towards smaller size particles. The 29 Si Magic angle NMR spectra indicate a trend towards a change from high Al occupancy to low Al occupancy silicate sheets as a result of the treatment with EcSS 3000 solution, resulting in structural changes in the expansive clay, namely, decrease in the negative charge, and destabilizing the silicate structure. And the zeta potential is reduced from -20.16 mv in the untreated sample to -14.49 mv in the treated one. The reduction in zeta potential indicates a much lower cation content in the interlayer spaces, and accordingly less water of hydration and less swelling. The results of the unidirectional compression tests of the treated and untreated samples show an increase in strength of the sample as a result of treatment. It is concluded that: EcSS 3000 increases ionic strength in the local area of injection, which results in: - Drawing water from other areas through osmosis. However, this water remains free (mobile), i.e. not migrating into the interlamellar spaces. - Depressing the thickness of the double layer providing less water for exchange with the interlamellar spaces in the clay. The soluble species in EcSS 3000 adsorb on the surface of the clay generating an electrosteric effect. In addition, a hydrophobic environment is created at the particles surfaces which tends to repel water. There is a trend towards a change from high Al occupancy to low Al occupancy in the silicate sheets, resulting in structural changes in the expansive clay: - decrease in the negative charge (evidence: zeta potential), and - destabilizing the silicate structure. 3

Background information Two soil samples were provided by Environmental Soil Stabilization LLC (ESSL), Arlington Texas, which were described as an EcSS 3000 treated brown silty clay soil (CB-7A) and an untreated brown silty clay soil (CB-7B). These two samples were collected at a depth of 2 to 4 feet below the ground surface on September 10, 2004 from an ESSL project site injected with EcSS 3000 electrochemical in August, 2004. The purpose of this report is to discuss the results of the tests carried out at Penn State to compare the properties of the two samples and to attempt to reveal the mechanism of action of the EcSS 3000 soil stabilizer on expansive clays. Introduction Expansive soils are known as Vertisols which are clayey soils that shrink when dry and swell when wet. Montmorrillonite is the major component of Vertisols. Figure 1 shows a map for the distribution of Vertisols in the United States. It is seen that Vertisoles are highly concentrated in Texas with the highest concentration in the area covering Dallas- Fort Worth, Waco, Austin and San Antonio. Methods Figure 1: Distribution of Vertisols in the United States. 4

Methods a) X-Ray Diffraction Finely ground powder samples were placed in an engraved, zero background sample holder. The surface of the powder was leveled using glass slide. In order to avoid preferred orientation of particles, which may produce inaccurate intensity values, a filter paper was placed on the surface of the powder and slightly pressed. Experience has shown that the texture of the filter paper prevents preferred orientation of crystals. The sample holder was then placed in a Scintag X-ray diffractometer and analyzed between angles 2 o to 70 o at a rate of 2 o per minute. Identification of clay structure by X-ray diffraction is based on the fact that each solid substance has its own characteristic atomic structure which diffracts X-rays in a characteristic pattern. The recognition of the pattern establishes uniquely the diffracting substance. The X-ray reflection takes place from the lattice planes according to Bragg s law: where λ = 2d sin θ... (1) λ is the wave length. θ is the incident angle, and d is the distance between lattice planes. b) Thermal Analysis A small portion of the as received soil samples was placed in the thermal analyzer and heated at a rate of 10 o C/min up to 1000 o C. The thermal analysis technique is used to characterize materials based on their behavior under various heat conditions. It includes Differential Scanning Calorimetry (DSC) which is used to measure the heat flows associated with physical and chemical changes during heating and Thermogravimetric Analysis (TGA) which is used to measure changes in weight of a sample with increasing temperature. c) Magic Angle Nuclear Magnetic Resonance Spectroscopy The solid state 29 Si Magic Angle Nuclear Magnetic Resonance Spectra (MAS NMR), referenced to tetramethylsilane (TMS) were determined as a single pulse NMR. The signals due to 1 H have been decoupled so that the signals due to Si-O-Si and Si-O-Al connectivities are not affected by the proton signals. The MAS NMR is based on splitting the nuclear spin in a magnetic field, followed by excitation by radio frequency. The energy difference between the ground and excited states is dependant on the shielding effect of electrons in the local bonding environment. The peaks of the resulting spectrum are assigned to different silicon/aluminum connectivity as follows: Q 0, Q 1, Q 2, Q 3, and Q 4 correspond to silica tetrahedron connected 5

to 0, 1, 2, 3, and 4 Si tetrahedra. Q n (0Al), Q n (1Al), Q n (2Al), Q n (3Al), Q n (4Al), correspond to silicon tetrahedra connected to 0, 1, 2, 3, and 4 Al tetrahedra. d) Zeta potential: Clays consist mainly of plate-like particles, which when in contact with water, usually have negatively charged faces. The physical properties of clay-water systems such as swelling are extremely sensitive to the nature of the electric double layer around the particles. Zeta potential measurements provide particularly relevant information regarding swelling. Zeta potential was measured using the microelectrophoretic mobility technique where the speed and direction of moving particles, under the influence of electric field, were measured and used to calculate the surface charge. e) Particle size analysis Particle size analysis in the nano size range was made by using Photon Correlation Spectroscopy (PCS) which relies on measuring the Brownian motion of small particles and relating this to the hydrodynamic diameter of the particle. Malvern Nanosizer S (0.6 nm to 6 mm) was used. f) Uniaxial compression: A soil consolidation apparatus was used to measure the strain percent of soil as a function of applied load. A soil specimen was taken from the larger sample and the upper and lower faces were trimmed flat. The weight of the specimen was determined and placed in the compression cell and a continuously increasing load is applied. The corresponding change in sample height, after equilibrium, was measured (ASTM D2435). Results and Discussion Structure of expansive clay In order to understand the mechanism of action of EcSS 3000 in stabilizing the clay, it is important to start with understanding the structure of expansive clay. Clays contain tetrahedral silicate sheets in which each Si tetrahedron is linked with three neighboring tetrahedra by sharing corners. The fourth tetrahedron corner points to a direction normal to the sheet and forms a part of an octahedral sheets (Al +3 ) in which octahedra are linked laterally, sharing their edges. Expansive clay has the montmorillonite structures (2:1 layer structure) in which tertrahedral silicate sheets are bound at both sides of the octahedral sheet. Isomorphous substitution of Al for Si in the silicate sheets and Ca and Mg for Al in the gibbsite sheets generates a net negative charge on the layers which is compensated by interlayer cations. Water molecules migrate to hydrate cations causing expansion (Fig 2). 6

Isomorphous Substitution Interlayer space Silicate sheet Gibbsite (Al) sheet Silicate sheet Interlayer space a) X-Ray Diffraction Figure 2: Structure of expansive clay The X-ray diffraction patterns of the treated and untreated samples are shown in Figure 3 together with that of standard clay (Wyoming montmorillonite) for comparison. From the X-ray results it is evident that the soil samples contain a high proportion of swelling clay, montmorillonite with a basal spacing of 14A. In addition quartz and calcite were identified. Furthermore Montmorillonite and calcite have lower intensities in the EcSS 3000 treated clay. b) Thermal analysis Results of the thermal analyses are shown in figures 4 and 5. Figure 4 represent a combination of Thermogravimetric analysis (TGA) and Differential Scanning Calorimetry (DSC). The former measures the weight loss as a function of temperature, whereas the later measures the heat evolved or absorbed as a result of chemical changes that occur in the sample when heated. It is evident that the untreated sample shows the following: An initial weight loss which is associated with an endothermic peak (heat absorption) up to 200 o C. This weight loss can be attributed to the free water (interparticle water). The shoulder in DSC in the range 200 o C-400 o C, indicates that water is lost from the interlayer spaces. 7

X-ray Q C M = Montmorillonite 14A I = Ilite C = Calcite Q = Quartz M M Q C C M Q C Q C Untreated C C EcSS 3000 Treated I I Wyoming Montmorillonite Figure 3: Results of X-ray diffraction. The weight loss in the range 400 o C-500 o C indicates that structural water is lost. The final weight loss region occurs in the range 700 o C-900 o C which is attributed to the decomposition of calcium carbonate. On the other hand, EcSS 3000 treated sample shows the following: An initial weight loss which is associated with an endothermic peak (heat absorption) up to 200 o C, attributed to the free water (interparticle water). This is similar to the untreated sample but with a larger magnitude indicating higher free water content. This may be due to the osmotic effects of the EcSS 3000 solution. The shoulder in DSC in the range 200 o C-400 o C does not exist indicating that water is tending to become free water. The weight loss in the range 400 o C-500 o C indicates that structural water is lost. 8

Thermal Analysis (TGA-DSC) 100 Treated Untreated 0.5 0.0 90 Interlayer water Dhydroxylation of structural water -0.5 Weight (%) 80 81.46% 79.31% 12.56% Carbonate (12.93mg) -1.0 Heat Flow (W/g) -1.5 70 Free water 2.301% Carbonate (1.701mg) -2.0 60-2.5 0 200 400 600 800 1000 Exo Up Temperature ( C) Universal V4.1D TA Instruments Figure 4: TGA/DSC results of the treated and untreated samples. The final weight loss region occurs in the range 700 o C-900 o C which is attributed to the decomposition of calcium carbonate. The treated sample shows a much less weight loss in this range indicating that a much less calcium carbonate is present. c) Particle size/size distribution The particle size distributions of the untreated and treated samples are shown in figures 6 and 7 for the untreated, treated samples, respectively. Figure 8 shows the two distributions for comparison. The Particle size distribution of the untreated sample (Fig. 6) shows a bimodal distribution with two peaks at 700 nm and 2000 nm. The average particle size is1030 nm. The EcSS 3000 treated sample, on the other hand, shows a multimodal distribution where a reduction in the two large size peaks is evident and a new small size peak at ~ 100 nm appears. The average particle size is 490 nm. 9

Differential Scanning Calorimetry 0.5 Treated Untreated 0.0 Heat Flow (W/g) -0.5-1.0-1.5 Interlayer water -2.0 Free water -2.5 0 100 200 300 400 Exo Up Temperature ( C) Universal V4.1D TA Instruments Figure 5: Differential Scanning Calorimetry (DSC) results of the treated and untreated samples. Size Distribution by Intensity 15. Intensity (%) 10. 5. 0 10. 100. 1000. 10000. Diameter (nm) Record 320: CB-7B' Figure 6: Particle size distribution of the untreated sample. 10

Size Distribution by Intensity 8. Intensity (%) 6. 4. 2. 0 10. 100. 1000. 10000. Diameter (nm) Record 321: CB-7A' Figure 7: Particle size distribution of the EcSS 3000 treated sample. Size Distribution by Intensity Intensity (%) 15. 10. 5. Untreated Treated 0 10. 100. 1000. 10000. Diameter (nm) Record 320: CB-7B' Record 321: CB-7A' Figure 8: Particle size distribution of the EcSS 3000 treated and untreated samples. 11

Magic Angle Nuclear Magnetic Resonance Spectroscopy Untreated = Spinning Side Bands Magic Angle Nuclear Magnetic Resonance Spectroscopy Q4(0Al) Treated Q3(0Al) Q3(1Al) Q4(1Al) Figure 9: 29Si spectrum of the Untreated sample: Bottom without SS bands. 12

Magic Angle Nuclear Magnetic Resonance Spectroscopy Treated = Spinning Side Bands Magic Angle Nuclear Magnetic Resonance Spectroscopy Q4(0Al) Treated Q3(0Al) Q3(1Al) Q4(1Al) 13

Figure 10: 29 Si spectrum of the Treated sample: Bottom without SS bands. d) 29 Si Magic angle NMR The 29 Si spectrum of the untreated sample (Figure 9) shows the following: Peak at -109 ppm, which is assigned to Q4(0Al) in quartz. Peak at -101 ppm, which is assigned to 4(1Al) in Si-O-Al, the linkage between Si sheets and Al sheets. Peak at -94.5 ppm is assigned to Q3(0Al) in silicate sheets. The Peak integrated area is 7.5. Peak at -88 ppm is assigned to Q3(1Al) in silicate sheets containing Al substitutes. The 29 Si spectrum of the EcSS 3000 treated sample (Figure 10) shows the following: Peak at -109 ppm, which is assigned to Q4(0Al) in quartz. Peak at -101 ppm, which is assigned to 4(1Al) in Si-O-Al, the linkage between Si sheets and Al sheets. Peak at -94.5 ppm is assigned to Q3(0Al) in silicate sheets. The Peak integrated area is 85. Peak at -88 ppm is assigned to Q3(1Al) in silicate sheets containing Al. The increase of the Q3(0Al) peak integrated area from 7.5 to 8.5 as a result of treatment with EcSS 3000 may indicate a trend towards a change from high Al occupancy to low Al occupancy silicate sheets, resulting in structural changes in the expansive clay, namely, decrease in the negative charge, and destabilizing the silicate structure. e) Zeta potential The clay particles in the untreated sample were found to carry a negative charge and their zeta potential was found to be - 20.16 mv. On the other hand, the EcSS 3000 treated sample was found also to possess a negative charge and the zeta potential is appreciably reduced to - 14.49 mv. The reduction in zeta potential indicates a much less cation in the interlayer spaces, and accordingly less water of hydration and less swelling. f) Unidirectional compression The results of the unidirectional compression tests of the treated and untreated samples are shown in figure 11. It is clear that the two curves are parallel. But the treated sample shows a much less strain percent indicating increase in toughness of the sample as a result of treatment. 14

Unidirectional compression test Strain % 35 30 25 20 15 10 5 Treated Untreated 0 1 10 100 1000 10000 100000 STRESS (kpa) Figure 11: Stress-strain curves of the treated and untreated samples. Conclusions EcSS 3000 increases ionic strength in the local area of injection, which results in: - Drawing water from other area through osmosis. However, this water remains free (mobile), i.e. not migrating into the interlamellar spaces.. - Depressing the thickness of the double layer providing less water for exchange with the interlamellar spaces in the clay. The soluble species in EcSS 3000 adsorb on the surface of the clay generating an electrosteric effect. In addition, a hydrophobic environment is created at the particles surfaces which tends to repel water. There is a trend towards a change from high Al occupancy to low Al occupancy in the silicate sheets, resulting in structural changes in the expansive clay: - decrease in the negative charge (evidence: zeta potential), and - destabilizing the silicate structure. Acknowledgement Thanks to Dr. D. Fritton, professor of soil physics, Penn State, for invaluable discussion on soil compression and properties of vertisols. 15