J . Chem. SOC., Faraday Trans. 1, 1987, 83, 1703-1709 Forces between Mica Surfaces in Aqueous KNO, Solution in the range 10-4-10-1 mol dm-3, showing Long-range Attraction at High Electrolyte Concentration Chris Toprakcioglu Cavendish Laboratory , Madingle y Road, Cambridge Jacob Klein*T Polymer Department, Weizmann Institute of Science, Rehovot 76100, Israel and Cavendish Laboratory, Madingley Road, Cambridge Paul F. Luckham Department of Chemical Engineering and Chemical Technology, Imperial College, Prince Consort Road, London S W7 2BT Interactions between mica surfaces immersed in 10-4-10-1 mol dm-, aque- ous KNO, have been measured: in contrast to earlier studies in this system, at the higher concentration range (> 4 x mol drn-,) the potential of the diffuse double-layer drops to values ry, < 30 mV.In 0.1 mol dmP3 KNO, only attraction is observed down to the range of ‘hydration forces’ (ca. 2 nm). The discrepancy with the earlier results is attributed to adsorption on the mica of surfactant originating in the Millipore filter used in our previous studies. The measurement of the forces acting between molecularly smooth mica surfaces immersed in aqueous electrolyte media has been the subject of active experimental interest. The pioneering investigation of Israelachvili and Adamsl and subsequent s t u d i e ~ ~ - ~ have revealed that for a variety of electrolytes and concentrations these forces are generally well described by DLVO with a double-layer repulsion predominating at long distances, and van der Waals attraction becoming appreciable closer in, while strong ‘hydration’ forces were found to operate at distances close to molecular contact.Similar experiments by Klein and Luckham*$ (although only with KNO, at two values of electrolyte concentration) performed as a preliminary to the determination of the forces between macromolecular layers adsorbed on mica, yielded substantially similar results. More recently, however, we performed a series of experi- ments which clearly demonstrate the presence of attractive forces (at high electrolyte concentrations) at surface separations where the earlier studies indicated a strong repulsion. We report here the results of this investigation of interactions between mica surfaces in aqueous KNO, as a function of electrolyte concentration.Experimental Materials The water used was first deionized and then doubly distilled from an all-quartz apparatus. It has a resistivity > 2 x lo6 i2 cm and a pH in the range 5.5-6 owing to dissolved atmospheric CO,. Potassium nitrate (analytical grade) was obtained from B.D.H. and t Permanent address and address for correspondence : Polymer Research, Weizmann Institute of Science, Rehovot 76100. Israel. 17031704 Forces between Mica Surfaces in Solution from Fisons, and electrolyte solutions were prepared by dissolving the salt in doubly distilled water. The mica used was best quality FS/GS, grade 2, Muscovite ruby mica, obtained from Mica & Micanite Ltd. Apparatus The apparatus used in these experiments was similar to that described earlier'? * and will not be described in detail here.The technique employs multiple-beam interferometry, from which the separation between two mica surfaces can be determined to a typical accuracy o f k 3 A and the force acting between the two surfaces to withink0.1 pN. Procedure All parts of the apparatus that came into contact with the electrolyte solution were thoroughly cleaned before each experiment. Metal and Delrin parts of the apparatus were left in toluene for 24-48 h, washed with absolute ethanol and ultrasonicated in 0.1 mol dm-, nitric acid. Glassware was immersed in chromic acid for 24 h at room temperature or in hot concentrated aqueous NaOH for a few minutes. All components were finally rinsed in doubly distilled water and further washed with filtered absolute ethanol (0.2 pm Gellman Teflon filters).They were then dried in a laminar flow cabinet. After assembly of the apparatus and measurement of the contact position of the mica surfaces in air, the surfaces were separated to ca. 3 mm and potassium nitrate solution, filtered through 0.22 pm Millipore Triton-free (TF) grade filters, was introduced into the cell until the two surfaces were immersed in the electrolyte. After allowing ca. 30 min for thermal equilibration, the force us. distance profile in the electrolyte was determined. The temperature of our experiments was 20f 1 "C. Results and Discussion Fig. 1 shows the force, F(D), vs. distance, D, profile between two mica surfaces (radius of curvature R % 7 mm) in aqueous solutions of KNO, at three different concentrations and with pH 5.7 k0.2.The force axis is normalized with respect to the radius of curvature of the mica surfaces. According to the Derjaguin approximationlo the quantity F/2R gives the interaction energy per unit area of two flat parallel surfaces a distance D apart obeying the same force law. At large surface separations the forces are repulsive owing to electrical double-layer interactions. At sufficiently short separation, however, the attractive van der Waals forces operating between the two surfaces exceed the double- layer repulsion (i.e. the potential curve has a 'primary' minimum) and the surfaces jump into contact. On separation of the surfaces, long outward jumps are observed (ca. 5 x lo4 A) which give a measure of the depth of this potential well, given that the spring constant of the leaf spring on which the lower mica surface is mounted is K = 110+10Nm-l.These results are consistent with DLVO theory,' and the observed forces are well described by the approximate DLVO expression for a 1- 1 electrolyte, incorporating a double layer on the mica surfaces: F 641 kT A --- tanh2 (zT) exp ( - K D ) - - - 2nR - K 12;nD2 where n is the number of ions per unit volume, k is Boltzmann's constant, T is the absolute temperature, e is the electronic charge, K is the Debye-Huckel parameter, Wd is the potential of the diffuse double layer and A is the Hamaker constant. Debye lengths I / K and potentials of the diffuse double layer t,u calculated from the experimental data of fig. 1 give the values 1/K=175*51 and Wd=105+10mV for 3x1OW4 mol dm-, KNO,, and l / ~ = 37+ 5 A and tyd = 1 lo+ 10 mV for mol dm-, KNO,.C .Toprakcioglu, J . Klein and P . F. Luckham 1705 Fig. 1. F(D)/R us. D profile between two curved mica surfaces (radius of curvature R z 7 mm) in aqueous KNO,: A, 3 x lo-* mol dm-, KNO, ; 0, mol dm-, KNO,. The Debye lengths for 3 x lop4 and mol dm-, were calculated from the slopes of the linear parts of the continuous lines for these concentrations, while the curved parts of the profiles are best-fit by eye. For 4 x mol dm-, the entire continuous line is a theoretical DLVO-type interaction with 1 / ~ = 20 A, ‘yd = 32 mV and A = 2.5 x J [eqn (l), see text]. The broken line corresponds to the force observed on further compression at this concentration (4 x rnol drn-,), attributed to ‘hydration’ effects.Inset: Variation of surface potential vd of mica surfaces with KNO, concentration c. The solid circles are experimental points from the same mica sheets. The vertical bar at 10-l mol dm-, KNO, indicates the range within which ly, must lie (see text). The broken line indicates the trend of the data. mol dm-, KNO,; 0, x , + , 0,4 x At these lower concentrations the results are in substantial agreement with those of previous investigations.’> *- At 4 x mol dm-, KNO, however, the force (and hence vd) is considerably lower. The figure shows results from consecutive approaches, while the continuous line is a theoretical DLVO interaction based on eqn (1). The parameters of this curve are the Hamaker constant of mica, A = 2.5 x J,’ vd = 32 mV and 1 / ~ = 20 A.On compression there is an indication of a small jump into a (primary) minimum and the emergence of an additional force which rises sharply on further compression (dashed line in fig. 1). On decompression there is a reproducible outward jump (ca. 2500 A). These observations appear to be consistent with the presence of ‘ hydration forces’ operating at short distances. These forces have been previously observed and attributed to solvent (water) structuring around cations adsorbed at the mica-water interface; they manifest themselves particularly at higher electrolyte concentrations and have been extensively investigated by Pashley and Israelachvili.2y Shown as an inset to fig. I are the calculated vd values as a function of the electrolyte concentration.At 10-1 mol dm-, KNO,, however, there is a complete absence of long-range repulsion (as shown in fig. 2). As the surfaces are brought closer, there is no indication of any interaction until a separation of 150-200 A is reached, when a small attractive force becomes evident. On further approach the magnitude of this attraction increases, until a jump into ‘contact’ (strictly s eaking, a new equilibrium position) is recorded from a distance in the range 5Ck100 On separation (without further compression) there is1706 0 I E % 5 5 -100 -200 Forces between Mica Surfaces in Solution I I I I / f w/ I I I Fig. 2. Force us. distance profile between mica surfaces at lo-' mol dm-, KNO,. Each set of symbols corresponds to a different experiment (different pairs of mica sheets).The continuous lines are theoretical DLVO-type interactions for different t,ud values [eqn (l), see text] (a) 25, (b) 30 and (c) 35 mV. The broken line represents pure van der Waals attraction. The solid stars (*) were obtained in an experiment where no filter was used. an outward jump of 300-1000 A which appears to be mica-dependent. Both inward and outward jumps are reproducible within the same experiment, and have been observed systematically in a large number of experiments at this concentration. The 'contact' position associated with the inward jump is typically within ca. 20 A out from air contact. Further compression is met with a stiff repulsion, and longer outward jumps are recorded on decompression; although we did not investigate the behaviour in this range in detail, both observations are consistent with the presence of hydration force^.^ Each set of symbols in fig.2 represents a different experiment with different mica sheets. Although our results are in good agreement with those of earlier studies1? for KNO, concentrations 3 x and lop2 mol dm-3 there is a discrepancy at higher concentra- tions, since earlier investigations reported repulsive forces only (with a weak secondary minimum in 10-1 mol dm-3 KNO,), and found t,vd to be essentially independent of electrolyte concentration.l* 8 $ In order to check the possibility of artefacts arising from surface impurities inherent in our cleaning procedure, we have also varied the latter using hot concentrated NaOH instead of chromic acid.No change in the results could be observed. Likewise, different batches of KNO, and mica produced similar results, while atamic absorption analysis of our electrolyte solutions showed the concentration of Fe and Cu (likely contaminant ions) to be below the detection limit of 0.1 ppm. The possibility of surface impurities is in any case improbable since the results are quite reversible and reproducible as a function of electrolyte concentration. Thus within the context of the same experiment (i.e. the same mica sheets), cycles of increasing and decreasing concentration could be performed with reproducible results at each concentration. The origin of the discrepancy appears rather to be contamination arising from the use of GS-grade Millipore filters in our earlier investigations.8* These contain up to 5 wt% water-extractable materia1,ll including Triton X- 100, a non-ionic polyethoxy-type surfactant (octylphenol with an average of 10 ethoxy units per molecule, but with a broadC .Toprakcioglu, J . Klein and P . F. Luckham 1707 I o4 lo3 E 1 ?i n a: 9 ,02 10 I \ j\ x \ \ \ \ \ b\ \ y.7 \ I I \ I 100 200 DlA 10 Fig. 3. Force us. distance profile for two mica surfaces in 10-' mol dm-3 KNO, filtered through a standard (GS grade) Millipore filter, 0.22 pm. (a) First approach after 1 h (0). (6) Subsequent compression-decompression cycles (v, A, m). The crosses ( x ) represent experimental points from fig. 1 of ref. (8). distribution) employed as a wetting agent12. This surfactant is water-soluble, and surface-tension measurements (kindly performed by Dr J.Mingins) on the first three 200 cm3 portions of filtered water yielded values of 65.0, 69.9 and 69.2fO.l mN m-l, respectively, at 293 K. These results indicate a low (< 5 ppm) level of Triton X-100 present in the water.13 A similar surfactant, Triton X-405, is known to adsorb on to mica;14 thus it is probable that the repulsion below 100 A observed in previous studies is due to adsorbed Triton X-100. In this context we note that the toxic effect, on cultured cells, of Triton X-100 eluted from unwashed Millipore filters has been earlier remarked on by Cahn,12 and more recently the presence of surfactant contamination arising from the use of Millipore filters has been reported by Israelachvili and Pa~h1ey.l~ We have therefore repeated experiments at 0.1 mol dm-3 KNO, using GS-grade Millipore filters instead of the TF (Triton-free)-type filters which were used in the measurements shown in fig.2. The first 150-200 cm3 of filtered water was discarded and the second 200 cm3 portion was used to prepare the electrolyte solution. The force-distance profile for this solution was then determined (fig. 3). On a first approach [curve (a)] repulsive interaction commences at a surface separation of 140 A and increases with decreasing D. All subsequent compression or decompression of the surfaces results in a reduced range of interaction [fig. 3(b)]. These profiles show an exponential decay of force with distance, with a decay length of ca. 12 A, deceptively close to the Debye length one might expect at lo-' mol dm-3 KNO,.No strong attraction was observed, although a weak secondary minimum attraction (not shown on the logarithmic plot of fig. 3) was noticed, in agreement with results reported in our earlier studies. The onset separation for the observed interaction in fig. 3(b) is reasonable, given the size and polydispersity of the polyethoxy chain of Triton X-100.131708 Forces between Mica Surfaces in Solution Although the interpretation of these experiments is not straightforward this should not obscure the fact that they indicate a repulsion below ca. 8-10 nm (typically with an exponential decay length of ca. 12 A) in contrast to the experiments performed using TF filters, which clearly show attraction in this range (fig. 2). As an additional check we also carried out experiments (in 0.1 mol dm-, KNO,) in which no filter was used: these are shown as solid stars in fig.2, and are essentially identical to the force profiles obtained using TF filters. These results lead us to suggest that the repulsive forces previously reported at 10-1 mol dm-, KNO, in studies using non-TF grade Millipore filters were due to the interaction of adsorbed layers of surfactant. For example, a plot of FIR us. D from an earlier study,8 also shown in fig. 3 (crosses), which was obtained after ca. 1 h following addition of the electrolyte filtered through the non-TF Millipore filter, is close to the results of the present study. Surfactant adsorption could also account, at least in part, for the hysteresis effects, as well as the anomalously high apparent surface potentials occasionally observed in previous investigation^.^ The original force-distance measurements of ref.(1) are also similar to the profiles of fig. 3, and to our own previous studies,8j9 and thus differ qualitatively from the results obtained in the present work using TF filters or no filter at all (fig. 2). While we cannot be certain why this is, this discrepancy indicates that contamination was present in some of the original experiments of Israelachvili and Adams,' as also suggested to us by Dr Israelachvili.16 Finally, it is appropriate in this context to note that, in separate experiments,I4 addition of a high-molecular-weight polymer (polyethylene oxide) to a solution of Triton X-405 in 0.1 mol dm-, KNO, in which mica sheets had been incubated results in the desorption of adsorbed Triton from the mica surfaces, and its replacment by the polymer; subsequent force measurements are then characteristic of interactions between the adsorbed polymer layers 1 4 9 l7 Conclusion The forces between mica surfaces in aqueous KNO, solutions are well described by DLVO theory in the range 3 x to 10-1 mol dm-, KNO,.The potential of the diffuse part of the electrical double layer vd decreases as a function of electrolyte concentration from ca. 100 mV at 3 x lo-* mol dm3 to ca. 30 mV at 4 x mol dm-, KNO,. In contrast to all earlier studies in this electrolyte system, we find that at higher electrolyte concentrations wd c 30 mV, with the result that attractive van der Waals forces are dominant at distances from 200 A down to the range of hydration forces (ca.20 A). We conclude that the repulsive forces at surface separations below 100 A observed at 10-l mol dm-, KNO, in our previous investigations were likely to have been caused by adsorbed surfactant originating from the filters employed in these studies. We are grateful to Dr J. Mingins of F.R.I. Norwich, for carrying out the surface-tension measurements noted in the text and for helpful discussions. We thank Prof. D. Tabor for interest and encouragement, Prof. A. Silberberg for pointing out to us ref. (12), and particularly Prof. J. N. Israelachvili for many suggestions and discussions and for useful correspondence. We also thank the S.E.R.C. for support to C.T. References 1 J. N. Israelachvili and G. E. Adams, J. Chem. SOC., Faraday Trans. 1, 1978, 74, 975. 2 R. M. Pashley, J . Colloid Interface Sci., 1981, 80, 153. 3 R. M. Pashley, J . Colloid Interface Sci., 1981, 83, 531. 4 R. M. Pashley, J . Colloid Interface Sci., 1984, 102, 23. 5 R. M. Pashley and J. N. Israelachvili, J . Colloid Interface Sci., 1984, 97, 446. 6 B. V. Derjaguin and L. Landau, Acta Phys. Chim. USSR, 1941, 14, 633. 7 E. J. W. Verwey and J. Th. G. Overbeek, ' Theory of the Stability of Lyophobic Colloids' (Elsevier, Amsterdam, 1948).C. Toprakcioglu, J . Klein and P . F. Luckham 8 J. Klein and P. F. Luckham, Nature (London), 1982, 300, 429. 9 P. F. Luckham and J. Klein, J. Chem. SOC., Faraduy Trans. 1, 1984,80, 865. 10 B. V. Derjaguin, KolloidZ., 1934, 69, 155. 11 Membrane Technology (Millipore Corp., Mass., 1979). 12 R. D. Cahn, Science, 1967,155, 195. 13 E. J. Colichman, J. Am. Chem. SOC., 1950, 72, 4037. 14 P. F. Luckham and J. Klein, J. Colloid Interface Sci., in press. 15 J. N. Israelachvili and R. M. Pashley, J. Colloid Interface Sci., 1984, 98, 500. 16 J. N. Israelachvili, personal communication. 17 P. F. Luckham and J. Klein, Macromolecules, 1985, 18, 721. 1709 Paper 6/ 1330; Received 24th June, 1986