A multiple equilibria model of the adsorption of oleate aqueous species at the goethite—water interface

A Multiple Equilibria Model of the Adsorption of Oleate Aqueous Species at the Goethite-Water Interface R. F. JUNG, ~ R. O. JAMES, 2 AND T. W. HEALY 3 Colloid and Surface Chemistry Group, Department of Physical Chemistry, The University of Melbourne, Parkville, Victoria 3052, Australia

ReceivedJanuary 5, 1987;acceptedMay 12, 1987 The interaction between monomeric and dimeric aqueous oleate speciesand ionizinggroups at the goethite-waterinterfacehas beenconstructedto analyzeexperimentalsurfacechangetitration and oleate adsorption isotherms. The binding constants that emerge and the pH--concentrationdistribution of surfaceoleatecomplexessupport the conceptof anion-neutralmoleculecoadsorptionor acid soapdimer adsorption in the regionwhereprecipitationof oleicacid and/or ferricoleateis precluded. ©1988Academic Press, inc.

INTRODUCTION The total distribution of ions and dipoles at aqueous interfaces continues to be an active area of fundamental research. When the interface is formed from an essentially insoluble insulating inorganic oxide such as SiO2, A1203, or Fe203 the considerable advantages of the electrochemical approach pioneered by Grahame (1) and others (2) are not available. Instread, interpretation must rely on derived electrical quantities (e.g., zeta potentials) or on analysis of adsorption isotherms of solute species. The most successful method of analysis of solute adsorption at the inorganic oxide-water interface is the surface group ionization model formulated first by Levine and Smith (3) and extended by Yates et al. (4) and especially by James et al. (5-7). A recent detailed study (8) of the adsorption of aqueous oleic acid-oleate surfactant species at the goethite (a-FeOOH)-water interface provides sufficient data for comprehensive testing of the surface group ionization model where the solute itself is involved in complex Present address: CSIRO Divisionof Fossil Fuels, P.O. Box 136, North Ryde, New South Wales 2113, Australia. 2Present address: Kodak Research, Rochester,NY. 3To whom correspondenceshould be addressed. 0021-9797/88 $3.00 Copyright © 1988 by Academic Press, Inc. All fights of reproduction in any form reserved.

hydrolysis equilibria. The essential fatty acid, cis-9-0ctadecanoic acid, or commonly "oleic acid" and the salt sodium oleate are important natural surfactants used in a range of applications involving wettability, dispersion, froth flotation, etc. The oxide in the form ofgoethite (a-FeOOH) is a well-characterized colloidal inorganic oxide that can be readily synthesized and purified. The titratable surface charge for the present goethite sample is reported elsewhere (8). We can represent the various aqueous or dissolved oleate species as HO1, O1-, O122-, and HOI~ and the bulk phases as HOl~l), NaOlts ), and NaHOl2(l). For the present we shall be concerned with solutions where the pH and total oleate concentration (O1T~q) are such as to ensure no precipitatiOn of the bulk phases. The goethite-water interface will be considered as a diprotic acid or amphoteric surface with FeOH °, FeO-, and FeOH~ groups, the members of which are determined by intrinsic surface dissociation constants. The model allows interaction of solute species with these groups and control of the oleate solution chemistry is through known hydrolysis constants and the availability of each species at each pH and total concentration. The set of binding constants that generates the best fit

544 Journal of Colloid and Interface Science, Vol. 122, No. 2, April 1988

ADSORPTION OF OLEATE AQUEOUS SPECIES

over a wide range o f concentration, pH, and available surface area is derived from the COMICS routine (9). The innovation in the present study of oleate adsorption on goethite resides in the need to invoke both monomer and dimeric solute species as adsorbates, the fractional concentrations of which are p H and total concentration dependent. THE ADSORPTION MODEL

The multiple equilibrium model in essence is as follows. The oxide surface is considered to be composed of positive, neutral, and negative sites, related by surface acidity constants, as shown by Eqs. [1] and [2], FeOH~ ~ F e O H ° + H +

[1]

FeOH ° ~ FeO- + H +

[2]

The adsorbed species is written in terms of a reaction involving the neutral groups of the surface and protons. Surfactant association at the interface is represented by the reaction of n surface sites simultaneously with n surfactant ions. The equations have been written in terms of FeOH sites because they form by far the most predominant surface species over the pH range considered. Solution equilibrium relationships are also used to determine the concentration of solution species in the adsorption system. The apparent surface acidity constants K1 (from Eq. [1]) and K2 (from Eq. [2]) were determined from the potentiometric surface charge titration data (8), considering the surface ionization as charge dependent (11), and as detailed earlier (15) it follows that PKl = pH - log(s0/s+)

[3]

PK2 = pH - log(s_/so),

[4]

and where s+ + so + s_ = 1 for which s+ is the fraction of FeOH~" sites, So the fraction of FeOH sites, and s_ the fraction of FeO- surface sites. The apparent surface acidity constants for goethite, based on 16.5 exchangeable protons]

545

nm 2 (12), were found from the observed titration data to be pKI = 5.4 and pK2 = 12.1 at 25°C and 10 -2 m o l e . d m -3 KNO3 as supporting indifferent electrolyte. The large ApKa of the average apparent Ka values denotes the fact that neutral groups predominate at the goethite surface in the region where oleate species adsorption has been determined. The derivation of the adsorption model is based on a number of experimental observations. (i) Adsorption decreases as pH increases and therefore H + is consumed or O H - is released. (ii) The plot of a percentage adsorbed versus pH curve is dependent upon the concentration, with the percentage adsorption increasing with concentration. Therefore there must be some associated or polymeric species adsorbing (13). (iii) The slope of the log(adsorption density, I"o3 vs log(equilibrium concentration, [Ol]oa) at pH < 9 is approximately 2. This suggests that the important adsorbed species has dimeric stoichiometry. The adsorption reaction may be written in the general form

xFeOH + yOl- + zH + [(FeOH)x(O1)y(H)z] (z-y)+ [5] where x, y, and z are stoichiometric coefficients. For a given y and [Ol]T~q, a higher z gives a steeper percentage adsorbed versus pH curve. The higher the value of y for a given pH, the wider the separation between the percentage adsorbed versus pH curves. The effect of polymeric adsorbed species on percentage adsorbed versus pH curves for a given x and z and excess surface sites is shown schematically in Fig. I. From the curves we can see that increasing the concentration ofadsorbate added to the system (C3 > C2 > C1) has no noticeable effect in the case o f m o n o m e r i c adsorption (Fig. lb), but increases the percentage adsorption when polymeric adsorption occurs (Fig. la). Journal of Colloid and Interface Science, Vol. 122,No. 2, April 1988

546

JUNG, JAMES, AND HEALY (a) y ;~2

TABLE I

(b] y--1

Solution and Interfacial Equilibria for the Goethite-Oleate System

pH

pH

FIG. 1. Schematicrepresentationofpercantageadsorbed as a function of pH for (a) polymericadsorption, and (b) monomeric adsorption. A number of adsorbed species were considered in modeling the oleate adsorption isotherms. It was found that the species FeOHOI-, FeOH2OI, (FeOHO1) 2-, and (FeOHO1)2H- gave a satisfactory fit to the curves. Other species such as higher surfactant aggregates and more protonated adsorption species were not required for the available experimental data. We have assumed that an FeOH surface site for oleate adsorption (55 A2/site) is not the same as the usual goethite surface site (6.1 A2/ site), which corresponds to the area per exchangeable proton (12). The 55 A2/site has been chosen since adsorption (8) and monolayer experiments with oleic acid (14) show this to represent a monolayer coverage for the oleate system. Surface adsorption equilibrium constants were estimated from the adsorption data using a method discussed by James et al. (15). This approach uses a distribution ratio D' which is defined as D' =

(oleate adsorbed) mole. m-2- S m2- dm-3 [O1-] mole. dm-3 solution

FeOH + O1- ~ FeOH + O1- + H+ ~ 2FeOH + 201- ~ 2FeOH + 201- + H+ ~ FeOH ~ FeOH + H+ ~ 201- ~ 201- + H+ ~

Ol- + H + ~- H O |

{K~,} {K2s} {K3,} {K4,} {K2} {(K0-'} {Kb} {KhD} {(K~)-1}

[71 [8] [9] [10] [111 [12] [13] [14] [15]

to the surface concentration ([ ]smole m-2). For convenience in this work, surface species and equilibrium constants involving them are all expressed in solution concentration units (since IX] m = [X]~. Sin). Kns represents the equilibrium constant for a surface adsorption reaction, such as is defined generally by Eq. [5]. The formal definition of the equilibria represented by the model is given in Table I. Equations [7] and [ 10] of Table I may be used to derive an expression for the distribution coefficient (D') from Eq. [6]; thus, D' = [FeOH](Kls + K2saH+ + K3s[FeOH][OI-] +/Gs[FeOH][OI-]aH+). [16] Plots of D'/[FeOH] as a function of an* at different equilibrium [O1-]'s were used to obtain estimates of the surface equilibrium constants, using the experimental adsorption data given elsewhere (8); see Table II. Kls and K3s TABLE II

n=l

= X

FeOHOIFeOH2OI (FeOHOI)~(FeOHOI)2HFeO- + H+ FeOH~O12HOI~

K=[OI-]m-I[FeOH]'~"Sm[H+] q, [6]

n~m

where m represents the stoichiometric coeffidents of O1- and [FeOH], and q the coefficient of H ÷ as defined in adsorption equilibria. S is the solid surface area concentration in m 2 dm -3. In Eq. [6] the solution concentration ([ ]mole dm -3) is expressed in different units Journal of Colloid and Interface Science, Vol. 122, No. 2, April 1988

Comparison of the Estimates of Surface Equilibrium Constants Derived Initially with the Refined Values Constant

log KI, log K2, log K3, log/G,

Graphically derived

2.85 (+0.15) 11.4 (+0.6)

11.4 (_+1.3) 21.7 (+0.2)

Refined with COMICS (+0.1)

2.8 10.0 10.8 21.6

ADSORPTION OF OLEATE AQUEOUS SPECIES i

!

i

547

TABLE III

HC~

Equilibrium Constants ~ Used in the Oleate-GoethiteWater Adsorption Model

•• lO-6

log KI~ = log K2~ = log K3, = log/(4, =

10-7 104

2.8 10.0 10.8 21.6

log/(2 - l o g KI log K b log K~D --log K~

= = = = =

-12.1 5.4 4.0 9.65 4.95

The stoichemistry is given in Eqs. [7] to [15] of Table

10-9

I. I

I

I

i0-6 10-5 [OLEATE ]T (tooldm-]]

i

10-/+

FIG. 2. Adsorption density of oleate of the goethitewater interface as a function of equilibrium total oleate concentrations. The points are derived experimentally while the curves result from the model discussed in this section.

were derived from the intercepts and K2~ and K4~ were derived from the slopes of these curves. In these calculations [FeOH] was assumed constant, since [FeOH] >> [O1]x and surface ionization was small. The [Ol-] was derived from [Ol]a- by standard multiple equilibria methods. Further refinements to the equilibrium constants of Eqs. [7] to [10] were made by

trial and error use of the computer program COMICS (9). The calculated curves from the model for log(adsorption density) versus log([oleate]T) at varying pH's are compared with the experimental data points in Fig. 2. Similarly the calculated percentage adsorbed oleate versus pH curves for varying oleate concentrations are compared with experimental points in Fig. 3. There is good agreement between the curves derived from the model and the experimental data for both types of plot. This agreement for both types of plot has apparently not been previously achieved in the description of the i

i

i

i

10_4

l ~.0H

........................

0~l~l

..%~1 ..'t'.'." .......

.,x.'..

x Z,:0

t

510_ 6

,"" ,,.

:.

01~."~, ,< 0

5 ~z L~

25x1~ 5 M

I

Added

..............................

::...v.i ~..........

"

"-.. """~.~lo~

10-101

a. 20

"~, ......H01(aq)

i

i

7

9

,'0

pH FIG. 3. Percentage oleate adsorbed on goethite versus pH for varying added oleate concentrations. The curves are calculated from the model while the points are derived from experiment.

pH FIG. 4, Concentration of solution (. • • ) and surface ( and ---) speciesas a function ofpH for the oleate-geothitewater adsorption model. The added oleate concentration is 1 × 10 -4 M a n d the goethite concentration is 115.2 m2/ dm 3. Journal of Colloid and Interface Science, V o l .

122, N o . 2, A p r i l 1988

548

JUNG, JAMES, AND HEALY 1 , ..........

10-5

,

i 1 .~H . . . . . . . . . . . . . .

•

x

~-

0tr~.t~._._~

//

#

0t-.!.t~L

,-

,,

I

--06

,

"%

|

pH

~)..

""

•

,,

. . - ~ "

6

7

'....

NX~N~f'I~OHOI~H -

9

B

10

11

pH

loG. 5. Concentration of solution (. • • ) and surface ( and ---) species as a function of oH for the adsorption model. The added oleate concentration is 3.43 × 10-7 g and the goethite concentration is 115.2 m2/dm 3.

uptake of a surfactant at the mineral-water interface. The values for the equilibrium constants used in the calculated model shown by Eqs. [7] to [ 15] are given in Table III. DISCUSSION

To explore the implications of the model the concentration of various solutions and

'

~

FIG. 7. The fraction of oleate in solution (o~(~)) present as particular species. The system oleate concentration is 10-4 M.

surface species at the highest ( 1 0 - 4 mole. dm -3) and lowest (3.43 × 10 -7 mole. d m -3) added oleate concentrations are as shown in Figs. 4 and 5. These results illustrate that the concentrations of FeOH surface sites are substantially constant over the pH and oleate concentration range considered. (FeOH O1)2H- is the most important absorbed species, particularly at lower pH's and higher oleate concentrations. FeOHO1- is significant in adsorption at higher pH's, as is (FeOHO1)2- for higher oleate concentrations. In the discussion of the model to this point no attempt to identify the form of the adsorption complexes has been made. One approach is to equate the protons and oleate anions added on the left-hand side of Eqs. [7] to [10] with the formation of an oleate adsorption

~a~'

0-8

|

08

i

)

HOt~

j

.

.

Ol-

.

9

10

"~ 06 04+

0"2 I

I

i

8

9 pH

10

0.2

7

FIG. 6. The fraction of oleate adsorbed, (a(~)) present as particular species. The results are derived from the adsorption model as discussed. The system oleate concentration is 10-4 M. Journal of Calloid and Interface Science, V o l .

122, No. 2, April

1988

8

pH

lOG. 8. The fraction of oleate adsorbed (t~(~)) present as particular species. The system oleate concentration is 3.43 × 10-7 M.

ADSORPTION OF OLEATE AQUEOUS SPECIES

species on the neutral goethite surface sites. Thus FeOH201 is equivalent to HOl(aos) and (FeOHO1)2H- is equivalent to HOl~(a~) etc. The fractional distribution of total oleate adsorbed, expressed as HO12-(ads), O12(ads), 2Ol(ads), or HOl(~d~), for the systems shown in the previous two figures is shown in Figs. 8 and 9. When presented in this form, the surfactant aggregate (dimeric) species show a marked tendency to concentrate at the interface, relative to the solution phase. In particular the HOl~ species (according to this interpretation) is spectacularly concentrated at the interface over a broad region of the pH scale. Study of the adsorption of the hydrolyzable metal species of cadmium(II) at the oxidewater interface (15) shows in a similar way that the hydrolyzed species CdOH~d~) is important over a much broader pH region than that for CdOH~aq) in bulk solution. This is analogous to the behavior of HO12 and HO1 in the present system. The foam fractionation studies of alkaline sodium oleate solutions by Eagland and Franks (16) also support the adsorption of HOl~-type species at interfaces. They found a significant proportion of acid species present in the fractionated foam, which was not present in an analysis of the bulk solution. Their experimental technique seems to preclude the possibility of artifacts such as that which might be due to CO2 contamination.

0t-

fie

O6

02 7

I

I

I

8

9

10

pH FIG. 9. The fraction of oleate in solution (a(~)) present as particular species. The system oleate concentration is 3.43 × 10-7 M.

549

The inferential proposals that the acid soap dimer is the surface active species (17, 18) or some other premicellar unit (19) are also consistent with the above output from the model. As stressed earlier, it is important in oleatesolid systems to confirm that the pH-O1T domain investigated is a true oleate aqueous solution region devoid of liquid oleic acid and or metal oleate precipitates. ACKNOWLEDGMENT This work was supported by the Australian Research Grants Commission. REFERENCES 1. Grahame, D. C., Chem. Res. 41, 441 (1947). 2. Parsons, R., "Modern Aspects of Electrochemistry" (B. E. Conway and J. O'M. Bockris, Eds.), Vol. 1, p. 103. 1954. 3. Levine S., and Smith A. L., Discuss. Faraday Soc. 52, 290 (1971). 4. Yates, D. E., Levine, S., and Healy, T. W., Trans. Faraday. Soc. 1 70, 1807 (1974). 5. Davis, J. A., James, R. O., and Leckie, J. 0., J. Colloid Interface Sci. 63, 480 (1978). 6. James, R, O., Davis, J. A., and Leckie, J. O., J. Colloid Interface Sci. 65, 331 (1978). 7. Westall, J., and Hohl, H., Adv. Colloid Interface Sci. 12, 265 (1980). 8. Jung, R. F., James, R. O., and Healy, T. W., Jr. Colloid Interface Sci. 118, 463 (1987). 9. Perrin, D. D., and Sayce, I. G,, Talanta 14, 833 (1967). 10. Schindler, P. W., Furst, B., Dick, R., and Wolf, P., J. Colloid Interface Sci. 55, 469 (1976). 11. Stumm, W., Huang, C, P., and Jenkins, S. R., in "Solid-Liquid Interfaces (Croat. Chim. Acta Special Publ. No. 1)" (Tezak and Pravdic, Eds.), p. 143, 1971. 12. Yates, D. E., Grieser, F., Cooper, R., and Healy, T. W., Aust. ,L Chem. 30, 1655 (1977). 13. Sillen, I. G., in "Coordination Chem" (A. E. Martell, Ed.), Amer. Chem. Soc. Monograph 1168, p, 491. Van Nostrand, New York, 1971. 14. Feher, A. E., Collins, F. D., and Healy, T. W., Aust. J. Chem. 30, 511 (1977), 15. James, R. O., Stiglich, P., and Healy, T. W., Discuss. Faraday Soc. 59, 142 (1975). 16. Eagland, D., and Franks, F., Int. Congr. Surf Activity 2, 539, Koln, September (1960). 17. Ananthpadmanabhan, K., Somasundaran, P., and Healy, T. W., Trans. Amer. Inst. Min. Met. 266, 2003 (1979), 18. Kulkarni, R. K., and Somasundaran, P., Colloids Surf. 1, 387 (1980). 19. Pugh, R. J., Colloids Surf. 18, 19 (1986). Journal of Colloid and Interface Science, Vol. 122, No. 2, April 1988