Environ. Sci. Technol. 2010, 44, 9470–9475
Adhesion of Oil to Kaolinite in Water EVGENIA V. LEBEDEVA AND ANDREW FOGDEN* Department of Applied Mathematics, Research School of Physics and Engineering, Australian National University, Canberra ACT 0200, Australia
Received June 16, 2010. Revised manuscript received October 28, 2010. Accepted October 28, 2010.
angles are measured using drop or plate methods (10), or intermolecular interactions are measured, e.g. using atomic force microscopy (AFM) (11). Extension of these approaches to measurement of NAPL wetting, spreading and adhesion on and to kaolinite in aqueous environments poses added complications. Primarily, any resuspension of kaolinite particles in the surrounding aqueous phase during measurement would invalidate results. Consequently, wettability and adhesion studies on such systems have been limited to model smooth substrates, e.g. quartz (13, 14, 1), mica (15-18), glass (13, 18-20), and calcite (6). For an oil drop in water on such a substrate, at equilibrium the work of adhesion is given by the Dupre equation (17) W ) γow(1 - cos θ)
Uniform coats of kaolinite particles on a flat glass substrate were prepared to be sufficiently smooth and thin to allow reliable measurement of contact angles of captive crude oil drops in a range of salt solutions, without any particle removal. The contact angle hysteresis was used to infer the extent of oil adhesion via rupture of the intervening water film and anchoring of charged groups to kaolinite. For sodium chloride solutions, adhesion decreases monotonically with pH and/or salinity, with strong adhesion only manifested under acidic conditions with salinity at most 0.1 M. Calcium chloride solutions at pH around 6 switch from strong adhesion in the range 0.001-0.01 M to weak adhesion at higher concentrations. For all mixtures of sodium and calcium chlorides investigated, a total ionic strength above 0.1 M guarantees a weak adhesion of oil to kaolinite. Results are qualitatively consistent with theoretical expectations of electrostatic interactions, with H+ and Ca2+ being potentialdetermining ions for both interfaces.
Introduction Efficient strategies for remediation of soil and subsurface contamination by nonaqueous phase liquids (NAPLs) must take into account their ability to adhere to minerals. Adhesion occurs if the NAPL a) comes into molecular contact with the naturally water-wet mineral, via destabilization of the intervening water film lining it, and b) subsequently interfacially bonds to the mineral, rendering it oil-wet (1). In this adhering state, the NAPL poses significant challenges for remediation, especially if it is located in small pores with high internal surface area, associated with clays. Among the various clay types, kaolinite has been found to have a high affinity for petroleum-based light or dense NAPL contaminants such as crude oil or coal tar and creosote (2, 3). Kaolinite is a layered aluminosilicate comprising a tetrahedral silica sheet bonded to an octahedral alumina sheet by sharing of oxygen atoms, with successive 1:1 layers held together by hydrogen bonds (4, 5). For these asphaltic NAPLs, the wettability altering components are thought to be chiefly the polyaromatic polar molecules termed asphaltenes (6, 7). Improved understanding of the interaction of asphaltic NAPL with kaolinite in aqueous phases of varying salinity also has relevance to kaolinite in filters to treat oily wastewater (8) and in bioremediation of oil spills (9). Wetting and spreading of aqueous phases on kaolinite in air have been quantified by preparing relatively flat, low porosity beds of the particles (10-12), on which contact * Corresponding author phone: 61-261254823; fax: 61-261250732; e-mail:
[email protected]. 9470
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 24, 2010
(1)
where γow is the oil-water interfacial tension, and θ is the contact angle, measured through the water. If the drop is first grown (i.e., water recedes, R), aged, and subsequently retracted (water advances, A), the change in adhesion is ∆W ) WA - WR ) γow(cos θR - cos θA)
(2)
The contact angle difference θA - θR, termed the hysteresis, is thus indicative of the oil-substrate adhesion developed over the aging period. Experimental protocols use either the real liquids for relatively short aging times (13, 15, 19, 20) or probe liquids after longer-time substrate aging in the real liquids and rinsing (13-16, 20). Adhesion has also been directly measured using the surface force apparatus (17) and AFM (18). These contact angle studies (1, 13-20) have contributed greatly to understanding of the intermolecular mechanisms by which wettability alteration and adhesion occur. The mechanisms include polar interactions, surface precipitation, acid/base interaction, and ion binding and are highly sensitive to the salts dissolved in the water and its pH. For the simplest situation of a 1:1 electrolyte, almost all studies on different combinations of crude oils and smooth silicate substrates yield strong adhesion at low salinity and pH and no adhesion at high salinity and pH. In the former case, the oil-water and water-silicate interfaces have opposite charge, thus the water film separating them is ruptured by their long-range attraction, in accordance with the Derjaguin-LandauVerwey-Overbeek (DLVO) theory (1, 19). Non-DLVO interactions are often invoked to explain the nonadhering latter case (18, 19). Kaolinite can give rise to even more complex intermolecular mechanisms for adhesion, thus there is a need to extend the studies of kaolinite layers exposed to aqueous phases in air, and smooth, impervious mineral substrates exposed to NAPL in water, to begin to address these realities. The current study investigated adhesion of NAPL, in particular a crude oil, to kaolinite in water as a function of dissolved salt, in particular sodium and calcium chlorides, and pH. To this end, smooth, thin, porous kaolinite coats, which completely cover their glass substrate and remain unperturbed in water and NAPL, were developed to facilitate assessment of NAPL drop adhesion via contact angle hysteresis. The measured hysteresis was compared to the surface charge of the kaolinite and NAPL in isolation in these same aqueous phases to analyze the electrostatic interactions causing adhesion and qualitatively evaluate the suitability of DLVO theory for predicting adhesion and designing remediation strategies. 10.1021/es102041b
2010 American Chemical Society
Published on Web 11/24/2010
TABLE 1. Matrix of 25 Salt Solutions at pH 4, Natural (pH 5.5-6.1, Denoted n) and 9, and the Corresponding Statistics of Zeta Potential of the Kaolinite Suspension and NAPL Emulsion, Receding and Advancing NAPL-Water-Kaolinite Contact Angles, and Their Difference (Hysteresis) pH
[NaCl], M
[CaCl2], M
ζ kaolinite, mV
ζ NAPL, mV
receding, degrees
advancing, degrees
hysteresis, degrees
n n n n n n n n n n n n n n n n n 4 4 4 4 9 9 9 9
0 0.01 0.1 1 0 0 0.01 0.1 1 0 0.01 0.1 1 0 0.01 0.1 1 0 0.01 0.1 1 0 0.01 0.1 1
0 0 0 0 0.001 0.01 0.01 0.01 0.01 0.1 0.1 0.1 0.1 1 1 1 1 0 0 0 0 0 0 0 0
-23.7 ( 0.3 -18.5 ( 0.7 -18.1 ( 0.8 -11.1 ( 1.9 -12.0 ( 0.9 -7.5 ( 0.2 -6.8 ( 0.4 -11.9 ( 0.4 -10.2 ( 0.9 0.9 ( 3.0 -2.6 ( 0.1 -3.4 ( 0.9 -6.2 ( 0.5 2.5 ( 0.8 1.6 ( 1.0 1.5 ( 1.4 1.3 ( 0.5 -15.8 ( 0.8 -16.7 ( 1.1 -14.5 ( 0.9 -5.1 ( 0.2 -54.6 ( 0.9 -47.6 ( 0.8 -39.9 ( 2.0 -15.1 ( 1.0
-66.6 ( 1.6 -51.2 ( 2.0 -17.1 ( 0.3 -12.5 ( 1.4 -44.3 ( 0.8 -27.5 ( 1.2 -13.9 ( 0.6 -16.3 ( 1.5 -12.9 ( 1.5 -7.8 ( 3.1 -7.6 ( 0.4 -7.6 ( 0.2 -9.6 ( 1.3 0.1 ( 1.4 -2.1 ( 1.1 1.6 ( 0.5 1.6 ( 0.9 21.6 ( 1.5 -7.1 ( 0.8 -5.7 ( 0.5 -9.8 ( 1.1 -127.0 ( 5.3 -78.9 ( 1.0 -24.4 ( 0.5 -16.0 ( 0.6
39 ( 4 35 ( 4 37 ( 1 44 ( 6 30 ( 2 30 ( 2 32 ( 4 32 ( 3 35 ( 4 37 ( 4 37 ( 3 40 ( 2 39 ( 4 36 ( 0 35 ( 5 28 ( 2 40 ( 1 33 ( 3 44 ( 16 43 ( 6 38 ( 1 32 ( 4 30 ( 2 36 ( 4 23 ( 2
74 ( 9 72 ( 21 62 ( 6 53 ( 5 87 ( 25 90 ( 5 52 ( 5 41 ( 0 36 ( 3 50 ( 5 46 ( 9 57 ( 13 40 ( 3 47 ( 1 51 ( 4 30 ( 1 49 ( 4 121 ( 5 110 ( 8 89 ( 29 50 ( 7 44 ( 7 62 ( 11 64 ( 6 30 ( 6
35 ( 13 36 ( 24 24 ( 6 9 ( 11 57 ( 23 60 ( 3 20 ( 7 8(3 2(4 13 ( 10 9 ( 12 17 ( 15 1(4 11 ( 1 16 ( 9 2(2 9(5 88 ( 3 66 ( 17 46 ( 23 13 ( 8 12 ( 3 32 ( 8 28 ( 8 7(8
Materials and Methods Preparation of Kaolinite Substrates. Well-crystallized kaolinite KGa-1b (Washington County, Georgia; Clay Minerals Society Source Clay Repository) comprises 96% kaolinite and trace dickite, 3% anatase, and 1% crandallite plus mica and/ or Illite and has cation exchange capacity 1-10 cmolc/kg and BET surface area 10 m2/g (4, 21). Elemental analysis by energy-dispersive X-ray spectroscopy is presented in the Supporting Information Table S1. No further cleaning of the kaolinite was performed. An aqueous suspension was prepared from 10 g of kaolinite in 150 g of deionized water from a Millipore Milli-Q reverse osmosis system (as used throughout, unless otherwise specified), with rapid magnetic stirring followed by sonication for 10 min. The pH was adjusted to 9.8 using NaOH and maintained in all subsequent steps. The suspension was centrifuged at 160 g for 3 min, after which the suspended upper phase was decanted (rejecting the sediment), then rapidly stirred, and sonicated for 10 min. This process of centrifugation, decanting, stirring, sonication, and pH adjustment was repeated at the centrifugal forces of 640 and 1000 g, for 3 min, and finally at 1000 g for 5 min. The resulting dilute suspension of fine kaolinite was heated at 85 °C under rapid stirring until its concentration reached 7 wt %. Approximately 0.05 g of suspension was pipetted as several drops onto a precleaned microscope glass slide (76 × 26 mm2) and thinly spread over the majority of its area. Immediately after, a heat gun (T ∼120 °C) was swept in one direction over the wet coat to rapidly immobilize the particles. After drying, each coated slide was flushed under tap water, then deionized water, and then dried. Salt Solutions and NAPL. The aqueous salt solutions were made from analytical grade NaCl and CaCl2 and degassed under vacuum for 15 min prior to any pH adjustment. All solutions were prepared at their unadjusted natural pH, in the range 5.5-6.1. Solutions without CaCl2 were also prepared in acidic (pH 4 ( 0.2) and alkaline (pH 9 ( 0.2) states, by adding HCl and NaOH. The NAPL is an asphaltic crude oil from the Minnelusa field in Wyoming. It has density 0.9062 g cm-3, viscosity 77.2 mPa.s, n-C7 asphaltene content 9.0 wt %, and acid and base numbers 0.17 and 2.29 mg KOH/g oil,
all at room temperature (22). Zeta potential of the fine kaolinite suspension and the emulsified NAPL were separately determined in each salt solution using a Zetasizer Nano-ZS (Malvern Instruments), based on standard procedures (23, 1). Results are summarized in Table 1 and plotted in Supporting Information Figures S1 and S2. Microscopy of Substrates. Kaolinite substrates, lightly sputter coated with platinum (and never reused), were imaged with a field emission scanning electron microscope (FESEM, Zeiss UltraPlus Analytical) in secondary electron mode at 1 kV. For AFM, a Nanoscope III (Digital Instruments) with high sensitivity Super A scanner was used in tapping mode at 0.89 Hz. The cantilever (Tap300Al, BudgetSensors) had resonant frequency 300 kHz and force constant 40 N/m. For height maps over much larger areas, a white-light profilometer (Wyko NT9000, Veeco) was operated with 20× microscope objective, 2.0× zoom, and numerical aperture of 0.4. Substrate pieces were mapped over 157 × 118 µm2 areas, at 0.25 µm/ pixel resolution, using vertical scanning interferometry, averaging 10 frames to reconstruct each map. Contact Angle Measurements. Captive pendant drop measurements of receding and advancing angle at the NAPL/ water/kaolinite contact were performed at 23-24 °C with a contact angle goniometer (KSV Instruments), using a rectangular prism fluid cell (Hellma OG). A piece of kaolinitecoated glass, ∼10 × 10 mm2, was mounted on a chuck, immersed in the salt solution-filled fluid cell for 1 h to equilibrate, and affixed to the fluid cell lid with coated side downward. An oil drop was pumped out upward at rate 0.8 µL/s from the stainless steel hooked syringe (Hamilton; outer and inner diameter 0.65 and 0.31 mm) and left for 10 min at the tip to equilibrate with the aqueous phase. The drop was contacted with the kaolinite layer and grown slightly more (to ∼4 µL), directly after which the drop profile in this water-receded state was imaged. Instrument software fits this profile to the Young-Laplace equation to determine the slope at the contact line, with angle taken through the water. The drop was left for 30 min to give oil the opportunity to partially adhere to kaolinite and then retracted at the same rate for the water-advancing contact angle to be similarly VOL. 44, NO. 24, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
9471
FIGURE 1. Images of the kaolinite coat, from a) AFM height mapping and b) corresponding phase mapping, c) FESEM (all with scale bars 500 nm), and d) white-light profilometry (local height in nm). measured. The procedure was duplicated for a second drop on a clean area of the piece. Averages and standard deviations are summarized in Table 1.
Results and Discussion Analysis of Kaolinite Substrates. Figure 1 displays representative high-resolution images, from a-b) AFM and c) FESEM, of the surface of a kaolinite-coated glass substrate. Kaolinite particles occur as single plates and stacks of several of these (with larger plates often overlain with smaller fragments), remaining hydrogen-bonded after sonication. From measuring 250 particles in FESEM images, their broadest dimension has number average of 280 nm, with standard deviation 150 nm. Although particle orientation is predominantly parallel to the substrate, many are tilted, some vertically to present only their edge. As kaolinite is a nonswelling clay, all coats possess some porosity, with pore throat sizes typically below 100 nm. White-light profilometry (Figure 1d) shows a random speckling of somewhat higher and lower subregions, corresponding to locally greater and lesser coat thickness. Overall surface roughness derives from these longer-range variations and finer-scale features of individual particles and their packing. Three such areas on each of four pieces were analyzed, giving an arithmetic-mean absolute roughness Ra averaging 137 nm with standard deviation 14 nm. From gravimetric measurements, the average wet layer thickness is ∼40 µm, corresponding to just over 1 µm for the dry coat. Wet preparation procedures in the literature give coats of around 20 µm thickness (10, 12, 24). Our approach attains much lower thickness without sacrificing coverage by combining a high quality suspension with rapid im9472
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 24, 2010
mobilization postapplication. The size fractionation, sonication and raised pH to render all kaolinite faces and edges strongly anionic, yields a very stable, fine suspension. This facilitates metering of a thin wet layer over the glass, and the comparatively high concentration and unidirectional drying at elevated temperature serve to quickly lock-in the structure. This minimizes nonuniformities from convective transport toward more rapidly evaporating regions (producing stick-slip bands of high-low deposition) or premature immobilization in the vicinity of larger particles and aggregates (rupturing the wet film) (25). The thin coats maximize smoothness and minimize internal stresses on liquid exposure. Incomplete coverings of kaolinite on smooth glass and alumina substrates were used to favor outward exposure of the silica and alumina faces, respectively, for force measurement (11). In our case, the high pH and concentration, and rapid immobilization, probably lead to a mixture of the two faces at the surface. Contact Angle. If the aqueous film does not rupture, the NAPL contacts the chemically homogeneous, rough surface of the water-filled and -sheathed kaolinite coat, and a Wenzeltype wetting pertains. If local rupture and adsorption occur, so NAPL contacts the chemically heterogeneous, rough surface of oil-wet kaolinite outer surfaces interspersed with residual bulk water menisci between particles, a Cassie-type model pertains. Both situations, and intermediates, occur in the results below; examples of low and high contact angle hysteresis are shown in Figure 2a-d. In addition to any wettability alteration, substrate topography also contributes to hysteresis, although usually comparatively little for roughness below 100 nm (26). We observe no stick-slip behavior on drop growth, in line with Shang et al. (10), and for salt solutions allowing little oil-kaolinite adhesion, the measured
FIGURE 2. a-d) Images during contact angle measurement of an oil drop on kaolinite in NaCl solutions of 1 M at pH 9 (after a: growth, b: retraction) and 0.005 M at pH 4 (after c: growth, d: retraction), giving average angle hysteresis of 7° (nonadhesive) and 88° (adhesive). e-f) FESEM micrographs, with scale bars 500 nm, of the same kaolinite coat area e) before, and f) after, a contact angle experiment in water at pH 4. After the experiment, the piece was immersed first in decahydronaphthalene (98%, Sigma-Aldrich), to dissolve nonadsorbed bulk oil, and then in n-heptane (99%, Riedel-de Haen), so only the asphaltene-based adsorbate remains. Both states e-f were imaged without a conducting coat. A higher resolution image of this “after” state is given in Supporting Information Figure S3.
FIGURE 3. Effect of NaCl concentration and acidity, for the three cases of pH 4, natural (n) and 9, on oil drop contact angle hysteresis on kaolinite. hysteresis approaches zero. The kaolinite particles remain immobile during drop growth and retraction, as evidenced by registered FESEM images of identical locations on the coat before and after the contact angle experiment (Figure 2e-f, for the sample with largest hysteresis in Table 1). Moreover, the contact angle hysteresis bears no overall correlation to the roughness of the individual kaolinite piece, determined after the contact angle experiment, but equally applicable to the “before” state due to the particle immobilization. Thus roughness does not contribute significantly, and hysteresis values reflect the intermolecular adhesion of NAPL to kaolinite. Figure 3 plots the contact angle hysteresis for NaCl-only solutions at four salinities, each at the three pH values, in Table 1. The wide range of values and trends in Figure 3 reflect the behavior of the advancing angle; receding angle varies little, with average 36.1° and standard deviation 6.1°. This invariance is expected for forced drop growth, since
oil-water interfacial tension is relatively constant over this matrix of salt solutions for crude oils very similar to ours (14, 27). Hysteresis increases with decreasing pH, somewhat weakly from pH 9 to the natural state, but much more pronounced from natural to pH 4. At pH 4 the oil-kaolinite adhesion increases strongly with decreasing salinity. This is supported by additional experiments at 0.005 M NaCl and pH 4, which gave hysteresis of 88 ( 5°. These four nonzero salinities at pH 4 exhibit a strong negative, linear correlation between hysteresis and logarithm of concentration (Figure 6). Hysteresis apparently reaches a plateau around 0.005 M, as its value is similar to pure water at pH 4 in Figure 3. At natural pH, oil-kaolinite adhesion, albeit weaker, also increases linearly with decrease in the logarithm of NaCl concentration from 1 M (Figure 6) but reaches its plateau earlier, around 0.01 M. In turn, for pH 9, maximum hysteresis occurs already just below 0.1 M in Figure 3, with salinity decrease below 0.01 M resulting in an adhesion reduction. The angles in Figure 3 are replotted in the matrix of Figure 4, together with annotations as to whether retraction of the oil drop leaves its bulk adhering to the kaolinite surface due to snap-off (adhesion, A), or results in clean removal (nonadhesion, N), or gives the intermediate situation in which the bulk is removed while small droplet(s) remain (transition, T). Compositions giving complete or partial adhesion are shaded. Generally, high hysteresis is associated with adhesion and low hysteresis characterizes nonadhesion; however, the correlation is not exact, with lower pH tending to give adhering droplet residues at comparable hysteresis values for which pH 9 results in complete removal. This classification allows comparison with literature adhesion maps on smooth mineral substrates, for which hysteresis values are not reported (1, 6, 13, 15, 19, 20). While no study used Minnelusa crude, A-93 Prudhoe Bay crude has similar properties (13, 15), and the boundaries of its maps on glass (13) and mica (15) VOL. 44, NO. 24, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
9473
FIGURE 4. Adhesion map for Minnelusa crude on kaolinite versus NaCl concentration and pH, with A denoting adhesion, N nonadhesion, and T transition state. Boundaries, above and right of which nonadhesion occurs, are compared to literature systems. at 25 °C are given in Figure 4. Some differences exist between these and our protocols; the former use pH buffers and only a 2 min drop-substrate contact time. All three systems, and the vast majority of other crude oil-brine-substrate systems (1, 13-20), exhibit strong adhesion for the combination of low pH and NaCl salinity and no adhesion at the opposite extreme of high pH and salinity. Distinctions reduce to the more precise location of their boundary and transition states. Kaolinite and mica behave similarly, while glass shows stronger adhesion and a smaller region of nonadhesion in Figure 4. The charges of the oil-water and water-kaolinite interfaces in Table 1 explain, within the DLVO framework, most of the adhesion trends in Figure 3, especially for lower salinity. At pH 4, polar oil components at the water interface are substantially protonated; Table 1 is consistent with a reported isoelectric point (IEP) at pH 4.1 for Minnelusa crude (27). The net charge on the oil-water interface is thus positive (10-4 M NaCl) or only very weakly negative (>0.01 M NaCl) with a considerable population of positively charged domains (19). Kaolinite is net negatively charged; Table 1 is in line with literature IEP values at pH 2-3 (23). The resulting electrostatic attraction is only slightly screened at low salt, and together with van der Waals attraction, leads to rupture of the water thin film and oil adhesion to substrate. Under these conditions, Figure 2f directly evidence the alteration of uppermost kaolinite surfaces toward oil-wetness. On pH increase from 4 to 6, kaolinite becomes somewhat more negatively charged, and the oil-water interface becomes substantially more anionic (Table 1), switching their DLVO electrostatics toward repulsion and lessening adhesion. Further pH rise to 9 increases only the magnitude of these large negative charges (Table 1), resulting in a smaller decrease in hysteresis in Figures 3 and 4. Increasing salinity shrinks the range of electrostatic interaction, i.e. the Debye screening length, and thus zeta potential magnitudes decrease in Table 1. At pH 4, the increased screening of locally positively charged NAPL domains from net negatively charged kaolinite reduces hysteresis, although further adhesion may occur over much longer times. The weaker hysteresis at pH 6 and 9 also exhibits this downward trend in Figure 3, despite the increasing salinity decreasing the DLVO net repulsion. One possible explanation is the onset of non-DLVO hydration repulsion (18) above 0.1 M. Alternatively, it could be speculated that the increasing overlap of electric double layers at lower salt provides the scope to induce attraction. One such mechanism is protonation of the NAPL interface by the H+ counterions 9474
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 24, 2010
FIGURE 5. Effect of NaCl and CaCl2 concentrations on hysteresis at natural pH, with A denoting adhesion, N: nonadhesion, T: small droplet(s) remain, and t: tiny droplet(s) remain. of the acid-dissociated mineral surface (1, 19). A second mechanism could involve repulsion-induced reorientation of surface NAPL groups to expose polyaromatic asphaltene sheets, which may bind Na+ ions and attract the mineral substrate. For kaolinite, the variety of charges on its faces and edges offers a third mechanism. Despite its low IEP, the amphoteric broken Si-O and Al-O bonds at edges remain negatively charged up to neutral pH (23), and some evidence points to alumina faces acting similarly (11). At natural pH, these edges and faces could attract deprotonated groups at the NAPL interface, and silica faces could attract the minority of protonated NAPL groups, to provide local adhesion. The presence of divalent ions in the aqueous phase is rarely addressed in the NAPL-mineral contact angle literature, as effects are ion- and NAPL-specific, and equilibration times can be lengthy (15, 20). However, they are an important reality in most environments. Table 1 shows the effect of calcium ions on the zeta potential of the two separate interfaces at natural pH. For both kaolinite and oil, Ca2+ reduces the zeta potential negativity via the combined effects of increased electrostatic screening by dissolved ions and ion binding at the interface. While CaCl2 concentrations of 0.1-1 M are required to reverse the net charge to positive, Ca2+ binding is still substantial at lower concentrations of 0.001-0.01 M, especially for kaolinite (in line with other studies (23)). This is apparent in Table 1 from the increase in zeta potential magnitude with rising NaCl, as the more weakly binding Na+ displaces more strongly binding Ca2+ to reassert negativity. Figure 5 plots hysteresis in Table 1 for solutions of NaCl and/or CaCl2 at natural pH. The receding angles again show little variation, with average 35.7° and standard deviation 4.3°. The letter beside each bar classifies the observed adhesion, now using the subcategory “t” between T and N to denote substrate retention of only a tiny oil droplet(s). The classification correlates well with hysteresis in Figure 5; the hierarchy A, T, t, and N has average angles 58.9°, 26.5°, 11.1°, and 6.7°, emphasizing that hysteresis provides a reasonably reproducible, indirect measure of adhesion. The most striking feature is the strong adhesion for the CaCl2only solutions at 0.001 and 0.01 M in Table 1, despite both interfaces being net negatively charged. From the zeta potential analysis, this appears to be due to attraction of anionic NAPL groups to kaolinite-bound Ca2+, and to a lesser extent attraction of anionic kaolinite to NAPL-bound Ca2+, combined with Ca2+ bridging of anionic sites on both interfaces. For 0.1 M CaCl2, the two interfaces have opposite net charges (Table 1), but due to their strong screening, adhesion is virtually nonexistent (although may develop over much longer times). The window of enhanced adhesion, within which multivalent cations can partially bind to one or both interfaces, or bridge them, at sufficiently low concentration for long-range attraction, is expected to be
FIGURE 6. Contact angle hysteresis versus ionic strength, for all 16 salt solutions at natural pH, and the four NaCl solutions at pH 4. quite general, although naturally dependent on the cation exchange and binding properties of the particular mineral substrate and NAPL (15). The trend to decreasing hysteresis with increasing ionic strength again generally applies for CaCl2 and its mixtures with NaCl. Figure 6 replots all data in Figure 5 and compares the NaCl-only subset with its counterparts at pH 4. At natural pH, all salt solutions with ionic strength above 0.1 M give hysteresis less than 17°, while all below 0.1 M give angles greater than 20°. The pH 4 line is much steeper than for natural pH, reinforcing that H+ is a potential determining ion, especially for the oil-water interface (Table 1). The two highest points for natural pH in Figure 6 (with hysteresis ∼60°) are the above-mentioned Ca2+ pair, also potentialdetermining, especially for the kaolinite-water interface (Table 1). Strong adhesion of NAPL to kaolinite in aqueous systems, at least over relatively short times, thus demands weakly screened electrostatics and a sufficiently high density of oppositely charged, or chargeable, domains on the two interfaces and would be qualitatively consistent with DLVO modeling properly incorporating acid/base charge regulation and ion binding.
Acknowledgments This study was funded by an ARC Discovery Grant (A.F.) and the member companies Total and Maersk of the Digital Core Consortium Wettability Satellite.
Supporting Information Available Elemental analysis of kaolinite (Table S1), and graphs of kaolinite and oil zeta potential, FESEM close-up of a kaolinite coat, comparison of coat roughness to hysteresis, graphs of receding and advancing angles (Figures S1-S5). This material is available free of charge via the Internet at http:// pubs.acs.org.
Literature Cited (1) Zheng, J.; Behrens, S. H.; Borkovec, M.; Powers, S. E. Predicting the wettability of quartz surfaces exposed to dense nonaqueous phase liquids. Environ. Sci. Technol. 2001, 35, 2207–2213. (2) Karimi-Lotfabad, S.; Gray, M. R. Characterization of contaminated soils using confocal laser scanning microscopy and cryogenic-scanning electron microscopy. Environ. Sci. Technol. 2000, 34, 3408–3414. (3) Durand, C.; Rosenberg, E. Fluid distribution in kaolinite- or Illite-bearing cores: cryo-SEM observations versus bulk measurements. J. Pet. Sci. Eng. 1998, 19, 65–72.
(4) Auerbach, S. M.; Carrado, K. A.; Dutta, P. K. Handbook of Layered Materials, 1st ed.; CRC Press: Boca Raton, 2004. (5) Bear, F. E. Chemistry of the Soil, 2nd ed.; Reinhold Publishing: New York, 1965. (6) Morrow, N. R. Wettability and its effect on oil recovery. J. Pet. Technol. 1990, 42 (12), 1476–1484. (7) Peters, C. A.; Luthy, R. G. Coal-tar dissolution in water-miscible solvents - Experimental evaluation. Environ. Sci. Technol. 1993, 27, 2831–2843. (8) Nandi, B. K.; Uppaluri, R.; Purkait, M. K. Treatment of oily waste water using low-cost ceramic membrane: Flux decline mechanism and economic feasibility. Sep. Sci. Technol. 2009, 44 (12), 2840–2869. (9) Warr, L. N.; Perdrial, J. N.; Lett, M. C.; Heinrich-Salmeron, A.; Khodja, M. Clay mineral-enhanced bioremediation of marine oil pollution. Appl. Clay Sci. 2009, 46 (4), 337–345. (10) Shang, J.; Flury, M.; Harsh, J. B.; Zollars, L. R. Comparison of different methods to measure contact angles of soil colloids. J. Colloid Interface Sci. 2008, 328 (2), 299–307. (11) Gupta, V.; Miller, J. D. Surface force measurements at the basal planes of ordered kaolinite particles. J. Colloid Interface Sci. 2010, 344 (2), 362–371. (12) Wu, W. Baseline studies of The Clay Minerals Society source clays: Colloid and surface phenomena. Clays Clay Miner. 2001, 49, 446–452. (13) Buckley, J. S.; Liu, Y.; Xie, X.; Morrow, N. R. Asphaltenes and crude oil wetting - the effect of oil composition. SPE J. 1997, 2 (35366), 107–119. (14) Xie, X.; Morrow, N. R.; Buckley, J. S. Crude oil/brine contact angles on quartz glass. Proceedings of the International Symposium of the Society of Core Analysts, Calgary, Canada, September, 1997. (15) Liu, L.; Buckley, J. S. Alteration of wetting of mica surfaces. J. Pet. Sci. Eng. 1999, 24, 75–83. (16) Yang, S.-Y.; Hirasaki, G. J.; Basu, S.; Vaidya, R. Mechanisms for contact angle hysteresis and advancing contact angles. J. Pet. Sci. Eng. 1999, 24, 63–73. (17) Drummond, C.; Israelachvili, J. Fundamental studies of crude oil-surface water interactions and its relationship to reservoir wettability. J. Pet. Sci. Eng. 2004, 45 (1-2), 61–81. (18) Basu, S.; Sharma, M. M. Measurement of critical disjoining pressure for dewetting of solid surfaces. J. Colloid Interface Sci. 1996, 181 (2), 443–455. (19) Buckley, J. S.; Takamura, K.; Morrow, N. R. Influence of electrical surface charges on the wetting properties of crude oils. SPE Res. Eng. 1989, 332–340. (20) Buckley, J. S.; Morrow, N R. An overview of crude oil adhesion phenomena. In Physical chemistry of colloids and interfaces in ´ ditions oil production; Toulhoat, H., Lecourtier, J., Eds.; E Technip: Paris, 1992. (21) Chipera, S. J.; Bish, D. L. Baseline studies of The Clay Minerals Society source clays: Powder x-ray diffraction analyses. Clays Clay Miner. 2001, 49, 398–409. (22) Tie, H. G.; Tong, Z. X.; Morrow, N. R. The effect of different crude oil/brine/rock combinations on wettability through spontaneous imbibition. Proceedings of the International Symposium of the Society of Core Analysts, Pau, France, September, 2003. (23) Alkan, M.; Demirbas, O.; Dogan, M. Electrokinetic properties of kaolinite in mono- and multivalent electrolyte solutions. Microporous Mesoporous Mater. 2005, 83, 51–59. (24) Shang, J.; Flury, M.; Harsh, J. B.; Zollars, L. R. Contact angles of aluminosilicate clays as affected by relative humidity and exchangeable cations. Colloid Surf., A 2010, 353, 1–9. (25) Denkov, N.; Velev, O.; Kralchevski, P.; Ivanov, I.; Yoshimura, H.; Nagayama, K. Mechanism of formation of two-dimensional crystals from latex particles on substrates. Langmuir 1992, 8 (12), 3183–3190. (26) Extrand, C. W. Contact angles and their hysteresis as a measure of liquid-solid adhesion. Langmuir 2004, 20, 4017–4021. (27) Buckley, J. S.; Morrow, N. R. Wettability and imbibition: Microscopic distribution of wetting and its consequences at the core and field scales; Final report, US DOE contract DEAC26-99BC15204; 2003; p 184.
ES102041B
VOL. 44, NO. 24, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
9475
