8776
J. Phys. Chem. 1993,97, 87764779
Contact Angle Hysteresis in Chlorinated Hydrocarbon-Water Mixtures C. D. Rugge' SRE Incorporated, 20 Chapin Road. P.O. Box 251, Pine Brook,New Jersey 07058
R. C. Ablert Rutgers University*Department of Chemical and Biochemical Engineering* Piscataway, New Jersey 08855-0909 Received: January 1 I t 1993; In Final Form April 21, 1993
Variations of two-phase capillary rise were conducted with water and a number of chlorinated hydrocarbons in glass capillaries cleaned with chromic-sulfuric acid. Contact angle hysteresis due to the advance or recession of a water-rganic interface was investigated, relative to the water phase. When water advanced, contact angles of 47.3-62.1 O were calculated. When water receded, equations based on the balance of interfacial and gravity forces within the capillary failed to yield single estimates of the interfacial angle. It was hypothesized that water-receding angles were approaching Oo due to a film of water. This hysteresis was minimized in the smallest capillaries.
According to a nationwide groundwater survey conducted by the Environmental Protection Agency in 1984, trichloroethylene, perchloroethylene,and 1,1,1-trichloroethanewere detected most frequently in municipal drinking supplies obtained from treated groundwater.' These chlorinated ethylenes and ethanes are largely immiscible with water and have specific gravities significantly greater than 1. Released volumes of these liquids are ultimately dispersed in the complex zones of an aquifer via the opposing forces of gravity and interfacial tension. Globs of nonaqueous phase organic tend to become entrapped between soil and sand particles, resulting in a constant source of groundwater contamination. The resultant of the force of interfacial tension acting along a solid surface is given by the cosine of the contact angle. Two possible contact angles formed between a dropof sparingly-watersoluble organic liquid and a solid surface submerged in water are shown in Figure 1. A large contact angle 8 (Figure la), as measured through the aqueous phase, implies that the drop has a greater affinity for the solid than does water. A small contact angle (Figure 1b) means that water has a greater affinity for the solid than does the drop. The chlorinated organics investigated in this study usually resemble Figure l b and are considered to be the nonwetting phase. The contact angle exhibits hysteresis and is dependent on the process of formation. When the water-organic meniscus advances over a solid surface (relative to the wetting phase) the resulting contact angle will be larger than when the meniscus recedes. Understanding the extent of contact angle hysteresis is important in predicting the migration path of a nonaqueous phase liquid through a groundwater system. For instance, displaced water recedes from the infiltrating face of a nonaqueous glob and advances to fill in the voids left by the trailing face. Immiscible fluid flow is further complicated by a postulated "stick-slip" mechanism in which the contact line is pinned by local surface inhomogeneities in a periodicallyunstable manner.2 Local inhomogeneity is thought to be due to molecular rearrangements induced at the solid surface by the contacting liquid.' Advancing and receding liquids may induce different inhomogeneities, thereby causing contact angle hysteresis. Two-phase capillary rise has been used to measure static contact angles between water-organic menisci and solid ~urfaces.4-~ This method is illustrated in Figure 2, where phases a and @ are immiscible liquids. The Ah of phase fi is related to the capillary inner diameter and fluid properties by the following expression: 0022-3654/93/2097-8776$04.00/0
(4 Flgure 1. Interfacial contact angles.
Capillary
3 +
2r
Figure 2. Capillary rise.
y a x c--
x Figure 3. Interfacial force balance.
Ah = (2A.y/r)g(p,t,- PB)(C@ - 4 . 3 The term A.y is considered to be the resultant of interfacial forces between the various phases and thecapillary walls (phasex), asshownin Figure 3. Resolution of the vectors in the horizontal direction yields the equilibrium expression y'I@ cos 8 = lyu/x - y@/xl.This equation cannot be satisfied unless .ya/8>ly/x - .y@/xl, in which case 8 goes to 0. In this situation, phase @ is said to completely wet the surface, separating phase a from solid phase x by a thin film. If 8 is not
0 1993 American Chemical Society
Chlorinated Hydrocarbon-Water Mixtures
The Journal of Physical Chemistry, Vol. 97, No. 34, 1993 8777
*,
L and radius r, inclined at angle with fluid 1 displacing fluid 2. This expression can be viewed as a force balance, with interfacial tension and the applied external pressure drop opposing viscous flow and fluid head. Equation I was modified for the case of advancing water, using equilibrium values of hl and h2 shown on Figure 4: pwgh,r/2+ p~gh,r/2- yw/'cos e"/' - yscos 6' = 0 (11) Equation I1 was rearranged to yield a linear equation for (pwhl + psh2) when plotted against l/r:
+ y8cos 6')-g2 -1r The slope of the line is proportional to (yWls Pis+ co8 e), (pwhl+ pshz) = (yWls cos Ow/8
I Water
CQS
and the angle PlScan be estimated given values of y' CQS 8 and ywls.
Figure 4. Water-advancing contact angle.
For the case of the water-recedingcontact angle, water trapped in a capillary inserted through a solvent-water interface will be displaced by movement of the solvent-water interface. The gravity and interfacial forces on the water plug at equilibrium are described by the following expression, where hl and h2 are equilibrium values shown on Figure 5.
+
p g h Z r / 2= pwg(hl + h2)/2 yW/'COS eW/' - ywc08 8" Since the water in the capillary rises to its equilibrium height and is assumed to remain trapped during penetration of the lower end of the capillary through the water-rganic interface, the first and third terms of the right-hand side of the equation shown above should cancel. The final expression is shown in eq 111:
Solvent
H
= hi
+
h2
Figure 5. Water-receding contact angle. zero, partial wetting by each of the phases occurs and there is no separating film. In this study, the hysteresis of contact angles between waterorganic menisci and a glass surface was studied in capillaries of various inner diameters. Two-phase capillary rise experiments were conducted with a variety of chlorinated hydrocarbons and water. In the presence of theseorganiccompounds,water behaves as the wetting phase; consequently, measurements were taken with water advancing or receding. The water-advancing and water-receding scenarios modeled the response of water to the trailing and leading faces, respectively, of an infiltrating nonaqueous phase liquid. In the water-advancing case, water was imbibed into a capillary that already contained a small plug of organic (see Figure 4). Water spontaneouslyrose into the capillary, pushing the organic plug ahead of the water-organic interface, and thus contacted a surface initially coated with organic. For the water-receding case, water was imbibed into a capillary. As shown in Figure 5, the capillary was then submerged through a water-rganic interface, bringing solventin contact with a surface initially coated with water. -ry
Equationsdescribing water-advancing and water-receding twophase capillary rise were derived from the dynamic description shown as eq I.*
d g [ p , x - p 2 ( -~ x)]sin 9 (I) This is a general expression for capillary flow in a tube of length
A plot of h2 versus 1/ r has a slope proportional to ywlsCQS PI', if the relationship is linear. Given values of yw/s, Pis can be estimated. The quantity y' CQS 8 can be determined from one-phase capillary rise heights, calculated by p&hr/2 = yocos B. For a series of measurements in capillaries of various radii, a plot of h versus l/r yields an estimate of y' CQS B from slope 2y co8 B/P&*
Experimeat.l Section
These equations were verified experimentally in capillaries of various diameters for several organics. Capillaries of 0.01350.0444 in. inner diameter were purchased from Drummond Scientific (Broomall, Pennsylvania), To verify manufacturer's stated dimensions, a plug of water was imbibed into four capillaria of each size. The weight of the plug was determined as the difference between empty and filled capillary weights, and plug length from meniscus to meniscus was measured with a steel ruler calibrated in 1/64-in. increments. The inner diameter was calculated on the assumption of a Oo contact angle and a water density of 1.0 g/cm3. Manufacturer and measured dimensions agreed to within O.OOO4 in. To ensure clean surfaces, capillaries were treated by the following protocol: Sparkleen and water soak, deionized water rinse; methanol soak,deionized water rinse, Chromerge (100%) soak (at least overnight), deionized water rinse, and oven-dried at 110 OC. An 80 OC Chromerge soak, followed by deionized water rinsing and boiling, hydroxylates the surface of the glass.9 It is likely that the room-temperature treatment followed in this study created a partially-hydroxylated polar surface. All interface heights were determined from menisci centers with a steel ruler calibrated in 1/64-in. increments. Capillaries were attached to the ruler with two small bands arranged in such a way as to ensure that the bands would be submerged only in the aqueous phase. Four capillaries of each size were aligned along the edge of the ruler and tested simultaneously. All capillary
8778 The Journal of Physical Chemistry, Vol. 97, No.34, 1993
Rugge and Ahlert
120
0.8 -
1
/ /
a
d
/
0.6
-
0.4
-
/ / /
> i
o*2 00
20
40
60
80
100
120
140
0
o
20
10
60
80
i m 120 140
160
Inverse Radius (1/in)
160
Figure 7. 1,1,1-Trichloroethane-wettedinterfacial contact angle determination.
Inverse Radius (Mn) Figure 6. One-phase capillary rise of water-saturated l,l,l-TCA.
rise experiments were conducted in a thermostated water bath held at 25 OC. Solventsusedin thisstudyincluded l,l,l-trichloroethane(ll1TCA), 1,l,2-trichloroethane (1 12-TCA), perchloroethylene (PCE), trichloroethylene(TCE), chloroform (CF), carbon tetrachloride (CT), and chlorobenzene (CB). All solventswere obtained from Fisher Scientific (Springfield, NJ) and used without further purification. Approximately 30 mL of each solvent was stored with 100 mL of deionized water for one week at ambient temperature to minimize diffusional losses during capillary measurements. For estimates of solvent-air interfacial properties, saturated organic was allowed to rise into clean capillaries. The height of the bottom of the liquid-air meniscus was measured relative to the level of solvent outside the capillary. For water-advancing contact angle determinations,solvent was imbibed into clean capillaries to wet the inner surface. The capillary was withdrawn from the solvent, leaving a small plug trapped at the lower end. The solvent-loaded capillary was immersed in water after excess solvent was removed from the outer surface of the capillary with a tissue. After an equilibration time of approximately 10 min, the positions of the solvent-air and solvent-water interfaces relative to the external water-air meniscus were recorded. This procedure was repeated for each solvent with six capillary sizes and four replicate measurements within each size. For determination of water-receding contact angles, water was imbibed into a clean capillary, with the water contained in a thermostated beaker. After water levels were stable, solvent was injected into the bottom of the beaker until the level of the newlyformed solvent-water interface outside the capillaries was well over the lower edges of the capillaries. The levels of the inner and outer water-solvent menisci were recorded after 10 min equilibration time. This procedure was repeated for all solvents with six capillary sizes and four replicate measurements within each size.
Results and Discussion The quantity 7' cos B was determined for each organic liquid from one-phase capillary rise measurements. The rise in each capillary was plotted against inverse capillary radius; a sample plot is shown in Figure 6 for 111-TCA. Since the data showed minimal scatter, the ordinary least squares method was used to determine the equation of the lines. All values of the regression coefficient were greater than 0.989. Values of the solvent-air interfacial tension components, summarized in Table I, were calculated from the slopes using pure liquid-air properties.10 In
TABLE I: Vertical Organic-Air Surface Tension Components TCE 1 12-TCA CF CB CT PCE 11 1-TCA
21.2 31.4 26.4 31.1 26.1 30.6 24.5
19.2 11.9 5.1 19.3 2.1 12.1 16.6
TABLE Ik Water-Advancing Interfacial Tension Components and Contact Angles solvent TCE 1 12-TCA CF CB CT PCE 1 1 1-TCA
y*lwcos Clw(dyn/cm)
&Iw (deg)
20.2 20.4 11.4 23.9 19.8 23.6 20.9
58.1 41.2 49.1 44.1 62.1 58.9 41.3
the absence of water, the organic species in this study wetted borosilicate glass. Contact angles ranged from 2.7O to 19.3O. A representative plot of (p,hl+ p,hz) versus 1/ r for 111-TCA with water advancing is shown in Figure 7. All organic species exhibited a linear relationship between (p,hl p,hz) and the inverse radius, although the data was somewhat scattered. To minimize the effect of outliers on the slopes generated for contact angle determination, robust regression was used. The subroutine "rreg" in the UNIX Statistical Program S was used to weight outliers lower than the rest of the data. The robust method used included a converged Huber estimate followed by two iterations of the Bisquare Method." Theerror bars indicate 95% confidence limits. Values of yllwcos $Iw, shown in Table 11, were determined from the slope; contact angles were calculated from published interfacial tensions.*2 Interfacial angles through the aqueous phase ranged from 47.3O to 62.1O. As shown in Figure 7, the experimentaldata exhibiteda constant slope and zero intercept, as predicted by eq 11. The reasonable values of the contact angles listed in Table I1 indicate that this equation, which includes only interfacial tension and gravity, describes all significant forces within the capillaries. According to water-advancing contact angle determinations, the waterorganic interface partially wetted the inside walls of glass capillaries. The nonaqueous phase thus did not behave purely nonwettingon glass surfacesfirst wetted by the nonaqueousphase. If there is no hysteresis, water-receding contact angles are expected to be similar to water-advancing anglesand independent of capillary radius; a plot of the water-receding value of hz versus inverse radius would be linear with a slope proportional to ys/w
+
The Journal of Physical Chemistry, Vol. 97, NO. 34, 1993 8779
Chlorinated Hydrocarbon-Water Mixtures
thicknesses may be formed. These films of water could separate the organic from the solid, resulting in 0 ' contact angles.
Complete Wetting
60
Conclusions 50
20 IO
By Water
0 0
20
40
60
80
100
120
140
160
Inverse Radius (1/in)
Figure 8. l,l,l-Trichloroethane water-receding
values; lines show expected results for partial wetting (water-advancing contact angle) and complete wetting (angle = O O ) . h2
TABLE I& Cosines of Water-Receding Interfacial Tensions as a Function of Radii radius (in.) TCE 112-TCA 0.92 0.0068 0.65 0.87 0.0109 0.93 a 0.0138 1.03 0.0157 a a a 0.0191 1.21 a 0.0222 0.91 partial 0.68 wetting 0.53
cos el.
CF 0.47 0.73 0.84 0.82 0.88 a
CB 0.94 0.91 a 0.92 0.67 a
CT 0.47 0.73 0.95 0.84 0.46 a
0.65 0.71 0.47
PCE 111-TCA 0.53 0.87 0.82 1.08 1.02 1.23 0.95 0.84 0.91 1.21 0.90 a 0.52
0.68
a Value rejected because initial water rise height was not within 10% of theoretical rise of pure water.
cos WW.However, this result was not observed. Figure 8 shows a representative plot for 11I-TCA, superimposed lines show expected results based on the water-advancing contact angle (labeled as "Partial Wetting by Water") and on a contact angle of zero (labeledas "Complete Wetting by Water"). In this figure, experimental values of h2 are greater than partial wetting values and often exceed complete wetting values. The slope method was discarded for the analysis of waterreceding measurements, and angles were calculated for each radius; see Table 111. Angles were only listed when the initial rise of water in the capillary (the sum of hl and hz in Figure 5 ) was within 10%of the theoretical rise of pure water. Cosines of water-advancing contact angles are included in this table for reference. The water-receding cosines were generally greater than wateradvancing estimates, implying that receding angles were smaller than advancing angles. The receding water-organic interface thus exhibited a greater tendency toward nonwetting than the advancing interface. When the solid surface was first in contact with water, the solvent appeared to be more nonwetting than when the surface was initially in contact with solvent. Table I11also shows that the tendency to behave as a nonwetting fluid on water-wetted surfaces may be dependent on capillary radius. In the smallest capillary tested, most organic liquids exhibited angles equivalent to or approaching water-advancing values. Thus, contact angle hysteresis was minimized in small diameter capillaries. The measurements reported in this study are not sensitive enough toquantify the exact natureof the radius effect. However, it has been reported that the thickness of the film trailing a wetting fluid in a capillary is a function of radius and the velocity of the meniscus.13 In larger capillaries, meniscus velocities are high during the approach to equilibrium, and water films of significant
Two-phase capillary rise, described by equations incorporating gravity and interfacial forces, yielded insights into the wetting properties of a group of chlorinated organics in the presence of water. When the organic-water interface advanced (relative to water), partial-wettingcontact anglesranging from47.3O to62.1 O, were calculated from the force balance equation. When the interface receded, non-wetting (0') and partial-wetting angles were determined for large and small capillaries, respectively. Contact angle hysteresis, as evident in this study, was attributed to a postulated film of water trailing the receding interface. These conclusions indicate that chlorinated organics in the presence of water should not universally be considered the nonwetting phase.
Acknowledgment. This research was funded by the New Jersey Hazardous Substance Management Research Center. This work would not have been possible without the laboratory assistance provided by Christopher Milos, Joseph Zukowski, and He-Sun Pak. Nomenclature concentration, mol/cm3 gravity, cm/s2 capillary rise height, cm radius, cm time, s fluid velocity, cm/s length, cm pressure density, g/cm3 interfacial tension, dyn/cm viscosity, cm2/s contact angle, degrees angle of inclination, degrees water-air property solvent-air property water-organic (interfacial) property saturated liquid or gas phase solid phase pure liquid or gas phase length variable, cm
References and Notes (1) Westrick, J. J.; Mello, J. W.; Thomas, R. F. J. Air Waste Manage. Assoc. 1984, 52. (2) Sauer, B. B.; Carney, T. E. Langmuir 1990,6, 1002. (3) Chen, Y. L.; Helm, C. A.; Israelachvili, J. N. J . Phys. Chem. 1991, 95, 10736. (4) Sigl, L.; Fenzl, F. Phys. Reu. Lett. 1986, 47, 2191. (5) Pohl, D. W.; Goldburg, W. I. Phys. Rev. Lett. 1982, 48, 1111. (6) Wiltzius, P.; Dierker, S.B.; Dennis, B. S.Phys. Lett. 1989,62,804. (7) Dixon, J. A.;Schlossman, M.; Wu, X . ; Franck, C. Phys. Rev. B 1985, 31, 1478. (8) Barrer, R. M. Discuss. Faraday Soc. 1948, 3, 61. (9) Rosebury, F. Handbook of Electron Tube and Vacuum Techniques; Addison-Wesley: Reading, MA, 1965; Chapter 1. (10) Riddick,J. A.; Bunger,W. B.;Sakano,T.K.OrganicSoluentsPhysical Properties and Methods ofPurification; John Wiley and Sons: New York, 1986. (1 1) Becker, R. A.; Chambers, J. M. S An Interactive Environment for Data Analysis and Graphics;Wadsworth Advanced Book Program: Belmont, California, 1984; Appendix 1. (12) Rugge, C. D.; Ahlert, R. C.; Zukowski, J. M. Spec. Sci. Tech. 1992, 15, 30. (13) Van Remooretere, P.; Joos, P. J. Colloid Interface Sci. 1991, 141, 348.
