Skip to main content
NCBI home page
EPA Author Manuscripts logoLink to EPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Mar 16.
Published in final edited form as: Near Surf Geophys. 2017 Dec;15(6):615–624. doi: 10.3997/1873-0604.2017049

Specific polarizability of sand–clay mixtures with varying ethanol concentration

Sundeep Sharma 1,*, Lee Slater 1, Dimitrios Ntarlagiannis 1, Dale Werkema 2, Zoltan Szabo 3
PMCID: PMC8925411  NIHMSID: NIHMS1783117  PMID: 35299702

Abstract

We utilise a concept of specific polarizability (cs), represented as the ratio of mineral–fluid interface polarization per pore-normalised surface area Sp, to demonstrate the influence of clay-organic interaction on complex conductivity measurements. Complex conductivity measurements were performed on kaolinite– and illite–sand mixtures as a function of varying ethanol (EtOH) concentration (10% and 20% v/v). The specific surface area of each clay type and Ottawa sand was determined by nitrogen-gas-adsorption Brunauer–Emmett–Teller method. We also calculated the porosity and saturation of each mixture based on weight loss of dried samples. Debye decomposition, a phenomenological model, was applied to the complex conductivity data to determine normalised chargeability (mn). Specific polarizability estimates from previous complex conductivity measurements for bentonite–sand mixtures were compared with our dataset. The cs for all sand–clay mixtures decreased as the EtOH concentration increased from 0% to 10% to 20% v/v. We observe similar cs responses to EtOH concentration for all sand–clay mixtures. Analysis of variance with a level of significance α = 0.05 suggests that the suppression in cs responses with increasing EtOH concentration was statistically significant for all sand–clay mixtures. On the other hand, real conductivity showed only 10% to 20% v/v changes with increasing EtOH concentration. The cs estimates reflect the sensitivity of complex conductivity measurements to alteration in surface chemistry at available surface adsorption sites for different clay types, likely resulting from ion exchange at the clay surface and associated with kinetic reactions in the electrical double layer of the clay–water–EtOH media. Our results indicate a much larger influence of specific surface area and ethanol concentration on clay-driven polarization relative to changes in clay mineralogy.

INTRODUCTION

Ethanol is now widely used as an additive to conventional fuel to provide cleaner emissions during internal combustion in various automotive engines. Ethanol enters groundwater through accidental spills, and it has emerged as a leading contaminant in groundwater (Gomez and Alvarez 2010). Powers et al. (2001a) described the potential risks to groundwater due to accidental spills of large volumes of gasoline containing ethanol during transportation. Increased concentrations of ethanol in groundwater can result in slower degradation of the benzene, toluene, ethyl–benzene, and xylene (BTEX) compounds present in gasoline. In groundwater containing ethanol and BTEX compounds, ethanol is preferentially biodegraded by microbes over BTEX compounds (Corseuil et al. 1998; Powers et al. 2001b; Lovanh, Hunt and Alvarez 2002; Ruiz-Aguilar et al. 2002; Österreicher-Cunha et al. 2007; Schaefer et al. 2010), resulting in rapid depletion of dissolved oxygen and other essential electron acceptors required for BTEX biodegradation.

Determining the location of subsurface contaminant plumes in porous media is essential for designing successful active or passive bioremediation processes. Non-invasive geophysical techniques offer the potential to help delineate organic contaminant plumes in the subsurface and monitor remediation processes. Olhoeft (1985, 1986) and Sadowski (1988) first suggested the potential of low-frequency electrical methods for delineating regions associated with clay-organic interactions in the subsurface. In particular, Olhoeft (1985) carried out the first study on detecting organic contamination using complex conductivity (CC). In this early study, montmorillonite samples were saturated using pore fluid with and without toluene and the CC of the resultant interaction was measured. The resultant increase in phase (up to ~100 mrad) with increase in toluene concentration was interpreted as the enhancement in surface polarization due to adsorption of organic cations onto clay surfaces. This change in phase (ϕ) as a result of clay-organic reactions prompted further research into electrical signatures of organic contaminants (Olhoeft 1985; Sadowski 1988; Olhoeft and King 1991).

However, later efforts by Brown, Sorenson and Brown (2004) were unsuccessful in reproducing ϕ responses associated with clay–toluene interactions reported by Olhoeft (1985). Such inconsistent findings encouraged further research to determine the geo-electric responses associated with clay-organic interactions. Ustra et al. (2012) demonstrated time dependence of CC measurements for clayey soil samples containing toluene. They noted a smaller yet significant effect of toluene content on the phase response shortly after sample preparation, followed by no significant relation between CC measurements and toluene content at electrical equilibrium after 40 days. Contrary to Olhoeft (1985), Ustra et al. (2012) found that the presence of toluene shortly after sample preparation suppressed (rather than enhanced) the CC response.

Personna et al. (2013) performed a series of laboratory experiments on sand–clay mixtures with varying concentrations of ethanol. Again, contrary to earlier studies by Olhoeft (1985) on CC measurements of clay-organic interaction, Personna et al. (2013) observed a clear decrease in the measured phase and imaginary part of CC for a sand–bentonite matrix with increasing concentration of ethanol in the pore-filling fluid. This decreasing magnitude increased with increasing ethanol concentration in the ethanol–water mixtures. Personna et al. (2013) attributed this behaviour to a variety of possible scenarios at the mineral–fluid interface, such as cation–dipole interactions resulting in preferential and strong adsorption of EtOH relative to water on clay surfaces.

The purpose of this study is to better understand the effects of ethanol concentration and clay mineralogy on the CC signatures of different sand–clay mixtures. We use the specific polarizability (cp) concept to investigate the effects of EtOH on the polarization in different sand–clay mixtures. The concept is based on normalising the imaginary conductivity by the pore volume normalised surface area (Sp) to define a parameter that primarily describes the control of fluid chemistry and/or mineralogy on the CC measurements (Weller et al. 2010; Weller et al. 2011; Weller et al. 2015). Variations in the cp for different clay types were hypothesised to result from differences in mineralogy for different types of clays used in our mixtures and variation in the EtOH concentration in the pore-filling fluid. Our main objective was to determine the cp variation associated with the interaction of EtOH with different clay minerals and associated adsorption capacity. We explore cp variation resulting from changes in the mineralogy between kaolinite (1:1) and illite (2:1) and its subsequent interaction with EtOH at different concentrations. We compare the variations in cp from our samples with cp calculated for bentonite (2:1)–sand mixtures measured by Personna et al. (2013).

CLAY MINERALS

Under natural groundwater conditions, a clay mineral generally carries a fixed net negative charge on the surface due to changes in total net charge resulting from isomorphic substitution by elements of similar size and lower valence. This negative charge is fulfilled by adsorption of cations on the surface of the clay mineral (van Olphen 1977; Sposito et al. 1999). In 2:1 clay minerals such as montmorillonite, illite, or bentonite, there are two tetrahedral sheets on each side of an octahedral sheet (T-O-T) and water molecules can be readily adsorbed onto the exchangeable cations (Figure 1). In contrast, a 1:1 clay mineral like kaolinite is characterised by an alternating two-sheet layer structure composed of one silica tetrahedral sheet and one alumina octahedral sheet (Spagnoli et al. 2010). As a result of chemical bonding between the silica and alumina sheets, the inner hydroxyl plane in 1:1 clay is inaccessible for ion interactions. For such clay types, only the outer groups, located along the unshared plane, are available for interactions with ions and organic molecules (Miranda-Trevino and Coles 2003) (Figure 1).

Figure 1.

Figure 1

Open in a new tab

Structural patterns for 1:1 and 2:1 clays. The kaolinite structure is stacked by hydrogen bonding, resulting in no interlayer spaces between two sheets. In illite, the interlayer space is occupied by potassium cations. In bentonite (and other 2:1 clays), the interlayer spaces are occupied by water and common substitutes such as sodium, magnesium, and calcium cations (modified from Craig 1974).

Saidy et al. (2012, 2013) showed an increase in the adsorption capacity for dissolved organic compounds onto clay in the order kaolinite (1:1) < illite (2:1) < montmorillonite (2:1). In 2:1 clay minerals, EtOH molecules, similar to water molecules, can be adsorbed onto the inner surface. Dowdy and Mortland (1967) and German and Harding (1969) demonstrated the affinity of EtOH and other alcohols for these adsorption sites between clay sheets. EtOH has been shown to be capable of replacing water in 2:1 clays in the inner hydration shell around cations such as Ca2+, Cu2+, and Al3+. Parke and Birch (1999) and Atamas and Atamas (2009) describe how the competition of EtOH for adsorption on available cation sites in a clay structure leads to complex intermolecular interactions. These interactions at the mineral–fluid interface can extensively alter mobility of ions in solution (Bhat and Shetty 2011). Therefore, they also have the potential to change interfacial electrical properties measured with CC, as considered here. The electrical double-layer (EDL) model is commonly used to describe the electric charge distribution at the mineral–water interface. The EDL consists of (a) a Stern layer formed as a result of direct ion sorption on the mineral surface and (b) a diffuse layer that consists of mostly freely moving excess counter ions extending beyond the Stern layer (Revil et al. 2012).

Okay et al. (2014) studied the effect of clay content and clay mineralogy on the CC response in a well-controlled laboratory experiment with water-saturated unconsolidated clays and sand–clay mixtures. They analysed two classes of samples of kaolinite and bentonite. Different weight fractions of kaolinite and bentonite samples were mixed with Fontainebleau sand (99.98% silicapure sand) to prepare various mixtures. Okay et al. (2014) observe generally higher imaginary conductivity values for bentonite–sand samples as compared to kaolinite–sand samples. Okay et al. (2014) attribute this behaviour to higher cation exchange capacity (CEC) for bentonite (CECkaolinite= <0.10 meq g−1, CECbentonite = 0.44 meq g−1) as compared to kaolinite.

ELECTRICAL PROPERTIES

The CC method characterises the capacitive and conductive properties of a material over a range of frequencies (typically 0.001–1000 Hz). In geophysical applications, the measured electrical properties are conveniently represented by the conductivity magnitude (|σ|) and phase shift (ϕ) of the measured voltage waveform across a sample relative to the current waveform measured across a resistor placed in series

σ∗=σ′2+σ′′2, (1)
ϕ=tan−1σ′′σ′, (2)

where σ′ represents the real part and σ″ represents the imaginary part of the CC (σ*).

These measurements can be presented in terms of real (in-phase) and imaginary (out-of-phase) components of the CC σ* by

σ′=σ∗cosϕ, (3)
σ′′=σ∗sinϕ. (4)

In a porous medium free of metallic particles, electrical conduction can follow two paths: (a) electrolytic conduction through fluid-filled interconnected pore spaces and (b) surface conduction within the EDL. The electrolytic and surface conductivity values are generally assumed to add in parallel (Waxman and Smits 1968; Vinegar and Waxman 1984; Lesmes and Frye 2001). With this assumption, σ′ and σ″ are expressed in terms of fluid conductivity (σw) and surface conductivity (σ″surf) by

σ′=1Fσw+σsurf′, (5)
σ"=σsurf", (6)

where F represents the formation factor, being the ratio of the conductivity of the saturating fluid to the conductivity of the fluid-saturated porous media in the absence of surface conduction (Archie 1942). The imaginary part (σ″) of CC is a direct measure of the interfacial properties (equation (6)) and is therefore sensitive to changes in surface electrochemistry.

COMPLEX CONDUCTIVITY MODELLING

Mechanisms that generate CC responses in soils are not yet entirely understood. Surface conductivity is a result of electrochemical processes governed by the chemical and physical characteristics of soils. The interpretation of the CC signature associated with clay–EtOH–water mixtures is therefore challenging. There is currently no published physicochemical model that can describe such interactions. Instead, phenomenological models such as the Cole–Cole relaxation models can be used to empirically describe the CC dependence on physical and chemical properties of soils (Cole and Cole 1941; Davidson and Cole 1951).

Similarly, the Debye decomposition (DD) procedure can be used to describe the CC spectra in terms of a few quantifiable parameters (Lesmes and Morgan 2001; Nordsiek and Weller 2008). As compared to Cole–Cole type models, the DD method fits a wider range of shapes of the phase spectra (Nordsiek and Weller 2008). They defined the DD in terms of complex resistivity

ρ∗(ω)=1σ0[1−∑k=1nmk(1−11+iωτk)], (7)

where σ0 is the DC conductivity, τ is the relaxation time, mk is the chargeability, ω is the angular frequency, and n is the number of individual Debye responses. We also impose positive chargeability values by solving the problem with a non-negative least-squares approach and invoke a chargeability normalisation that ensures that the sum of chargeability values in equation (7) does not exceed 1. This approach relies on the decomposition of the CC spectra into n discrete Debye responses with specific chargeability (mk) and relaxation time (τk) (for more details, refer to Zisser, Kemna and Nover 2010 and Nordsiek and Weller 2008). The integral or total chargeability (m) is then obtained by

m=∑k=1nmk. (8)

Chargeability represents the ratio of polarization to conduction within the sample. To obtain a global direct estimate of the polarizability of the material over the measured frequency range, normalised chargeability must be computed (Lesmes and Frye 2001; Slater and Lesmes 2002),

mn=mσ0. (9)

The concept of specific polarizability was first introduced by Weller et al. (2010). Evidence from previous research suggests strong dependence of pore-normalised specific surface area (Sp) on CC responses for various sediments. Sp is typically computed from

Sp=Sm×ρd×(1−Φ)Φ, (10)

where Φ is the porosity, ρd is the mineral density (g/m3), and Sm is the specific surface area (m2/g), which can be measured by nitrogen adsorption using the Brunauer–Emmett–Teller (BET) method (Brunauer, Emmett and Teller 1938).

Weller et al. (2010) investigated the relationship between pore-normalised specific surface area (Sp) and imaginary conductivity measured at 1 Hz or normalised chargeability mn determined from DD for an extensive sample set. They defined measures of specific polarizability,

cp=σ′′Sp (11)
cs=mnSp, (12)

where cp is determined from a single-frequency CC measurement and cs is defined from the DD of broadband CC data (Weller et al. 2011). Weller et al. (2010) showed that this σ′′←→Sp relationship fits an extensive database of sandstone and unconsolidated sediments. They present evidence that cp depends slightly on clay content, being higher for samples containing clay (cp = 11.5 × 10−12 S) than sandy material (cp = 7.2 × 10−12 S). Weller et al. (2010) discussed the concept of specific polarizability as a representation of polarization magnitude per unit pore-volume-normalised surface area and assumed it to be exclusively determined by the chemistry and mineralogy of the polarized grain–fluid interface. Weller et al. (2015) further showed strong evidence for this concept for CC measurements on clay–sand mixtures where Sp was more reliably measured using the methylene blue method. Therefore, the specific polarizability concept appears well suited to represent the dependence of the CC on the electrochemical factors controlling the induced polarization (IP) responses independent of the total polarizable surface area of the sample and is utilised here.

METHODS

Sample mixtures were prepared by carefully combining the various types of clay with Ottawa sand. The mixture for kaolinite was prepared by adding 4% w/w sample to 96% w/w Ottawa sand, and 2% illite mixture was prepared by adding 2% w/w sample to 98% Ottawa sand. The small clay fraction was chosen to ensure enough clay within the sand matrix to allow for a measurable electrical response while also preventing a large reduction in permeability. This procedure was similar to that adopted by Personna et al. (2013) in order to ensure consistency in experimental procedures and to permit comparison against previous results obtained for bentonite following similar procedures. The nitrogen-gas-adsorption BET method (Brunauer et al. 1938) was used to determine specific surface area (Ss) of kaolinite (9.73 ± 0.03 m2/g), illite (34.54 ± 0.20 m2/g), bentonite (256.76± 0.45 m2/g), and Ottawa sand (0.049 ± 0.002 m2/g).

Cylindrical PVC columns were dry packed with a mixture of Ottawa sand (density (ρ) = 2.64 g/cm3, d10 = 0.096 mm) and (1.) kaolinite (1:1 clay, ρ = 2.62 g/cm3, d10 = 0.0010 mm) or (2.) illite (2:1 clay, ρ = 2.77 g/cm3, d10 = 0.0013 mm). The PVC column schematic is summarised in Figure 2.

Figure 2.

Figure 2

Open in a new tab

Experimental setup schematic showing the sample holders constructed using cylindrical transparent PVC (inner diameter = 0.025 m and length = 0.028 m). The locations of two coiled Ag–AgCl current electrodes (C1 and C2) and two Ag–AgCl point potential electrodes (P1 and P2) are shown. The current electrodes are embedded in the end caps on the top and bottom of the column setup. The potential electrodes are uncoiled point electrodes placed in the end caps.

The mixtures were dry packed into a PVC column with a 0.45-μm cellulose membrane on either side to ensure no loss of clay during the subsequent saturation process. In addition, the flow rate was kept low to ensure no redistribution of clay within the column. We found that dry packing promoted a more homogenous mixture of clay and sand and minimised redistribution of clay associated with variations in grain sizes of the clay relative to the sand mixture. The saturating fluid was either 300-μS/cm background tap water solution or the background solution mixed with pure ethanol (ethyl alcohol of 200 proof and 99.98% assay v/v from Pharmco-AAPER) in different concentrations by volume. Personna et al. (2013) used a mixture of 2% (w/w) bentonite clay with 98% (w/w) Ottawa sand, with column packing procedures and setup identical to ours. The solution used by Personna et al. (2013) was either background water or a water–ethanol mixture made with the same pure ethanol (ethyl alcohol of 200 proof and 99.98% assay v/v from Pharmco-AAPER) used in our study. The water used in their experiment was well characterised, containing major ions of natural groundwater and having an electrical conductivity value of 266.3 μS/cm and comparable with the conductivity of our background solution (300 μS/cm).

The experimental treatments applied in our CC measurements are summarised in Figure 3. The kaolinite mixture contained 4% (w/w) clay with 96% (w/w) Ottawa sand. The illite and bentonite mixtures contained 2% (w/w) clay with 98% (w/w) Ottawa sand. CC measurements with 2% kaolinite were also acquired but responses were unreliable due to the small polarization associated with the low-surface-area kaolinite, preventing a reliable DD fitting. As cp represents the polarization normalised for the available surface area of the pores, the higher percentage concentration of the kaolinite does not impact the analysis. Three different experimental treatments were applied (Figure 3): (a) sand–clay mixture saturated with the background solution (EtOH blank), (b) sand–clay mixture saturated with the solution containing 10% EtOH mixture (EtOH 10%, v/v), and (c) sand–clay mixture saturated with the solution containing 20% EtOH mixture (EtOH 20%, v/v). Measurements were also conducted on sand samples with water and 10% EtOH. Samples were saturated with their respective saturating solutions and placed in the same saturating fluid at the end of each measurement to allow repeated observations, required to monitor any temporal changes in CC response due to possible non-equilibrium effects (Personna et al. 2013).

Figure 3.

Figure 3

Open in a new tab

Summary of experimental treatments applied to samples. Personna et al. (2013) applied a similar experimental treatment to bentonite clay samples, with EtOH blank being tap water instead of KCl.

Saturation and porosity of the samples were determined from weight loss between saturated and dried samples after completion of CC measurements. Sample porosities and saturations are given in Table 1. For each sample mixture, pore-volume-normalised specific surface area Sp (μm−1) was calculated using

Sp=[(Ssx×ρx×Wx)+(SsOS×ρOS[1−Wx])]×1−ΦΦ, (15)

where is in m2/g, Φ is the porosity of the mixture, W is fraction by weight, ρ is density (g/cm3), OS refers to Ottawa sand, and x refers to the clay mineral used in the mixture. The physical characteristics of all clay minerals used are summarised in Table 1.

Table 1.

Summary of measured physical properties of clay and sand–clay mixtures.

Sample D10 Particle size (mm) Porosity (%) Surface area (m2/g)
Ottawa sand 0.6–0.7 - 0.0485 ± 0.0015
Kaolinite (1:1) 0.0010± 0.00011 34.3 9.7295 ± 0.0250
Illite (2:1) 0.0013±0.00012 38.7 34.5386 ± 0.2002
Bentonite (2:1) 0.0030±0.00010 37 256.7876 ± 0.4485
Sand–clay mixture Sp (μm−1) Saturation (%)
4% Kaolin + 96% Sand 2.18±0.015 96.37 ± 1.54
2% Illite + 98% Sand 3.60±0.020 88.35 ± 2.36
2% Bentonite + 98% Sand 5.40±0.010 -
Open in a new tab

ANI4661 dynamic signal analyser was used for all CC measurements. The conductivity magnitude (|σ*|) and phase shift (ϕ) were measured relative to a reference resistor. CC measurements were acquired for 51 frequencies between 0.01 and 1000 Hz and were available at 41 frequencies between 0.1 and 1000 Hz for the bentonite sample of Personna et al. (2013). Silver–silver chloride electrodes were used for current injection and for recording potential measurements. The DD approach was used to fit measured CC data to compute a specific integral chargeability (mk) and relaxation time (τk) to determine the CC model parameters (equations (7), (8), and (9)). Analysis of variance (ANOVA), performed with ezANOVA (www.cabiatl.com/mricro/ezanova/), was used to test the statistical significance (level of significance α = 0.05) of the changes in the CC data as a function of EtOH concentration.

ANOVA compares the means between the groups in question and determines how statistically different their means are by testing the null hypothesis (i.e., the means are statistically equivalent). A statistically significant ANOVA result expresses that the null hypothesis is false.

RESULTS

Figure 4 shows the temporal variation in the electrical response for 2% w/w illite sample saturated with 20% v/v EtOH. Figure 5 summarises the specific polarizability (cp) spectra for the three different clay–sand mixtures: 4% kaolinite, 2% illite, 2% bentonite, and pure Ottawa sand. The cp spectra shown in Figure 5 represent measurements that were taken 20 to 25 days after column preparation and pore fluid injection. This time lapse since sample preparation was necessary for the system to reach equilibrium (Figure 4). After this time, no more significant change in specific polarizability was observed for bentonite, illite, and kaolinite containing mixtures. All the results are reported by Sharma (2016).

Figure 4.

Figure 4

Open in a new tab

Temporal variation in the electrical response (σ″) for 2% w/w illite sample saturated with 20% v/v EtOH for the entire measured frequency range.

Figure 5.

Figure 5

Open in a new tab

Specific polarizability (cp) responses for different clay minerals: (a) bentonite (green), (b) illite (red), (c) kaolinite (black), and (d) Ottawa sand as a function of ethanol concentration (v/v) (EtOH 0%, EtOH 10%, and EtOH 20%).

We reiterate that the use of specific polarizability is critical to our analysis as it removes the effects of variable polarization between the clay minerals due to variations in total surface area, allowing us to focus on differences that can be attributed to variations in interfacial chemistry associated with clay organic reactions. We observed a clear variation in the specific polarizability (cp) of the sand–clay samples with increasing EtOH concentration (Figure 5). The kaolinite mixture shows the strongest decrease in cp with increasing EtOH concentration over the entire frequency range (Figure 5). The mixture containing illite also shows similar suppression in specific polarizability with respect to increasing EtOH concentration across the entire frequency range. However, the dependence of imaginary conductivity on the frequency complicates data interpretation for cp as imaginary conductivity values at different frequencies will result in different cp estimates. Therefore, more emphasis was placed on the variation of cs as it is determined from normalised chargeability, which is a global estimate of polarization for the sample and therefore captures the frequency variation. All sand–clay mixtures show similar changes in cs with increasing EtOH concentration in the pore-filling fluid (Figure 6).

Figure 6.

Figure 6

Open in a new tab

Specific polarizability (cs) versus measured EtOH concentration for different sand–clay mixtures.

The percent relative changes in cp, cs, and real conductivity for the 20% EtOH samples were calculated based on the 10% EtOH sample as the reference (Table 2). The overall percentage suppression is similar between all three clay types (Table 2). The relative changes in specific polarizability based on increasing concentration are always higher than relative changes in the real conductivity component (spectra not shown for brevity) and as summarised in Table 2. The mixture containing bentonite shows a slightly larger relative decrease in cs with increasing EtOH concentration, but the degree of suppression is comparable with the suppression response for illite and kaolinite mixtures. Table 2 further emphasises that all three clays show a similar degree of suppression in cs responses with increasing EtOH concentration. The ANOVA for the level of significance α = 0.05 indicates that the suppression of clay-driven polarization response with increasing EtOH concentration observed in Figure 6 is statistically significant.

Table 2.

Percentage changes (%) in cp (1 Hz), cs, and real conductivity (1 Hz). Negative values indicate decrease relative to values obtained for the 10% EtOH sample.

Sample mixture cs cp σ′
20% EtOH 20% EtOH 20% EtOH
2% Bentonite + Sand −64.6 −73.3 12.5
2 % Illite + Sand −60.1 −68.5 −1.5
4% Kaolinite + Sand −42.9 −87.3 8.2
Open in a new tab

DISCUSSION

We observe clear and statistically significant variations in the clay-driven polarization response for different clay–sand mixtures with increasing EtOH concentration (Figure 5). We observe suppression in cs for all clays with increasing EtOH concentration (Figure 6) that is similar for each clay mineral. The suppression effect of EtOH on polarization for the bentonite mixtures has been previously interpreted in terms of (a) preferential adsorption of EtOH relative to water onto clay and resultant changes in clay structure and (b) complex intermolecular interactions between EtOH and water resulting in changes in mobility of ions in the EDL at the mineral surface (Personna et al. 2013). Changes in specific polarizability can thus be attributed to geochemical alterations on clay surfaces and changes in pore fluid chemistry, which are not observed for EtOH present in pure Ottawa sand. Suppression could also be occurring due to organic interactions with the silica surface for clean sands, but the small phase angles recorded made it difficult to confidently interpret the data.

Personna et al. (2013) focused on cation–dipole interactions, among others, to potentially explain the adsorption of EtOH on bentonite clay resulting in the alteration of surface chemistry. Similarly, suppression in clay-driven polarization for illite can be explained by preferential adsorption of EtOH onto clay surfaces. However, the relative reduction in cs with increasing ethanol concentration is comparable for all clay types (Table 2). The ANOVA results (α = 0.05) indicate that the differences between cs values are significant for all the sand–clay mixtures. In this experiment, the clay mineralogy (2:1 versus 1:1 clay) appears to have no significant effect on the polarization suppression magnitude once the effect of the pore-volume-normalised surface area is accounted for.

We initially hypothesised that the suppression of polarization would be greatest for bentonite followed by illite and then kaolinite. This hypothesis was based on large differences in surface area for each clay mineral. However, specific polarizability (cs) is not significantly different between the three clay minerals at all ethanol concentrations showing that surface polarizability is comparable for each clay mineral (Table 2). Thus, the mineral–fluid electrochemistry results in a very similar electrical response for the three clay minerals examined. Kaolinite and illite clay exhibit a similar polarization per pore-normalised surface area to that previously observed for bentonite. However, the clay organic interactions might be significantly different between the clay minerals. For non-swelling 1:1 kaolinite clay, the surfaces may act like oxides and thereby interact with polar ethanol molecules (Harris, Wells and Johnson 2001). For illite, there might be a combination of adsorption sites available at the surface and between individual clay sheets. The adsorption of ethanol between the individual clay sheets (Figure 1) can only be hypothesised and requires X-ray diffraction to measure such effects.

As shown in Table 2, the relative changes in specific polarizability (cs and cp) are larger than relative changes in real conductivity. This observation emphasises the sensitivity of CC to alterations in surface geochemistry of different clay types over traditional resistivity methods where the effect of pore fluid chemistry often dominates the surface effects. The presence of EtOH in the pore-filling fluid results in suppression of the polarization response for sand–clay mixtures, presumably due to adsorption of EtOH onto clay surfaces. The total adsorption is directly related to the available surface-area-to-pore-volume ratio of the mixture. CC measurements on sand saturated with water and 10% v/v EtOH were also acquired but the responses were too low to be confidently interpreted within the context of determining relative changes partly because of the sensitivity limits of the instrumentation.

Our findings are consistent with recent studies on clay-organic interactions carried out by Ustra et al. (2012). However, we observe greater clay-driven polarization suppression in our EtOH solutions as opposed to that recorded by Ustra et al. (2012) for toluene. Ethanol, being a polar molecule, has a higher tendency for adsorption onto clay surfaces as compared to non-polar toluene. The suppression in cs response is a direct result of the increase in EtOH concentration in the pore-filling fluid (Figure 5). Our data indicate that CC measurements are sensitive to clay-driven suppression in polarization in response to adsorption of EtOH onto the internal and external surfaces of different clays.

CONCLUSIONS

Clay-driven polarization changes are observed for both illite and kaolinite that are largely consistent with that previously observed only for the study of the response for bentonite. The relative reduction in specific polarizability cs with increasing ethanol concentration is comparable for all clay types. Pure sand showed no polarization response to the presence of EtOH (within the sensitivity limits of the instrumentation), supporting the hypothesis that clay-organic interactions cause the suppression effect observed in these datasets with the clay minerals. The cs estimates reflect the sensitivity of the CC measurements to investigate alteration in surface chemistry at the available surface adsorption sites (both internal and external) for different clay types resulting from various chemical ion exchange and kinetic reactions in clay–water–EtOH media. Our results indicate that the available surface area and the concentration of the ethanol control the IP response for clays in the presence of ethanol, and these factors supersede any minor effect of clay mineralogy. This method cannot differentiate between individual clay minerals based on electrical measurements, but it is sensitive to clay-organic interactions and might ultimately be used to delineate contaminant plumes in the subsurface.

ACKNOWLEDGEMENTS

The United States Environmental Protection Agency (EPA) partially funded this project under student services contract (EP-13-D-000143) to Sundeep Sharma. Any mention of trade names, products, or services does not imply an endorsement by the US Government or the EPA. The EPA does not endorse any commercial products, services, or enterprises. The views expressed in this article are those of the authors and do not necessarily represent the views or policies of the EPA. The authors would like to thank Martin Briggs (US Geological Survey) for his critical suggestions for this manuscript and especially thank Yves Personna for letting them use his data for comparison. Sundeep Sharma would like to thank you his colleagues Jeff Heenan and Casey McGuffy for all their help and support on this project and especially thank all faculty and administrative members in the Department of Earth and Environmental Sciences, Rutgers University–Newark, for all their support during his time there as a graduate student. Lastly, he would like to thank his parents, Lata and Krishan Sharma, for their continued love and support.

REFERENCES

  1. Archie G 1942. The electrical resistivity log as an aid in determining some reservoir characteristics. Transactions of the AIME 146(99), 54–62. [Google Scholar]
  2. Atamas N and Atamas A 2009. The investigations of water-ethanol mixture by Monte Carlo method. World Academy of Science, Engineering and Technology 55, 4 p. [Google Scholar]
  3. Bhat JI and Shetty MK 2011. Evaluation of limiting molar conductance, Walden product, association constant and thermodynamic properties of sulfacetamide sodium in water+EtOH mixtures. Journal of Molecular Liquids 160(3), 140–143. [Google Scholar]
  4. Brown SR, Sorenson JR and Brown TI 2004. A laboratory study of the complex electrical resistivity response of soils. Symposium on the Application of Geophysics to Engineering and Environmental Problems 2004, pp. 528–539. Society of Exploration Geophysicists. [Google Scholar]
  5. Brunauer S, Emmett PH and Teller E 1938. Adsorption of gases in multimolecular layers. Journal of the American Chemical Society 60(2), 309–319. [Google Scholar]
  6. Cole KS and Cole RH 1941. Dispersion and absorption in dielectrics I. Alternating current characteristics. The Journal of Chemical Physics 9, 341. [Google Scholar]
  7. Corseuil HX, Hunt CS, Ferreira dos Santos RC and Alvarez PJ 1998. The influence of the gasoline oxygenate ethanol on aerobic and anaerobic BTX biodegradation. Water Research 32(7), 2065–2072. [Google Scholar]
  8. Craig RF 1974. Craig’s Soil Mechanics. New York, NY: Spon Press. [Google Scholar]
  9. Davidson D and Cole R 1951. Dielectric relaxation in glycerol, propylene glycol, and n-propanol. The Journal of Chemical Physics 19, 1484. [Google Scholar]
  10. Dowdy RH and Mortland MM 1967. Alcohol-water interactions on montmorillonite surfaces. I. Ethanol. Clays and Clay Minerals 15(1), 259–271. [Google Scholar]
  11. German W and Harding D 1969. The adsorption of aliphatic alcohols by montmorillonite and kaolinite. Clay Miner 8, 213–227. [Google Scholar]
  12. Gomez DE and Alvarez PJJ 2010. Comparing the effects of various fuel alcohols on the natural attenuation of benzene plumes using a general substrate interaction model. Journal of Contaminant Hydrology 113(1–4), 66–76. [DOI] [PubMed] [Google Scholar]
  13. Harris RG, Wells JD and Johnson BB 2001. Selective adsorption of dyes and other organic molecules to kaolinite and oxide surfaces. Colloids and Surfaces A: Physicochemical and Engineering Aspects 180(1), 131–140. [Google Scholar]
  14. Lesmes DP and Frye KM 2001. Influence of pore fluid chemistry on the complex conductivity and induced polarisation responses of Berea sandstone. Journal of Geophysical Research 106(B3), 4079–4090. [Google Scholar]
  15. Lesmes DP and Morgan FD 2001. Dielectric spectroscopy of sedimentary rocks. Journal of Geophysical Research 106(B7), 13329–13346. [Google Scholar]
  16. Lovanh N, Hunt CS and Alvarez PJJ 2002. Effect of ethanol on BTEX biodegradation kinetics: aerobic continuous culture experiments. Water Research 36(15), 3739–3746. [DOI] [PubMed] [Google Scholar]
  17. Miranda-Trevino JC and Coles CA 2003. Kaolinite properties, structure and influence of metal retention on pH. Applied Clay Science 23(1–4), 133–139. [Google Scholar]
  18. Nordsiek S and Weller A 2008. A new approach to fitting induced-polarisation spectra. Geophysics 73(6), F235–F245. [Google Scholar]
  19. Okay G, Leroy P, Ghorbani A, Cosenza P, Camerlynck C, Cabrera J et al. 2014. Spectral induced polarisation of clay-sand mixtures: experiments and modeling. Geophysics 79(6), E353–E375. [Google Scholar]
  20. Olhoeft GR 1985. Low frequency electrical properties. Geophysics 50(12), 2492–2503. [Google Scholar]
  21. Olhoeft GR 1986. Direct detection of hydrocarbon and organic chemicals with ground penetrating radar and complex resistivity. NWWA/API Conference on Petroleum Hydrocarbons and Organic Chemicals in Ground Water, Dublin, OH, November 12–14, 1986, pp. 284–305. [Google Scholar]
  22. Olhoeft G and King T 1991. Mapping subsurface: organic compounds noninvasively by their reactions with clays. US Geological Survey Toxic Substance Hydrology Program, Proceedings of Technical Meeting, Monterey, CA, pp. 552–557 [Google Scholar]
  23. Österreicher-Cunha P, Guimarães JRD, do Amaral Vargas E and Pais da Silva MI 2007. Study of biodegradation processes of BTEX-ethanol mixture in tropical soil. Water, Air, and Soil Pollution 181(1–4), 303–317. [Google Scholar]
  24. Parke SA and Birch GG 1999. Solution properties of ethanol in water. Food Chemistry 67(3), 241–246. [Google Scholar]
  25. Personna YR, Slater L, Ntarlagiannis D, Werkema D and Szabo Z 2013. Complex resistivity signatures of ethanol in sand-clay mixtures. Journal of Contaminant Hydrology 149, 76–87. [DOI] [PubMed] [Google Scholar]
  26. Powers SE, Hunt CS, Heermann SE, Corseuil HX, Rice D and Alvarez PJ 2001a. The transport and fate of ethanol and BTEX in groundwater contaminated by gasohol. Critical Reviews in Environmental Science and Technology 31(1), 79–123. [Google Scholar]
  27. Powers SE, Rice D, Dooher B and Alvarez PJ 2001b. Will ethanol-blended gasoline affect groundwater quality? Environmental Science & Technology 35, 24–30. [DOI] [PubMed] [Google Scholar]
  28. Revil A, Karaoulis M, Johnson T and Kema A 2012. Review: some low-frequency electrical methods for subsurface characterization and monitoring in hydrogeology. Hydrogeology Journal, 1–42 [Google Scholar]
  29. Roberts J and Wildenschild D 2004. Electrical properties of sand–clay mixtures containing trichloroethylene and ethanol. Journal of Environmental & Engineering Geophysics 9, 1. [Google Scholar]
  30. Ruiz-Aguilar GML, Fernandez-Sanchez JM, Kane SR, Kim D and Alvarez PJJ 2002. Effect of ethanol and methyl-tert-butyl ether on monoaromatic hydrocarbon biodegradation: response variability for different aquifer materials under various electron-accepting conditions. Environmental Toxicology and Chemistry 21(12), 2631–2639. Society of Environmental Toxicology and Chemistry. [PubMed] [Google Scholar]
  31. Sadowski RM 1988. Clay-organic interactions. MSc thesis, Colorado School of Mines, USA. [Google Scholar]
  32. Saidy AR, Smernik RJ, Baldock JA, Kaiser K, Sanderman J and Macdonald LM 2012. Effects of clay mineralogy and hydrous iron oxides on labile organic carbon stabilisation. Geoderma 173, 104–110. [Google Scholar]
  33. Saidy AR, Smernik RJ, Baldock JA, Kaiser K and Sanderman J 2013. The sorption of organic carbon onto differing clay minerals in the presence and absence of hydrous iron oxide. Geoderma 209–210, 15–21. [Google Scholar]
  34. Schaefer CE, Yang X, Pelz O, Tsao DT, Streger SH and Steffan RJ 2010. Aerobic biodegradation of iso-butanol and ethanol and their relative effects on BTEX biodegradation in aquifer materials. Chemosphere 81(9), 1104–1110. [DOI] [PubMed] [Google Scholar]
  35. Sharma S 2016. Specific polarizability of sand-clay mixtures with varying ethanol concentration. Master’s dissertation, Rutgers University–Newark, USA. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Slater L and Lesmes D 2002. Electrical–hydraulic relationships observed for unconsolidated sediments. Water Resources Research 38(10), 31. [Google Scholar]
  37. Spagnoli G, Fernández-Steeger T, Hu H, Feinendegen M and Azzam R 2010. Potential calculation according to the Gouy and the Stern model for kaolinite and smectite. Giornale di Geologia Applicata 13, 87–91. [Google Scholar]
  38. Sposito G, Skipper NT, Sutton R, Park S, Soper AK and Greathouse JA 1999. Surface geochemistry of the clay minerals. Proceedings of the National Academy of Sciences of the United States of America 96(7), 3358–3364. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Ustra A, Slater L, Ntarlagiannis D and Elis VR 2012. Spectral induced polarisation (SIP) signatures of clayey soils containing toluene. Near Surface Geophysics 10(6), 503–515. [Google Scholar]
  40. van Olphen H 1977. An Introduction to Clay Colloid Chemistry. For Clay Technologists, Geologists, and Soil Scientists, 2nd edn, pp. 57–120. New York, NY: John Wiley & Sons. [Google Scholar]
  41. Vinegar HJ and Waxman MH 1984. Induced polarisation of shaly sands. Geophysics 49(8), 1267–1287. [Google Scholar]
  42. Waxman MH and Smits LJM 1968. Electrical conductivities in oil-bearing shaly sands. Society of Petroleum Engineers Journal 8(02), 107–122. [Google Scholar]
  43. Weller A, Slater L, Nordsiek S and Ntarlagiannis D 2010. On the estimation of specific surface per unit pore volume from induced polarisation: a robust empirical relation fits multiple data sets. Geophysics 75(4). [Google Scholar]
  44. Weller A, Breede K, Slater L and Nordsiek S 2011. Effect of changing water salinity on complex conductivity spectra of sandstones. Geophysics 76(5), F315. [Google Scholar]
  45. Weller A, Slater L, Huisman JA, Esser O and Haegel F 2015. On the specific polarizability of sands and sand-clay mixtures. Geophysics 80(3), 3–7. [Google Scholar]
  46. Zisser N, Kemna A and Nover G 2010. Relationship between low-frequency electrical properties and hydraulic permeability of low-permeability sandstones. Geophysics 75, E131–E141. [Google Scholar]

RESOURCES