Characterization of the Wetting and Dewetting Behavior of Powders †

Experimental procedures for characterizing the wetting behavior of powders are reviewed. The fundamental processes involved in wetting-passage from one state of two-phase equilibrium (solid/gas and liquid/gas) through a three-phase condition (solid/liquid/gas) to a second two-phase condition (solid/liquid and liquid/gas)-are evaluated. A brief discussion of the use of chemical agents such as surfactants to control wetting/dewetting behavior is also included. Characterization procedures, ranging from direct measurement of three-phase contact angles to indirect measures based on observation of the behavior of particles at liquid/gas interfaces are described. For large pieces of solid, contact angles can be determined by direct observation of liquid drops or gas bubbles in contact with polished surfaces. The problems associated with applications of this approach to packed beds of powder are discussed. A procedure for estimating apparent contact angles from the relative partitioning of small particles across an interface is presented. Estimation procedures based on static and dynamic measurements of liquid penetration into powder beds are evaluated. While indirect methods do not generally provide values of well-defined quantities such as contact angles, when appropriately selected they can yield quantitative information directly relevant to practical applications. Various indirect methods including film flotation, the Hallimond tube, bubble pick-up, induction time, immersion/sink time, imbibition time and wetting rate are described. Investigation of wetting phenomena at the molecular scale using techniques such as atomic force microscopy is discussed.


Introduction
Wetting phenomena are of critical importance in practical applications involving mixtures of powders and liquids.Wetting can be defined as the process of displacement at a solid surface of one fluid by another.In general, the fluids may be a liquid and a gas or two immiscible liquids, though the term is most commonly applied to the displacement of a gas by a liquid.In the case of powders, most applications are directed at the preparation of a suspension of the particles in a liquid or coating the particles with liquid in a paste.The use of liquids for dust control is another important application.The reverse process −displacement of liquid by gas−is important in applications such as filtration where gas displacement is used to reduce the liquid content of filter cakes and froth flotation where solid-solid separations are achieved following selective attachment of particles to gas bubbles.Wetting behavior is essentially determined by interaction forces at the different solid-fluid and fluid-fluid interfaces.However, the approach to an equilibrium configuration of the phases can be strongly affected by the viscosity of the fluids.In fact, such viscous effects can make the final approach to equilibrium exceedingly slow leading to hysteresis phenomena.Surface modifications of the solid−for preparing socalled "wettable powders"−or the addition of surface active agents to the liquid are widely used to enhance wettability.Characterization of wetting behavior is an impor tant component of these applications.In principle, wetting behavior could be predicted from a knowledge of the interfacial tensions involved.Unfortunately, however, solid-fluid interfacial tensions are not amenable to simple measurement.Theoretical treatments [1][2][3][4] based on the intermolecular forces involved in wetting phenomena have provided some predictive capability, especially when the interactions are dominated by London-dispersion types of forces. Esentially, this approach involves breaking down the interfacial tensions according to the contribution of each type of force acting−polar (e.g., acidbase), non-polar, etc. Prediction of wetting behavior requires knowledge of these contributions for each of the phases involved.The lack of such information seriously limits the general application of this approach at this time.Consequently, it is usually necessar y to apply less direct methods−including pure empiricism−to characterize wetting behavior.Our objective in this article is to present a critical review of the various techniques available for evaluating the wetting behavior of powders.

Background
Three distinct stages in the wetting of a solid surface can be defined 5) .The first stage, adhesional wetting, refers to the establishment of a three-phase contact at the solid surface.Spreading wetting involves displacement of one fluid by the other at the solid surface.Finally, immersional wetting represents the complete transfer of a solid particle from one fluid phase to the other.Obviously, the three stages must proceed sequentially and, depending on the various interaction forces, may be interrupted after any stage.In the case of porous media such as a bed of powder, capillary forces, due to curvature of the interface between the fluid phases, also play a role, either enhancing or impeding the spreading phenomenon.This type can be defined as capillary wetting.The process of complete wetting represents passage from a condition of two-phase equilibrium (solid-gas and liquid-gas) through a state of threephase contact (solid-liquid-gas) to a final stage of twophase equilibrium (solid-liquid and liquid-gas).If conditions favor spreading, wetting proceeds spontaneously to the fully immersed state.Otherwise a final state of partial wetting−three-phase equilibrium−is reached.Thermodynamically, the free energy changes involved in the different stages can be expressed as fol-lows: where ΔGA, ΔGS and ΔGI are the free energy changes (per unit area) corresponding to adhesion, spreading and immersion respectively.γSL, γSV and γLV are the respective solid-liquid, solid-vapor and liquid-vapor interfacial tensions.Capillar y wetting involves changes in the solid-liquid and solid-gas interfaces only and is, therefore, a special case of immersional wetting and the same thermodynamic relationship applies at equilibrium.It should be noted, however that capillary forces can significantly affect the approach to equilibrium, i.e., the kinetics of wetting.
The thermodynamic relationships can also be expressed in terms of the work of adhesion WA and the spreading coefficient SLS.The work of adhesion is defined as the work required (per unit area) to separate the solid from the liquid so that, WA = -ΔGA.Similarly, the spreading coefficient is defined such that SLS = -ΔGS.The conditions corresponding to three-phase equilibrium (partial wetting) can be conveniently characterized by means of the contact angle θ defined, through the liquid phase, as shown in Fig. 1.
Thermodynamically, the three-phase equilibrium can be expressed by Young's equation: By substitution from Equation 4, the free energies given by Equations 1-3 can also be expressed in terms of the contact angle.Thus, ∆G S = γ LV (1 − cos θ) (6) The variation of these free energy changes with contact angle is illustrated in Fig. 2. It should be emphasized that Equations 4-7 are valid only when the three-phase contact is stable, i.e., when Other wise, there is no contact angle, spreading is spontaneous and the equilibrium condition corresponds to two-phase contact.A zero contact angle describes a limiting condition such that Similar arguments apply to contact angles of 180 .While 180 angles do not occur in solid-liquid-gas systems, the full range is possible for solid-liquidliquid contact.When the contact angle is finite (0 ≤ θ ≤ 180 ), the two phase contact is unstable in the presence of the third phase−liquid droplets or gas bubbles adhere to the solid surface.
Spreading is a prerequisite for immersional wetting, and ΔGS can be regarded as an activation energy for wetting by immersion.For contact angles less than 90 , transfer of a solid particle from the gas to a liquid phase (immersion) leads to a lowering of the free energy and is, therefore, favored.However, in order to reach the immersed state, additional energy must be supplied to overcome the increase in free energy associated with spreading.The three-phase equilibrium resulting from adhesion represents the most stable configuration (minimum free energy) in all cases.As a result, spreading is the critical stage in the overall wetting process.Zisman 6) showed, empirically, that plots of the cosine of the contact angle for a series of liquids on a given solid surface against the liquid surface tension typically showed a linear relationship.Extrapolation of the plots to zero contact angle (cosθ = 1) yields the value of the critical surface tension γc which is characteristic of the solid.The contact angle becomes zero when γLG = γc.Liquids whose surface tension is less than γc wet the solid completely and the threephase contact is no longer stable.It follows that γc represents the limit of applicability of Young's equation and relationships such as Equations 5 -7 derived from it.An example of a Zisman-type plot for a set of liquids on a bituminous coal surface is given in Fig. 3. Extrapolation of the quantity (γSV -γSL)/γLV to values greater than unity is, of course, arbitrary.
Substantial control over wetting behavior can be obtained through the addition of surface-active agents to the liquid phase.So-called wetting agents enhance spreading by reducing the liquid-gas or solid-liquid interfacial tensions (or both).On the other hand, the collectors used in mineral separations by froth flotation promote bubble attachment to the solid surface by reducing the solid-liquid interfacial tension through adsorption.Wetting behavior can also be controlled by appropriate chemical or physical modification of the solid surface either to enhance or impede spreading by the liquid.Mineral flotation is a process whereby one mineral is separated from another by an organic reagent that adsorbs on the selected mineral to make it hydrophobic.The attachment of an air bubble to a solid immersed in water is governed by relations similar to Equations 2-4 and 5-7.The free energy change, ΔGflot, for the attachment of an air bubble to a solid is simply ∆G which means that any time the contact angle is greater than zero, the bubble should attach.
The development of suitable reagents or surface treatments for practical applications to the control of wetting or dewetting behavior requires reliable, quantitative procedures for evaluating wetting characteristics.

Characterization of Wetting/Dewetting Behavior
Techniques for evaluating wetting characteristics range from direct measurement of contact angles to indirect measures based on observation of the behavior of particles at liquid-gas interfaces.

Direct obser vation
The classical technique for determining contact angles is direct observation of a sessile drop of liquid on a highly polished, clean surface of the solid.Problems of simultaneously focussing on the solid surface and that of the liquid drop can be avoided by measuring the (double) angle between the surface of the actual drop and that of its reflection in the polished surface.Hysteresis effects, whereby the measured angle varies according to whether the liquid is advancing or receding at the surface, are commonly encountered.The causes of these effects are not fully understood but probably include pinning of the threephase boundary at surface inhomogeneities, natural imperfections or induced patterns, or contamination.Viscous effects due structuring in the liquid adjacent to the interface may also contribute.
An alternative approach (restricted to transparent liquids) is to immerse the solid in the liquid and observe the angle in a captive bubble brought into contact with the solid.The use of sessile drops is generally preferred for measuring liquid-advancing angles while the captive bubble approach offers advantages for liquid-receding angles.If the liquid surface tension is known accurately, uncertainties in direct contact angle measurements can be reduced by comparing observed drop (or bubble) shapes to theoretical profiles calculated using the tables of Bashforth and Adams 8) and Blaisdell 9,10) .Obviously, the direct methods are impractical for measurements on fine powders.Measurements on large pieces of the same material may be inappropriate due to surface modifications that might have occurred during preparation of the powder.In some cases, it is possible to press the powder into a sufficiently coherent pellet for measurements to be made.However, since the use of binders during pellet preparation must be avoided, it is rarely possible to obtain a high polish on the surface and problems arise due to surface roughness and penetration of the liquid into pores.Hysteresis of the measured contact angle is particularly common when powder compacts are used.Nevertheless, this approach is useful for evaluating changes in wetting behavior where absolute values of the contact angle are not required − in studies of the effect of reagent concentration, for example.
Kossen and Heer tjes 11) described a procedure based on the contact angle of a liquid drop resting on the composite (solid + liquid) surface of a powder bed saturated with the liquid.Interpretation of the results involves some questionable assumptions regarding the structure of the porous surface and the general applicability of the method appears not to have been demonstrated.

Interface partitioning
When a small par ticle resides at the inter face between two fluids, its relative immersion in each depends on the contact angle.If gravity can be neglected relative to the surface tension forces, the equilibrium location of a spherical particle at the interface will be as shown in Fig. 4. The interface between the two fluids meets the particle surface at the contact angle, which can be calculated from the fraction of the cross-sectional area that projects into the upper phase.
From the geometry of Fig. 4, the area A1 is given by: where R is the particle radius and θ is the contact angle in radians.The total cross-sectional area A is equal to πR 2 so the relative area is given by Using Equation 10, the contact angle can be calculated from a measured value of the relative area A1/A.The use of simple image-analysis techniques permits measurements to be made very rapidly.This method is especially useful for determining distributions of wettability in collections of particles 12) .The need to minimize the effects of gravity restricts the use of this technique to particles smaller than about 1 mm, depending on relative density.Lower size limits are related to optical considerations.Measurements have been reported on particles as small as about 100 μm 12) .Placing a dry particle at the interface provides a measure of the liquid-advancing angle while prewetted particles give the liquid-receding angle.
Examples of contact angle distributions for spherical polymethylmethacrylate beads and irregular polyethylene and bituminous coal particles are shown in Fig. 5.It can be seen that the homogeneous spheres have a very narrow distribution, probably reflecting minor variations together with experimental error.The homogeneous but irregularly shaped polyethylene particles show a somewhat broader distribution while that for the coal is much broader, presumably due to heterogeneity of the material.

Liquid penetration
While direct measurements of contact angles on fine, micron-size particles are generally impractical, quantitative estimates can be obtained from studies of the penetration of fluid into (or expulsion from) a powder bed.If the pores in a packed bed are regarded as a system of fine capillaries, the capillary pressure in the bed can be described by a form of the Laplace equation: where rp is the effective pore radius.The capillar y pressure can be determined from measurements of the equilibrium capillary rise in the bed, or of the applied pressure needed either to prevent (θ < 90 ) or to cause penetration (θ > 90 ) into the bed 13) .The effective mean pore radius can be estimated from 14) where ε is the porosity of the bed and Sv is the volume specific surface area of the powder.Alternatively, the pore radius can be eliminated by measuring the capillary pressure for a wetting liquid (θ = 0), so that: where γw and Δpw refer to the wetting liquid.
Common to each of these methods is the problem of preparing a uniform bed of particles.This is especially difficult in the case of capillary rise measurement where very thick beds−up to several meters for fine powders−may be required.The use of Equation 13requires a high degree of reproducibility as well as uniformity in bed preparation, since measurement with the wetting liquid must be on a separate bed.Direct estimation of the pore radius from Equation 12is subject to additional errors due to the approximations involved and the sensitivity of the equation to the bed porosity.
The rate of capillary rise can be used as an alternative to the equilibrium height, thereby substantially reducing the required bed depth.In this case, the capillar y pressure is substituted into standard expressions for the rate of flow through a porous bed leading to the Washburn equation 15) .
where h is the height of liquid in the bed after time t, μ is the viscosity of the liquid and Rf is a geometric factor related to the effective mean pore radius in the bed.The slope of a plot of h 2 against time is a measure of the quantity Rf cosθ.Again, it is common to use measurements with a wetting liquid (cosθ = 1) to estimate Rf.It should be noted, however, that the capillary pressure given by Equation 11 is valid at equilibrium or for cases where the contact angle is finite.For wetting liquids, the γLV cosθ term should be replaced by (γSV -γSL}, which is only equal to γLV at the critical surface tension γc. It follows that estimates of rp based on different wetting liquids may vary.Wolfrom et al., 16) have suggested a procedure for avoiding this uncertainty through the use of two or more wetting liquids.Obviously, the problem of preparing uniform beds reproducibly is especially critical in these applications.A standardized procedure for bed preparation has been suggested 16) .It should also be noted that this technique is not suitable for liquids of high viscosity for which bubble entrapment can lead to significant errors and poor reproducibility.
Procedures based on the use of thin layers of particles deposited on a substrate have also been described 17) .Uniformity of the pore structure may be particularly difficult to achieve using this approach.Again, standardization of the specimen preparation procedure and thorough evaluation of reproducibility are necessar y if meaningful results are to be obtained.
The use of short columns of powder is advantageous with respect to bed uniformity and minimization of gravity effects for vertical orientations.However, accurate measurement of penetration depth is more difficult.The latter can also present problems with opaque particles where direct observation of the liquid interface becomes difficult.The use of gravimetric methods with continuous weighing of a particle bed in contact with a fluid reservoir can substantially improve the accuracy of depth measurement.At the same time, however, the results are more sensitive to bed porosity and uniformity.The mass of liquid in the bed is given by where ρ is the liquid density and Ab is the crosssectional area of the bed.Thus, the equivalent to Equation 14 is: where the geometric factor c is given by: which is significantly more sensitive to variations in porosity than Rf.
Examples of capillary rise measurements on beds of glass beads are presented in Fig. 6.The quantity μm 2 /γLVρ 2 is plotted to permit direct comparison of the results, allowing for differences in liquid density and viscosity.It can be seen from the replicate test results that quite good reproducibility can be obtained.
Statistically, the results for hexane and decane (γLG = 18.4 and 23.8 dynes/cm respectively) fall on a single line, from which the constant c can be obtained.However, if the constant is estimated from just one of the set of hydrocarbon tests, contact angle estimates for the other tests can range from zero to as much as 28 , indicating the very high sensitivity of this technique to small errors, particularly for determining low contact angles.On the other hand, the use of a value of the constant obtained from any of the hydrocarbon results gives a much narrower range (66.5−69.3 ) for the contact angle for formamide (γLV = 58 dynes/ cm).
In general, the penetration methods can provide a reasonably quantitative comparison of the wetting of a solid by different liquids, but actual estimates of contact angles should probably be regarded as approximate at best.These methods normally provide an estimate of the liquid-advancing contact angle.However, the capillary pressure approach can also be used, by studying expulsion of liquid from a bed, to investigate liquid-receding angles.

Indirect Measures of Wetting Behavior
Information on wetting characteristics can also be acquired from obser vations of the behavior of particles at liquid-gas interfaces.While these tend to be somewhat qualitative and do not yield values of, for example, contact angles, they can provide useful measures of relative wettability and can be used to estimate quantities such as the Zisman critical surface tension γc.

Film Flotation
In order to devise a method for characterizing the wetting behavior of heterogeneous coal particles, Fuerstenau and Williams 18) developed a novel technique that they called film flotation, wherein a given sample of particles is portioned into hydrophobic and hydrophilic particles using aqueous methanol solutions of different surface tensions, with which the surface tension of the liquid can be varied from 72.8 to 22.5 mN/m.The film flotation technique simply involves placing the test solution in a small vessel 75 mm in diameter and 20 or 30 mm in depth.Having the lower part of the vessel conically shaped, separation of particles that sink is readily achieved by draining some of the liquid.Generally enough particles are added so that a monolayer remains on the surface.In the case of coal, this may range from 0.06 to 0.3 gram, depending on the particle size being tested.
Their initial work was with coal particles, which are by nature heterogeneous.As the surface tension is varied, particles sink under conditions when the contact angle just approaches zero, that is the process is controlled by spreading wetting.This yields surface tension-vs-percent hydrophobic fraction distribution diagrams, from which the distribution and the mean critical wetting tension of the particles can be determined.Careful experimentation showed that virtually identical results are obtained with aqueous solutions of methanol, ethanol, propanol, tertiary butanol, and acetone 18,19) .Since interfacial forces are so dominant, particle size does not have an appreciable effect, as found for 425 × 300-μm, 300 × 212, 212 × 150, 150 × 106, 106 × 75 and minus 75 × 53-μm Cambria #78 coal particles 19) .
Because of the heterogeneity of coal, the wetting behavior of particles in the sample may change continuously from that of hydrophobic organic materials to those of hydrophilic inorganic matter.To illustrate this, Fig. 7 presents the cumulative distribution of hydrophobic Cambria #78 coal particles as a function of their wetting surface tension, as obtained by film flotation with aqueous methanol solutions.From the results given in Fig. 7, the frequency distribution of the critical wetting surface tension of Cambria #78 coal particles can be determined since every point along the curve in Fig. 7 must represent particles for which the contact angle is zero.The frequency distribution plot is presented in Fig. 8.These two figures clearly show the heterogeneous nature of the surface of coal particles.In contrast, ideal homogeneous particles will have the same critical wetting surface ten-sion, that is, the cumulative plot is a vertical line that gives the sharp demarcation between sink and float surface tension 19) .
From such a distribution, four wetting parameters have been defined.The critical wetting surface tension of the most hydrophobic particles in the assembly, γc min , is the surface tension of the liquid at which none of the particles remain (float) at the liquid surface.The critical wetting surface tension of the most hydrophilic par ticles in the powder, γc max , is the surface tension of the liquid at which all particles remain at the liquid surface.The mean critical wetting surface tension of all particles, γ -c , can be calculated from the film flotation frequency distribution using the equation: where γc is the critical wetting surface tension of the Fig. 7 Partitition curve obtained from the film flotation response of Cambria #78 coal as a function of the surface tension of aqueous methanol solutions (data from Jia [29]).
Fig. 8 Frequency histogram for the distribution of the wetting surface tension of 100×150-mesh Cambria #78 coal particles (data from Jia [29]).
particles and f (γc) is the frequency distribution function.The standard deviation of the frequency distribution function, σγc, reflects the heterogeneity of the surface and is given by High values of σγc correspond to more heterogeneous materials.Fig. 7 and 8 show the three wetting surface tensions that can be determined from film flotation experiments.
The critical wetting surface tension is an important parameter that can be used as an index of the wettability of a solid.As defined by Zisman 6) , it is the surface tension of a liquid that just forms a zero contact angle on the solid.Both theoretically and experimentally for film flotation the critical wetting surface tension of hydrophobic particles can be taken to be the surface tension of the liquid at which the particle just sinks into the liquid.The contact angle of those particles that sink at the given surface tension is essentially zero since the effect of particle size and density are sensibly negligible in the practical size range.
Since the contact angle of a small particle cannot unambiguously be determined directly, it is useful to find a way to assess its value from other kinds of experiments.In the approach developed here, it is necessary to know the value of γc for the individual particles 20) .Since any point along a film flotation plot (a plot of the fraction of particles floating versus the liquid surface tension) represents conditions for zero contact angle at a specific surface tension, the distribution of wetting surface tensions obtained from film flotation results is the starting experimental information needed to evaluate contact angles.
The theoretical basis for this begins with the Neumann-Good Equation of State 21) : where γSV, γSV and γSV are the interfacial energies (in erg/cm 2 ) at the solid/vapor, solid/liquid and liquid/vapor interfaces, respectively.Substituting this into the Young Equation: The contact angle θ of a particle for a liquid of a given surface tension can be calculated with Equation 22using the values of γc measured by film flotation.The basis for this is taking γc as being equivalent to γSV , an assumption discussed by Neumann et al 21) .Many materials are nonhomogeneous, probably the most significant being coal.Because coal is a complex mixture of carbonaceous substances and inorganic minerals.the surface of coal will be a patchwork assembly of hydrophobic and hydrophilic areas.
To simplify calculations, only two types of sites will be considered: hydrophobic (paraffin-like) sites with a water contact angle of 105 and hydrophilic sites with a water contact angle of 0 20) .Using the Cassie equation 22) , it is possible to evaluate the fraction of hydrophobic sites (areas of low surface energy) on the surface of coal particles αHB and the fraction of hydrophilic sites (areas of high surface energy) αHL.Under these conditions, and the contact angle of the composite surface can be written as With these assumptions, the surface composition of heterogeneous coal particles can be calculated using Equations 23 and 24 from their water contact angles calculated from film flotation results using Equation 22.
This can be illustrated by conducting film flotation tests with Cambria #78 coal: as-received, wax coated, and oxidized.If the coal is coated with vaporized paraffin wax, the hydrophilic sites on the coal are covered with paraffin, the mean critical wetting surface tension should decrease and the contact angle should approach that of pure paraffin.On the other hand, if the coal is oxidized by heating in air, hydrocarbon sites are converted to hydrophilic oxygen functional groups and the mean critical wetting surface tension increases.
Although the film flotation cur ves are not presented here, Table 1 gives a summary of the wetting parameters determined in this manner, which are as would be expected by the treatments discussed above.
To further verify the approach of using film flotation experiments to determine contact angles of particles, the mean contact angles of particle assemblies were compared with the values reported for direct measurements on flat surfaces of bulk material 20) .Table 2 gives a summary of contact angles of a variety of materials calculated from film flotation data with those measured by captive-bubble and/or sessiledrop methods on flat surfaces of the same material.
From the results given in Table 2, it can be seen that the two sets of values are in quite good agreement with each other, especially for the more homogeneous materials.Film flotation indeed through this approach allows for estimating the distribution of hydrophobic and hydrophilic sites on heterogeneous particles, such as coal, and provides a way of determining the contact angle on particles.

Modified Hallimond Tube
The modified Hallimond tube is a microflotation device designed so that gas bubbles (air or nitrogen) can be introduced at a constant rate into a bed of particles (about one gram of particles in 100 ml of water).Fig. 9 schematically shows the design and construction of this simple device, which allows determining the amount of material floated during a fixed time period under closely controlled operating conditions with high-purity materials 23) .
In addition, an important attribute of this device is that there is no change in the solution composition by froth removal as occurs in normal flotation testing.Initially, particles are at rest or are being slowly stirred on the frit.When the gas flow is started, the hydrophobic particles attach to bubbles, rise to the top of the liquid level, and fall back down into the collection tube when the bubbles break at the surface.
For a given series of experiments, the flotation time is fixed at some constant value.This technique permits ready determination of conditions that can lead to the dewetting or hydrophobization of normally hydrophilic solids.The technique can similarly be used to delineate how changing pH or adding a wetting agent can affect naturally hydrophobic particles, such as graphite, talc and molybdenite.Typically, a fairly narrow particle size is used in Hallimond tube experiments, such as 65 × 100 mesh (150 × 208 μm) particles.
The flotation of quartz will be taken as an example of the use of the modified Hallimond tube to determine factors that control wettability by the addition Fig. 9 Schematic drawing of the modified Hallimond tube which can be used for controlled testing of flotation behavior and study of dewetting phenomena(after Fuerstenau et al., [23]).  of a surfactant.Clean quartz is hydrophilic but can be made hydrophobic by the addition of a cationic surfactant that has a sufficiently long hydrocarbon chain on its molecule.Detailed studies have been carried out on the adsorption of the cationic surfactant dodecylammonium acetate (DDAA) and its effect on zeta potentials, contact angles and Hallimond tube flotation response 24) .Fig. 10 shows how Hallimond tube flotation response of 208 × 295-μm quar tz particles is related to the concentration of DDAA in solution.With increasing additions of DDAA in solution, the positively charged dodecylammonium ions adsorb at the negative surface of quartz.Because of the tendency for hydrocarbon chains to escape from water by associating together, long-chained surfactants form micelles in solution at higher concentrations, called the critical micelle concentration (the CMC).A micelle may be spherical or rod shaped, with the chains inside the micelle and the charged head groups oriented towards the aqueous solution.
Similar association of adsorbed surfactant ions takes place at the solid-water surface at higher adsorption densities, but because the charged head groups are oriented towards the solid surface, the aggregated adsorbed surfactant ions essentially are similar to only half of a micelle, and hence are called hemimicelles 24) .As can be seen in Fig. 10, the adsorption of DDA ions increases sharply at about 10 -4 M because the chains of the aminium ions begin to associate into hemimicelles at the surface, making the solid more hydrophobic.This is reflected not only in an increase in the flotation of the quartz particles but also in the abrupt changes in the other interfacial phenomena given in Fig. 10.
Flotation is a dewetting process governed by inter-facial tensions (Equation 8).In Fig. 10, the results of measurement of contact angles on a polished quartz cr ystal in DDAA solution were not in terms of the contact angle in degrees but its cosine.In the original paper 24) , the results were plotted with contact angle in degrees.Actually, in accordance with Equation 8, flotation response should be related to the contact angle through the quantity (cosθ -1).For correlation purposes, we want a parameter that increases with increasing flotation recover y, which means simply plotting the results in terms of the quantity (1 -cosθ).Although not shown here, there is quite good correlation between the behavior of these two complex three-phase systems, namely that Hallimond tube flotation can be used to assess the effect of surfactants on the wettability of particles and to estimate the magnitude of the contact angle for any flotation recovery if it is known for one flotation experiment.

Oil Flotation
Oil flotation, in which an oil phase is substituted for the gaseous phase, has been extensively investigated as a means for recovering fine particles.An example is the oil flotation of fine, naturally hydrophobic molybdenite powder without the addition of a surfactant.Before flotation can take place, either with oil droplets or air bubbles, the inter vening film between a particle and an oil droplet or an air bubble must thin and rupture.Any electrical double layer repulsion would clearly hinder this process by introducing a kinetically inefficient step.With most materials, pH has a marked effect on the magnitude of the surface charge and surface potentials.To illustrate the effect of pH on the oil flotation of a naturally hydrophobic material, molybdenite powder of 8.0 m 2 /g was dispersed in the aqueous phase using an ultrasonic bath Fig. 10 Correlation among contact angle, adsorption density, flotation response and zeta potential for quartz as a function of the concentration of dodecylammonium acetate at pH 6-7 (after Fuerstenau et al., [24]).
Fig. 11 The effect of pH on the oil flotation of fine naturally hydrophobic molybdenite without the addition of a surfactant.Also plotted is the effect of pH on the oil/water contact angle of molybdenite, expressed in terms of the flotation dewetting relation, (1 -cos θ), (adapted from Wie and Fuerstenau [25]).
with the pH adjusted to the desired value 25) .After adding iso-octane as the oil phase and stirring the suspension for a conditioning period of five minutes, the mixture is placed in a small cell where the aqueous phase and organic phase are allowed to separate and the two liquids collected separately in beakers.
The amount of molybdenite in the organic and aqueous phases can be determined simply by drying and weighing the samples.The recovery of molybdenite by oil flotation as a function of pH is given in Fig. 11, which shows that oil flotation recovery is nearly 100 percent at pHs below approximately 4 or 5.In alkaline solutions, the recovery decreases, due to the more negative zeta potentials on both the molybdenite and the oil droplets.
Here again, oil flotation can be used to follow how conditions change the oil-water-solid contact angles, which were measured on freshly cleaved faces of a molybdenite cr ystal.The angle of contact between a free iso-octane droplet was measured across the water phase with a microscope-goniometer.At pH 4, the contact angle was determined to be about 150 degrees and in alkaline solutions it decreased to 100 degrees.Since this process involves the attachment of a solid particle immersed in an aqueous phase to an oil droplet, the process is again governed by interfacial tensions in accordance with Equation 8.For correlation purposes, we want a parameter that increases with increasing flotation recover y, which means simply plotting the results in terms of the quantity (1 -cosθ).The results given in Fig. 11 show very close correlation between the contact angle of oil and the oil flotation recovery.For this type of system, if the contact angle is known for one condition, then it should be possible to estimate the contact angle for any other recovery.

Liquid-Liquid Extraction of Nano-Particles
How the addition of a surfactant controls the wettability of nano-sized particles can be assessed by determining the distribution of particles between water and hexane.Fuerstenau and Colic 26) investigated the wettability of 100-nm particles of anatase and hematite using this technique.First, the oxide particles are equilibrated with the surfactant at the desired pH in water, then 5 ml of hexane is added to the aqueous dispersion, and the vials agitated for one hour.After phase separation, the oxide content in each phase can be determined either gravimetrically or by a light scattering technique.The results presented here are for monodisperse spherical hematite particles (α -Fe2O3) that were prepared by thermal hydrolysis of acidified solutions of ferric chloride.Their diameter was 120 ± 20 nm, and after thorough washing their point of zero charge (the pzc) was found to occur at pH 9.5.Since the hematite surface is positively charged at pH 3, the surfactant needed to control wettability under these conditions must be anionic, in this case sodium dodecylsulfate (SDDS).In Fig. 12, the percentage of hematite extracted into the hexane phase is plotted as a function of the concentration of SDDS in the aqueous solution.This plot shows that the conditions for maximum hydrophobicity of these nanosize particles is quite narrow.In Fig. 12, the electrophoretic mobility of these same particles is plotted as a function of SDDS concentration at pH 3 (here mobilities have not been converted to zeta potentials).Both of these plots can be explained in terms of dodecylsulfate adsorption.
At pH 3, the ionic strength is millimolar, so adding SDDS at 10 -5 M does not change ionic strength and the electrophoretic mobility (which is directly related to the zeta potential) remains constant while adsorption is controlled only by the attraction of an anion by the positive oxide surface in exchange for nitrate ions (from the acid) in the double layer.However, when the hydrocarbon chains of adsorbed surfactant ions begin to associate at the surface, the zeta potential (mobility) decreases sharply.Under these conditions, the ionic heads are oriented to the surface and the hydrocarbon chains out towards the solution, causing the particles to become increasingly hydrophobic in this region.With further increase in the SDDS con-Fig.12 The effect of sodium dodecylsulfate on the wettability of hematite particles at pH 3 and 0.001 M sodium nitrate, expressed in terms of the percentage of hematite particles transferred into hexane from the aqueous phase.Also plotted is the electrophoretic mobility of the particles as a function of the surfactant concentration under the same conditions, showing maximum hydrophobicity when the mobility is zero (adapted from Fuerstenau and Colic [26]).
centration, the zeta potential becomes zero when the number of adsorbed ions exactly equals the positively charged surface sites.Maximum hydrophobicity occurs when the mobility or zeta potential and the net interfacial charge are all zero.With increased addition of the SDDS, continued adsorption of the surfactant ions takes place in reverse orientation because the heads of the surfactant ions are now repelled from the negatively interface (as shown by the reversal of the zeta potential).Thus, the particles become hydrophilic again, preferring the aqueous phase over the organic phase, as can be seen by the decrease in extraction into hexane.These experiments show that conditions for maximum hydrophobicity of nanosized particles can be determined simply by finding conditions under which the surfactant brings the zeta potential to zero.To ensure full wetting through formation of an adsorbed bilayer, high concentrations of the surfactant must be added the solution so that a first layer with the charged heads of the surfactant ions being oriented down and a second layer with charged heads being oriented towards the aqueous solution.To accomplish this, the surfactant concentration in solution should be raised above the critical micelle concentration (the CMC) of the given surfactant.

Bubble-Particle Pickup and Induction Times
For bubble-particle attachment the liquid film between the solid and the gas must thin and rupture.
The stability of such films is controlled by electrical double layer forces, attractive van der Waals dispersion forces, repulsive hydration forces, and the existence of attractive hydrophobic forces between particles and air bubbles.Wetting films are not stable on hydrophobic surfaces and hence the attachment between a particle and an air bubble is almost instantaneous.However, partially hydrophobic surfaces interact with water molecules through hydrogen bonding, giving rise to a liquid film that may exhibit some stability over a finite period so that a certain amount of time may be required to achieve the formation of a contact angle.In mineral flotation, the minimum time required for thinning and rupture of the liquid film leading to a stable bubble-particle interface is called the induction time.Perhaps the simplest method devised was the bubble pickup apparatus of Cooke and Digre 27) .Their device was essentially a test tube with a stopper in the top through which a glass tube extends to near the bottom.On top of the glass tube a rubber hose was attached and clamped at the upper end, with the rubber hose acting as an air chamber for generating bubbles.In their investigation, which was concerned with how oleate and calcium salts controlled the wettability of quartz, they used minus 1-mm quartz particles: 0.05 g of wet quartz in 30 ml of aqueous solution.This would also yield a rough measure of induction times, which ranged from about 10 seconds down to 1 second in their system.They found complete pickup, that is when all particles adhered to a bubble that had been pressed against the particle bed and withdrawn, was readily discernible with an induction time less than one second when the contact angle was about 60 degrees in their system.Partial pickup occurred when the contact angle was in the 20-degree range.With this simple device, they were able to correlate the limits of full pickup conditions with solution chemistry reasonably well 27) .The Cooke bubble-particle pickup apparatus provides more or less yes/no information on conditions for mineral-bubble attachment.Various devices have been designed and built over the years to measure the very short induction times that relate directly to flotation kinetics, and these operate similarly to the bubble pickup device but with the contact time being controlled electronically.A fairly simple apparatus in which a captive bubble is moved against a bed of particles for successively longer periods of time until particles become attached to the bubble has been developed by Yoon and Yordon 28) .In their device, the vertical movement of the bubble holder is controlled by a power driver.An electronic pulse generator controls the driver, which allows both the amplitude and time of contact to be set.Contact time in this device can be varied between 30 microseconds and 2 seconds.Typically, a captive bubble of 2 mm diameter is produced on the end of a capillar y tube and the particle bed raised until the particles are about 0.1 mm beneath the bubble.Generally, experiments are carried out with particles fractionated between two consecutive sieves.The bubble is pushed downward by the power driver for the set time and the bubble is observed through a microscope to determine if any particles are attached to the bubble.Fig. 13 is an example of the dependence on solution conditions of the measured induction times for a naturally hydrophobic bituminous (Cambria #78) coal 29) .In this case, the induction time measurements were carried out in 0.002 M sodium nitrate as a function of pH with 100 × 150 mesh (106 × 150 μm) particles with the Yoon-Yordon induction time apparatus 28) .In this Fig. the zeta potential of the coal, as calculated from electrophoretic mobilities, is also plotted as a function of pH.The zeta potential reverses sign at pH 4.7, indicating that the adsorp-tion of hydrogen and hydroxyl ions is responsible for the charging of the surface of the coal particles.At pH 4.7, the double layer on the coal particles is absent.Clearly, maximum hydrophobicity of Cambria #78 coal, as represented by the minimum in induction time, corresponds to conditions where the zeta potential of the coal is zero.At this pH, the electrical double layer does not have any effect on film thinning processes at the interface.
Other work has been published with measurements of induction times on quartz particles as a function of the concentration of dodecylammonium chloride at constant pH 28) .The results show that there is a marked decrease in induction time under conditions where hemimicelles form, that is conditions that lead to greater hydrophobicity of particles, and that the induction time increases sharply as the critical micelle concentration in solution is reached.Under these conditions, the quartz surface is covered with the equivalent of a bilayer of adsorbed aminium ions and is fully hydrophilic.This is important for finding optimal conditions for planarization in electronic wafer processing, for example.

Immersion/Sink Time
The 'sink time' was originally defined by Draves and Clark 30) as the time taken by a skein of cotton to sink after it had been immersed in a surfactant solution.A variation of this test is to measure the concentration of a surfactant required to give a 'sink time' of 25 seconds.The procedure was later modified by Walker et al. 31) to determine the wetting ability of various surfactants for coal dusts.In this technique coal particles were dropped individually onto the surface of surfactant solutions of progressively dilute concentrations.The most dilute concentration in which coal would sink 'instantaneously' was determined.These authors did not define the term 'instantaneously'.Results of such measurement would be sensitive to the operator.Other variations of this technique have also been used.Some investigators 32, 33, and 34) reported the time for wetting a given amount of powder, whereas others have reported wetting rates 35) .Also there are differences in the manner in which time of wetting was measured.Some investigators have determined the time for the last trace of powder to be wetted 32,36) whereas others have determined the time for the vast majority of the particles to sink 34) .The sink time method has been used by investigators to measure the wettability of a wide variety of solids: organic chromium pigment 37,38) flour sulfur and molybdenite 39) .Some difficulties in accurate determination of sink time are likely to occur for powders that are heterogeneous with regard to their wettability.

Wetting Rate
To overcome some of the difficulties mentioned in the previous paragraphs, Mohal and Chander 40) modified the Walker method to develop a procedure for determining the 'wetting rate' of powders.They used the pan of a microbalance placed beneath the surface of the liquid to measure the amount of powder wetted as a function of time.A schematic of the apparatus is given in Fig. 14.A barrier was placed at the surface to direct the particles towards the center of the pan and prevent them from contacting the sides of the pan.A fixed weight of the powder, typically 40 mg, was placed on a spatula attached to a vibro-feeder.
The powder was dropped from a height of 0.5 cm above the barrier within a period of 1 second and the mass of the wetted powder was recorded as a function of time.The rate of wetting was estimated from the slope of the mass vs. time plot as illustrated in Fig. 15.Results were obtained for wetting a sample  of hvA bituminous coal with an aqueous solution of 4 × 10 -4 M Triton X-100 (an ethloxylated octyl phenol).
The authors used the results to demonstrate that differences in the choice of the criterion for wetting can lead to different values for wetting rates.The linear portion of curve A corresponds to the condition in which wetting rate is constant and is defined as the 'initial wetting rate'.If the definition by Kost et al. 34) is used, the rate would be based on wetting of the vast majority of particles, which is schematically shown by line B in Fig. 15.Similarly, if one uses the definition in which the criterion of wetting is the disappearance of the last traces of particles, which is illustrated by line C in the figure, the lowest rate of wetting is measured.The main advantage of the wetting rate approach is that the wetting behavior of heterogeneous systems can be determined unambiguously.
For heterogeneous systems one could define two separate wetting rates if the powder was a mixture of materials with different wettabilities, or one could determine a wetting rate distribution, from slope of the plot if different particles exhibited a range of wettabilities.Mohal and Chander 40) observed several types of wetting rate cur ves that are illustrated in Fig. 16.Different types of wetting curves were obtained depending on the coal type and concentration of the wetting agent.The main disadvantage of this technique is that the particle size for experimentation must be in a narrow range of values.If the particles are too large, gravity will sink the particles, and if they are too small dispersion will prevent settling of particles on the balance pan.

Imbibition Time
Mohal and Chander 41) determined the imbibition time of powders by microscopic observation of particles as they pass from air into liquid.A high-speed motion analysis system was used to photograph the sequence of events during imbibition of particles into a liquid.By analyzing the photographs the imbibition time, the time taken by a particle to be fully imbibed into the liquid was measured, to a resolution of 83 μs (microseconds).A sequence of frames for quartz (hydrophilic) and coal (hydrophobic) is shown in Fig. 17.The quartz particle is wetted by water and it pulls the interface downwards due to hydrodynamic resistance and surface tension as seen in frames 1-7 on the left in Fig. 17.In subsequent frames, 8-12 on the left, the particle has gone into the liquid and the interface relaxes and moves upwards.The hydrophobic coal particle has an initial downward pull due to the kinetic energy of the particle, frames 1-5 on the right, but the particle moves slightly upwards once the kinetic energy is dissipated, frames 6-12 on the right.In this case the coal particle remains floating at the interface.
The imbibition times for quartz and coal samples are presented in Fig. 18 as a frequency distribution.Crushed quartz particles are wetted by water and the measured time corresponded to the time to travel a distance of one diameter under gravity.Wetting of coal by a liquid of lower surface tension (60% metha-nol in water) gave three distinct peaks which were identified with ash-minerals (mainly hydrophilic), a hydrophobic component that gave imbibition times of the order 80 ms, and a middling fraction with an imbibition time of about 3 ms.The results clearly show that this method could be used to determine the distribution of wettability in a given sample.One can consider that the distribution would be 'a true distribution' because the procedure involves observation of one-particle-at-a-time in the sample set.Thus, any distortions due to particle crowding at the interface are eliminated.Some of the techniques described in previous paragraphs suffer from this disadvantage.
Another way to represent the imbibition time distribution is to present the data as a cumulative plot.An example is given in Fig. 19 for wetting of three different coals by a liquid containing 4 × 10 -4 M Triton N-101 (an ethoxylated nonylphenol).In separate wetting tests, wetting rates of 24 mg/s, 21 mg/s and 14 mg/s were measured for the hvA bituminous coal, anthracite, and sub-bituminous coals, respectively.These results confirmed at least a qualitative correlation between imbibition time and wetting rate.

Atomic Force Microscopy
Direct measurement of the forces acting between particles, which is challenging because the forces are weak and the surface separations must be controlled and measured to with 0.l nm, has now become Fig. 17 A sequence of photographic frames (1-12) showing imbibition of quartz (left) and coal particles (right) into liquid water.The quartz particle has a downward pull at the liquid-air interface (frames 1-7 on the left).In subsequent frames (8-12 on the left), the particle has gone into the liquid and the interface relaxes and moves upwards.The hydrophobic coal particle has an initial downward pull due to kinetic energy of the particle (frames 1-5 on the right) but the particle moves slightly upwards once the kinetic energy is dissipated (frames 6-12 on the right) (after Mohal and Chander [41]).
Fig. 18 Frequency distributions of the imbibition times for quartz and coal.The quartz was wetted by water, whereas the hvA bituminous coal was wetted by a mixture of 60% (by volume) methanol in water (after Mohal and Chander [41]).

Fig. 19
Cumulative frequency distributions of the imbibition times for a hvA bituminous coal, anthracite and a sub-bituminous coal.The wetting liquid was an aqueous solution of 4×10 -4 M Triton N-101 (after Mohal and Chander [41]).
quite widely done.The impetus for doing this came from the development of the surface force apparatus (SFA) in the late 1960s by Tabor, Winterton and Israelachvili 42) for measuring van der Waals forces between molecularly smooth mica surfaces in vacuum or in air.Later modifications have permitted making measurements in liquids and with other molecularly smooth materials.The interaction forces between two cur ved surfaces of mica mounted in cross-cylinder configuration are measured using highly sensitive force-measuring springs and the separation distance is determined by an optical technique (usually to better than 0.l nm).The force is measured through the expansion and contraction of a piezoelectric crystal.
An extension or offshoot of the SFA is the atomic force microscope, readily available commercially, where the force is measured not between two macroscopic surfaces but between a fine tip and a flat surface.Several different researchers have attached micron-sized quartz or other spheres to the AFM tip and measured long-range electrostatic repulsive forces and hydrophobic attractive forces.Yoon et al. 43) measured interaction forces between silanated glass spheres and a silanated silica plate and observed major attractive forces with increasing contact angles.In a recent AFM study, Zhang and Yoon 44) measured the interaction force between glass spheres (12 to 20 μm diameter) and a silica plate in the presence of alkyltrimethylammonium chlorides (CTAC) with 12, 14, 16 and 18 carbon atoms in their alkyl chains.Contact angles were also measured on the flat silica plates.The experiments were carried out at various surfactant concentrations, with maximum hydrophobicity appearing to occur under conditions where the amount of CTAC adsorbed in the Stern layer exactly equals the surface charge, namely when the zeta potential is brought to zero, that is, the point of zeta reversal, pzr.At higher concentrations, where the zeta potential would have reversed sign, electrical double layer repulsion forces reduce the total interaction force.Fig. 20 is a plot of their results giving the force-distance curves for the glass sphere/silica plate system with the four different surfactants at the concentrations of their strongest attractive forces, at what must be the pzr for each of these systems.These curves are fitted to a single-exponential force law, namely: where F is the interaction force, R is the radius of the glass sphere, H is the separation distance between the glass sphere and the silica plate, and C and D are fitting parameters.In this system, C was found to be a constant with a value of 1.35.D is considered to be the decay length.As can be seen by the plots given in Fig. 20, the attractive force between the hydrophobic sphere and plate extends farther out as the chain length of the surfactant is increased, meaning that the apparent hydrophobic force is greater, the longer the chain on the surfactant.Table 3 summarizes the results for conditions of strongest attraction with Fig. 20 Normalized force / separation distance curves to illustrate the effect of chain length of alkyltrimethylammonium chloride surfactants on the hydrophobic force acting between a glass sphere and a silica plate.Each surfactant is at the concentration that gave maximum hydrophobicity (after Zhang and Yoon [44]).
the four surfactants.This table clearly shows, as expected, that the longer the chain, the lower the bulk concentration required to attain maximum hydrophobicity.Since micelle formation in solution results from the same hydrophobic phenomena, the critical micelle concentration (CMC) of the four surfactants is also give in Table 3.The bulk concentration of a given surfactant required to achieve maximum hydrophobicity at the surface is about one-hundreth the CMC.Even though hydrophobic attraction changes markedly, the magnitude of the contact angles under those conditions does not change very much.These results suggest, once more, that the most powerful way to find the condition for maximum hydrophobicity due to the adsorption of a surfactant is to ascertain conditions under which the zeta potential is brought to zero, namely to determine the pzr from electrophoresis measurements.

Conclusions
Wetting of a solid refers to the displacement of one fluid by another at the solid surface and is determined by the different interfacial tensions involved.The balance among these surface forces defines the contact angle which is typically used to characterize wetting behavior.While interfacial tensions between fluid phases (liquid-gas, liquid-liquid) are readily measured, those involving a solid are not generally amenable to direct measurement.Consequently, prediction of contact angles based on the properties of the individual phases is not generally possible, although estimates can be made under certain conditions.
Control of wetting characteristics is desirable for many applications.Surface-active agents can be used to enhance wetting (reduce contact angle) by lowering the liquid-gas or liquid-liquid interfacial tension.Adsorption of appropriate reagents at the solid surface can enhance the wetting of lyophobic solids or reduce that of lyophilic surfaces.Characterization of wetting behavior is critical to the selection of appropriate reagents for specific applications.Contact angles on solids are commonly measured by direct obser vation of drops or bubbles in contact with the surface.Unfortunately, this approach is impractical for very small particles and it cannot necessarily be assumed that the wettability of a fine particle is the same as that of the polished surface of a large piece of the same material.Measurements on compacted powders involve uncertainties due to surface irregularities and liquid penetration.The relative disposition of a particle between the two phases at a liquid-gas or liquid-liquid interface can be used to determine an effective contact angle (the actual angle if the particle is a sphere).Since the method is based on the assumption that gravity is negligible relative to the surface forces, its use is generally limited to particles smaller than about 1mm.Measurements on particles as small as 0.1 mm have been reported.Measurements of the extent and/or rate of penetration of liquid into packed beds of powder can provide estimates of contact angles.Uncer tainties in the estimates arise primarily from problems in the reproducible preparation and characterization of uniform particle beds.
Appropriately chosen indirect methods for characterizing wetting behavior can offer advantages in simplicity of measurement and direct relevance to specific applications.Methods such as immersion/sink time, wetting rate and imbibition time are directly applicable to problems associated with the dispersion of powders in liquid.Film flotation, the Hallimond tube, liquid-liquid extraction and particle-bubble pickup and induction time provide information of direct relevance to solid-solid separation processes based on selective wetting control.In some cases, the indirect methods can yield estimates of contact angles or quantities such as the Zisman critical surface tension.
The limited ability to predict wetting behavior based on material properties continues to be a primary constraint in practical applications.Development of a more complete understanding of the quantitative role of the various interaction forces in establishing interfacial tensions is a critical need.At the same time, improved procedures for the measurement of criteria such as contact angles, especially on fine powders, are needed to establish the basic parameters necessary for predictions to be made.

Nomenclature
A, A1 Cross-sectional areas defined in Fig.

Fig. 1
Fig. 1 Contact angle at a solid surface as measured through the liquid phase.

Fig. 2 Fig. 3
Fig. 2 Relative free energy changes involved in different stages of wetting.

Fig. 4
Fig. 4 Location of a solid particle at the interface between two fluids

Fig. 13
Fig.13 Correlation of the effect of pH on the induction time and zeta potential of as-received 100×150-mesh Cambria #78 bituminous coal in 0.002 M sodium nitrate as a function of pH (adapted from Jia[29]).

Fig. 15
Fig.15 The wetting rate test results for a sample of coal (hvA bituminous) wetted by a surfactant solution (4×10 -4 M Triton X-100).The results are typical of those observed in many experiments (after Chander et al.,[1987]).

Fig. 16 A
Fig.16 A schematic representation of the types of wetting curves obtained by various combinations of type and concentration of the surfactant, and the type of hydrophobic characteristics of the solid (after Chander et al.,[1987]).

Table 2
20)tact angles of various materials calculated from film flotation results with 100 × 150 μm particles and measured by captive bubble (CB) or sessile-drop (SD) methods on flat surfaces (after Fuerstenau et al.,20))

Table 3
43)mary of the results for measuring attractive force curves by AFM measurements between a glass sphere and a silica plate with alkyltrimethyl-ammonium chloride solutions, including the concentration for maximum attractive force (the pzr), the decay length of the attractive forces, the contact angle on the silica plate, and the CMC of the surfactants (after Yoon et al.,43))