Adsorption of Butane-l,4-diol at the Hg-Aqueous Solution Interface Transition with Polarization between Two Ideal Adsorption Modelst BY FERNANDO PULIDORI, GIANNA BORGHESANI AND RODOLFO PEDRIALI Chemical Institute, Instrumental Chemical Analysis, The University, Ferrara, Italy AND ACHILLE DE BATTISTI AND SERGIO TRASATTI* Laboratory of Electrochemistry, The University, Via Venezian 21, 20133 Milano, Italy Received 2nd May, 1977 The adsorption of butane-1,4-diol on a polarized Hg electrode from both NaF and Na2S04 solutions has been studied by means of electrocapillary and capacity curves. Analysis has been carried out both at constant charge and at constant potential. Irrespective of the choice of the electrical variable, adsorption has been found to conform to congruent Langmuir isotherms on the negative side of the adsorption maximum, and to non-congruent isotherms on the other side.Results suggest that molecules adsorb flat on the surface. Non congruence may be described in terms of a saturation coverage linearly decreasing with electric field. The inner layer capacity at constant amount adsorbed has been determined up to saturation coverage over a wide range of electric field. Results are discussed in terms of a polarization-dependent adsorbate-solvent interaction on the surface. A detailed molecular model of the adsorption layer is suggested. Available data in the literature indicate that organic molecules adsorb with congruence 3-5 of isotherms either with respect to both charge and potential, or with respect to neither of the two electrical variables.Whereas the latter result is under- standable, it has been suggested 5* that the former is electrically unreasonable. Nevertheless, accurate experimental results '-lo suggest that it may in fact be possible. Damaskin,ll following up his early suggestion^,^* l2 has pointed out that simultaneous congruence is as a rule exhibited by substances with relatively high values of C1, the capacity at saturation coverage. The intrinsically low adsorbing power of such substances does not permit experimental data to be collected at very high coverages. Non congruence may thus be only apparent. It follows that the rationale behind simultaneous congruence with respect to both and E can only be discovered by improving the accuracy in collecting experimental data and by extending observations up to coverages close to saturation.Butane-l,4diol (BD) has been chosen because diols belong to the group of substances expected 2* l o p l3 to adsorb flat on the surface with apparent simultaneous congruence with respect to both electrical variables. Actually, the adsorption of this compound has already been investigated by Garnish et all0* l4 However, they restricted their analysis to concentrations ranging from 0 to 0.23 mol dm-3. Thus t Presented at the XI National Congress of the Italian Association of Physical Chemistry, S. Margherita Ligure, 9-11 December 1976. 7980 A D S ORP TI 0 N A T Hg-S 0 L U TI 0 N I N T ER FA C E the coverage was never above 0.7, which is probably too low to evidence non congruence, in the light of Damaskin’s arguments.’ ’ Furthermore, their data were derived only from electrocapillary curves, which are now known lP lS* l6 to give inaccurate results at positive rational potentials. EXPERIMENTAL Electrocapillary curves were determined for 3 1 concentrations, and capacity curves for 11 concentrations, of BD between 0.001 and 1 mol dm-3 in aqueous solutions containing 0.25 mol dm-3 NaF. Electrocapillary curves were also made for 9 concentrations of BD in 0.1 mol dm-3 Na2S04 aqueous solutions. This was done to investigate whether such a substitution of the supporting electrolyte, which improves the accuracy and reproducibility of electrocapillary data, is of any assistance in the diagnosis of the adsorption mechanism owing to the weak specific adsorption of sulphate i0ns.l’ Equipment and experimental procedure, including the account taken of capillary wear effects ** l8 in surface tension measurements have been described previously.7* l3 Hg was purified using standard procedures.BD was purified by four recrystallizations from the melt followed by double distillation in vacuu in the presence of anhydrous Na2S04. Solutions were prepared by weighing the desired amount of BD then adding the electrolyte and triply distilled water. Molar fractions (x) were calculated by assuming the density to be equal to 1 and independent of BD content. The maximum concentration investigated corresponds to about 0.02 molar fraction. This is sufficiently small for the solvent mixture to be assumed to follow Henry’s law.19-21 Deviations of the activity coefficient ofthe organic substance from unity can only affect the absolute value of AGZd, not the curvature of the (surface pressure, log x) curve.The maximum concentration is indeed so small that the medium effect on the activity of the supporting electrolyte is probably negligible.22* 23 The analysis was carried out at constant concentration of the supporting electrolyte. RESULTS Tables 1 and 2 summarize all relevant data for the solutions used in this work. Comparison of electrocapillary curves with twice integrated capacity curves showed that the calculated interfacial tension deviated from the experimental curve at both positive and negative rational potentials. Experiments with samples of different purity showed that the cathodic effect depended on the extent of purification. Small traces of impurities capable of depressing the negative desorption peak were also observed 8 * 2 4 3 2 5 with other substances.Electrocapillary curves on the other hand were apparently unaffected by the impurity content. Extensive and careful purifica- tions could not eliminate the effect completely. However, since in capacity curves there is no pronounced desorption peak at positive rational potentials, it is thought that discrepancies in that region are those usually observed and attributable to contact angle effects.15* l6 For the above reasons the analysis of data was carried out by relying only on electrocapillary curves at negative rational potentials, and on both sets of data at positive rational potentials.Procedures for the treatment of data were the same as those described in detail in previous paper^.^'^. 13# 26 Fig. 1 shows the zero charge potential shift upon adsorption of BD. As usually ob~erved,~-~* 26 data from electrocapillary curves in NaF lie 10 to 15 mV more negative than those found with the streaming electrode. This is due to some depression of the positive branch of the electrocapillasy curve.” l6 With Na,SO, the latter effect is less serious and the two sets of data are closer to one another. The shape of the dependence of AE, = on x,,, may give some qualitative indication of the nature of the isotherm. Substances obeying a Frumkin isotherm with particle- particle attraction have been found * * 9* 26 to exhibit on such plots the typical 27PULIDORI, BORGHESANI, PEDRIALI, DE BATTISTI AND TRASATTL 81 TABLE 1 .-DATA FOR BUTANE-1 ,q-DIOL SOLUTIONS USED IN THIS WORK. SUPPORTNG ELECTRO- 0 0.001 0.006 0.010 0.013 0.0165 0.02 0.022 0.028 0.032 0.038 0.048 0.06 0.07 0.08 0.09 0.1134 0.120 0.140 0.143 0.170 0.175 0.201 0.24 0.27 0.305 0.38 0.420 0.500 0.650 0.802 1 .Ooo 0 0.017 97 0.1078 0.1798 0.2338 0.2968 0.3598 0.3959 0.5040 0.5762 0.6845 0.8653 1.083 1.264 1.446 1.627 2.054 2.175 2.541 2.596 3.092 3.184 3.664 4.387 4.947 5.602 7.01 8 7.780 9.317 12.25 15.29 19.36 0.4300 - - - - - - - - - - - - - - - 0.3850 - - 0.3770 - - 0.3600 0.3430 0.3316 0.3410 0.3234 0.3300 0.3134 0.3180 0.3100 - 23.45 - - - - - - - - - - - - - - I 17.92 - - 16.96 - - 16.38 16.47 15.53 14.30 15.38 14.04 14.60 12.64 12.72 - 426.7 425.5 425.8 425.9 425.5 425.3 425.5 424.3 424.2 423.7 423.5 423.1 422.5 422.4 419.3 421.5 419.S 420.1 417.0 419.0 417.2 417.4 416.7 415.1 41 5.0 413.4 413.3 41 1.2 410.2 408.2 406.5 403.7 0.4439 0.4457 0.4401 0.4428 0.4368 0.4281 0.4308 0.4245 0.4325 0.4253 0.4216 0.41 35 0.4080 0.4120 0.4060 0.4041 0.41 30 0.391 1 0.3304 0.3857 0.3579 0.3798 0.391 9 0.3785 0.3 674 0.3689 0.3544 0.3573 0.3543 0.3575 0.3453 0.3333 S-shaped pattern.Fig. 1 suggests that BD may obey either a Langmuir isotherm or a Frumkin isotherm with particle-particle repulsion. Plots of surface pressure against CT or E showed that the adsorption maximum is probably concentration dependent. A more sensitive approach is the derivation of TABLE 2.-DATA FOR BUTANE-1,4-DIOL SOLUTIONS USED IN THIS WORK.SUPPORTING ELECTRO- LYTE, Na2S04 0.1 rnol dm-3 0 0.006 0.02 0.05 0.09 0.14 0.17 0.305 0.5 0.8 427.0 426.2 425.4 423.1 421.5 419.7 417.8 414.1 410.6 407.0 0.4383 0.4325 0.4290 0.41 56 0.3904 0.3762 0.3584 0.3561 0.3319 0.320682 A D S ORP T I 0 N A T Hg-SO L U T I 0 N INTERFACE omax and Emax from the coordinates of the intersection point of charge against potential curves.2* l 2 Fig. 2 shows the concentration dependence of omax. Its value tends to about -3.7 to -3.8 pC cm-2 for 8 -+ 1, whereas for 8 -+ 0, to a first approximation, amax was taken as - 3 pC cm-2. This value was successively refined to -2.8 pC cm-2 on the basis of additional data as described below. Accordingly, the value of Emax was found to shift from -0.56 V at low coverage to about -0.59 V - 5.0 7 -4.0- $ i3 f -2.0- x - 1.0- t - *-:- - -9- - -*- - Q-.-- -O- ---? -3.0*3* 0 1 I 1 0.2 0 .4 0.6 0.E BD molar fraction, x x lo2 FIG. 1 .-Zero charge potential shift of a Hg electrode upon adsorption of BD from aqueous solutions of different electrolytes. (0, A) 0.25 mol dm3 NaF, electrocapillary maximum ; (0) 0.25 rnol dm-3 NaF, streaming electrode ; ( x ) 0.1 mol dm-3 Na2S04, electrocapillary maximum. at surface saturation. Around the adsorption maximum isotherms are congruent with respect to neither of the electrical variables. Constancy in omax or Emax is a necessary even though not sufficient condition for congruen~e.~~ Congruence was tested to a first approximation by superimposition of surface pressure against log x plots.2* Congruence was found reasonably to be good at each charge and potential.However, preliminary calculations on the assumption of 3 FIG. 2.-Concentration dependence of the charge of maximum adsorption of BD on a Hg electrode. Base solution : (0) 0.25 rnol dm-3 NaF ; (A) 0.1 rnol dm-3 Na2S04.PULIDORI, BORGHESANI, PEDRIALI, DE BATTISTI A N D TRASATTI 83 coverage, 0 FIG. 13.-Change in inner layer potential drop at a Hg-aqueous solution interface with coverage with adsorbed BD. Base solution : (0) 0.25 mol dm-3 NaF, capacity data ; ( x ) 0.25 mol dm-3 NaF, electrocapillary data ; (A) 0.1 mol dm-3 Na2S04, electrocapillary data. Difference between linear and non linear behaviour is emphasized. congruence with respect to charge showed that the potential shift at charges between - 2 and + 2 undoubtedly exhibited some non-linear variation with coverage.This is shown in fig. 3. This was thought to indicate '* lo non congruence at these charges I I 1 1 I I I I I t I I 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 0 BD molar fraction, x x 10' (6) constant potential. FIG. 4.-Test of the Langmuir isotherm for BD adsorption on a Hg electrode. (a) Constant charge;84 ADSORPTION AT Ha- 0 SOLUTION INTERFACE and possible congruence at more negative charges where plots were strictly linear up to 0.9 coverage. The nature of the isotherm was investigated by determining surface excesses by differentiation of surface pressure against log x plots.28 In the range of congruence, a Langmuir isotherm is followed with rs = 4.1 mol cm-’. Plots of x l r against x, as a test for this isotherm,’ showed a strictly linear behaviour (fig.4). The value found for rs corresponds to a covered area of about 0.41 nm2 per molecule. This value is in good agreement with the area projected by a molecule of BD lying flat on the surface ’’ and is also consistent with the value of 0.31 nm2 found for ethylene Acid was derived by calculating the theoretical surface pressure curve with rs = 4.1 x rnol cm-2 and fitting to the experimental points at the various charges. Results showed that at charges more positive than -3 ,uC the surface pressure curves flatten down slightly. This could be recognized in this work only because results extend to high values of @. At low and intermediate values of <D the coiidition of congruence may be taken as apparently met and this explains the findings of’ Dutkiewicz et a!.,1° who ineasured @ up to 1.2 pJ cm-2 whereas here CP values up to 2.4 pJ cm-2 at the point of maximum adsorption were measured.each additional CH2 group contributing N 0.05 nm2. (TI& cm-2 FIG. 5.-Charge variation of the surface saturation concentration in the Langmuir isotherm for adsorption of BD on a Hg electrode. Flattening of the surface pressure curve may indicate either particle-particle repulsion, or decrease in I-‘,. Tests of the Frumkin isotherm at +2 pC cm-2 with rs = 4.1 x 10-lo actually suggested a positive value for the interaction parameter which, according to the isotherm in the form : corresponds to particle-particle repulsion. The tests of the Langmuir isotherm shown in fig.4 suggest that this isotherm is probably also better followed at positive changes with rs decreasing with charge becoming more positive. Fig. 5 shows that I‘s decreases linearly with charge, as observed *. 26 with other organic substances. Fig. 6 shows the charge dependence of AG& as found by superimposing surface pressure curves at different 0 and E to those at cmax and Em,,, respectively. The same features as those reported by Dutkiewicz et aL1* can be recognized. At constant Ethere is no criterion by which one might assume non congruence at positive rational potentials. However, whereas for potentials negative to -0.56 V a strictlyPULIDORI, BORGHESANI, PEDRIALI, DE BATTISTI AND TRASATTI 85 0 I 2 3 4 (E- ~max)21V2 FIG. 6.-Change in standard free energy of adsorption of BD on a Hg electrode (a) at constant charge and (b) at constant potential.(0) Charges more negative than - 4 pC cm-2 ; potentials more positive (A) and more negative (0) than -0.55 V. ( . . . ) According to Dutkiewicz et a1.lo (- - -) Calculated according to eqn (6a) (see text). quadratic potential dependence 2* 30 of log p has been found, at more positive potentials the relationship is more complex and, in any case, the points lie above the straight line. This confirms the findings of Dutkiewicz et aPO Values of AGtd at charges more positive than -4 pC cm-2 were derived by fitting surface pressure curves calculated with the given value of rs to the experimental t I 1 I t I I I lo 5 0 -5 -10 -15 u/pC cm-2 FIG. 7.-Charge dependence of the standard free energy of BD adsorption on a Hg electrode.(- - -) According to the same quadratic dependence as at negative charges. (0) Experimental points.86 AD SORP TI 0 N A T Hg-S 0 LU T I 0 N I NTER F A CE points. Fig. 7 shows that AG,Od decreases quadratically with charge for CT more negative than -4, whereas on the other side it decreases more slowly. Assuming strict congruence with respect to charge at more negative values than -4 pC cm-2, values of r at other charges were calculated from the Langmuir isotherm with the equation : by introducing at each charge the appropriate values for Ts and p. Fig. 8 shows the variation!of AYq5 with coverage at constant charge. This plot supports the view that isotherms are congruent at 0 < - 4 pC cm-2 and not congruent at CT > - 4 pC cm-2.The final value of -2.8 pC cm-2 for omax was derived by plotting the slopes of the straight lines in fig. 8 as a function of CT. Fig. 9 shows that the slopes change r = [Bxor,/(l +Pxor,>lrs (2) L - - * * A L*. - ' - * * h e . a . * - . - 4 - 0.2 I- - 0.3 - 0.4 - I 0 -1 4 -0.9 - I I I I I I 2 3 4 BD surface concentration, I'x 10lO/mol cm-2 FIG. 8.-Change in inner layer potential drop of a Hg electrode upon adsorption of BD from aqueous solutions. Base electrolyte : (0) 0.25 mol dm-3 NaF, capacity data ; (0) 0.25 mol dm-3 NaF, electrocapillary data ; (A) 0.1 mol dm-3 Na2S04, electrocapillary data. Figures by the fines indicate the charge on the metal. The arrow indicates the position of the adsorption maximum.PULIDORI, BORGHESANI, PEDRIALI, DE BATTISTI AND TRASATTI 87 0.2 - I c( 8 k -0.1 - 3 2 0.1- *i 0 - > \ h -0.2- W -0.3 - 0 -0.4 I 5 0 -5 -10 - 15 o/pC cm-2 according to eqn (3) from fig.6(a). The arrow indicates the position of omax. FIG. 9.--Initial slopes of the lines in fig. 8 plotted as a function of charge. (---):Calculated ti E 0 Y - b 0.2 v 0. I 0 - -0.2 -- -5 I 2 3 4 BD surface concentration, r x 101O/mol cm-2 FIG. 10.-Change in charge at constant inner layer potential drop upon adsorption of BD on a H g electrode. The arrow indicates the position of am=.88 ADSORPTION AT Hg-SOLUTION INTERFACE linearly with Q at charges more negative than -4. The linear portion in fig. 9 is quantitatively related to the slope in fig. 6(a) by the equation :30* 31 The straight line in fig.9 has been calculated with eqn (3) from the slope of the straight line in fig. 6(a). Points for 0 > -4 have been derived from the initial slope of the curves in fig. 8. Since :309 31 a26y#prao = - 2.3 m ( a 2 log p/aa2)). (3) ( a ~ y $ p r ) , = -2.3 R T ( ~ log ppQ) (4) the lower values for aAy4/aa are an indication of lower slope of the charge dependence of AG,Dd at the given charge, which is in agreement with the results in fig. 7. 0 0.03 1 I I I 2 I 3 4 I BD surface concentration, r x 101o/mol cm-2 FIG, 11.-Reciprocal of the inner Iayer capacity at constant amount adsorbed plotted as a function of BD surface concentration. Figures indicate the charge on the metal. Fig. 10 shows the change in B as a function of r at constant inner layer potential drop.Plots are apparently strictly linear between -0.7 and -0.2 V whereas at -0.1 V (and probably at 0 V) corresponding to charge around -3.5 pC there is some evidence for non congruence. This is expected from the concentrationPULIDORI, BORGHESANI, PEDRIALI, DE BATTISTI AND TRASATTI 89 TABLE 3 .-PARAMETERS OF BUTANE-1 ,~-DIOL ADSORPTION ON POLARIZED Hg ELECTRODES FROM AQUEOUS SOLUTIONS OF NOT SPECIFICALLY ADSORBED ELECTROLYTES this work ref. (10) isotherm oniax rs Langmuir -2.8 pC cm-2 (r 3 0) to - 3.7 (r+ rs) 4 . 1 x 10-lo mol cm-2 (a neg. to - 3 pC cm-2) 4.08 x 10-lo mol cm-2 (a = - 3) - 3.71 kcal mo1-1 - 3.69 kcal mo1-l 0.0076 p C 2 cm4 (a neg. to - 3) 1 . 9 4 V-2 (E neg. to -0.55 V) 0.14 to 0 . 1 5 V 11.7f 0.3 pF C M - ~ [4.25-0.16(0+4)]~ 10-l' (0 POS. to -3) Langmuir - 2.5 pC cm-2 4.37 x mol cm-2 - - 3.72 kcal mol-1 0.0065 pC-' cm4 1.78 V-2 0 .1 (estd.) - - dependence of Em,,. At 0.1 and 0.2 V there is no evidence for any curvature and strict linearity may again be assumed. The inner layer capacity at constant amount adsorbed can be related straight- forwardly to the structure of the adsorbed layer. From fig. 8, values of AY4 were plotted as a function of charge at constant amount adsorbed and the resulting curves differentiated at constant charge. l 2 Results are shown in fig. 11 in the form of plots I01 I I I I I 2 3 4 BD surface concentration, r x 101O/mol cm-2 FIG. 12.-Inner layer capacity at constant amount adsorbed plotted as a function of BD surface concentration. Figures indicate the inner layer potential drop.90 A D SORPTION A T Hg-S 0 L U TI 0 N I NT ERF A CE of l/Ci against r.All curves converge towards a common value of Ci at r = Ts. This is the capacity for a monolayer of BD and its value has been found to be 11.7 _+ 0.3 pF cm-2. Table 3 summarizes all the adsorption parameters. Results of Dutkiewicz et are also reported for comparison. Linear relationships, as expected from congruence with respect to charge, are found for charges between -4 and - 12. At charges more positive than -4 the plots are non linear. In order to investigate congruence with respect to potential, values of Ci at constant A:+ were derived by differentiating charge against potential curves at a constant amount adsorbed, as obtained from fig. 8. Results are shown in fig. 12.At potentials between 0 and 0.2 V linear relationships between Ci and r may be assumed. Between -0.1 and -0.3 V plots are certainly non linear whereas for more negative potentials, while they are still non linear, the plots depart only very slightly from linearity. Results in fig. 12 may explain the apparent simultaneous congruence of isotherms with respect to both electrical variables at negative charge and potentials. However, they also indicate that plots like those in fig. 10 are much less sensitive than those in fig, 8 so that the former can hardly be used as a criterion for non congruence. This is implicit also in the work by Dutkiewicz et al." It is easy to show that the small curvature in the plots of Ci against r at potentials more negative than -0.2 V is a consequence of an almost undetectable curvature in the plots in fig.10 at the same AY4. Scatter of experimental points may well obscure this effect. DISCUSSION With reference to the plate capacitor formula, the adsorption of BD appears to be better described ' 9 2* 30 by a model at constant dielectric constant, E, on the negative side and by a model at constant thickness, d, of the adsorbate layer on the positive side of the adsorption maximum. The apparent congruence of isotherms over all the charge range as claimed by Dutkiewicz et is presumably to be related to the relatively low coverages reached in that investigation. This work supports Damaskin's continual caution 4* '* l 2 regarding the reliability of simultaneous congruence with respect to both o and E.The prerequisite for either of the models to hold throughout is that the solvent behaviour should be field independent.12 In reality Co, the capacity of the interfxe at 9 = 0, almost doubles from the extreme negative end to the extreme positive end of the range investigated. Thus, Cl/Co = 0.7 at strong negative charges and the decrease in capacity upon adsorption may entirely be accounted for by a change in thickness, with dH20 = 0.28 and dsD = 0.42 nm. This implies that EBD N &H20. Actually, the bulk dielectric constants of the lower diols 32-34 are comparable with, but smaller than that of water.35 However, the high frequency dielectric constant (more relevant here) of propylene-1,2-diol 34 is only slightly less than E for water.35 It is expected that the same is true of BD.Co increases as the charge becomes more positive as a result of increase in orientation polarizability of the solvent.36 Thus, the effect of decrease in E progres- sively prevails over the effect of increase in thickness and the model describing adsorption turns from one of two series capacitors to one of two parallel capacitors. The non-linear dependence of AY4 on r in fig. 8 at positive charges is evidence for strong reorientation of the molecules. This is in fact implied in the two parallel capacitor 37 In the present case, reorientation of water molecules upon BD adsorption is thought to take place as their position with respect to the adsorbate molecules becomes critical as a consequence of rotation under the action of the fieid.9. 36PULIDORI, BORGHESANI, PEDRIALI, DE BATTISTI AND TRASATTI 91 The curvature of the quadratic dependence of Actd on charge or potential is a measure of the difference in polarizability per molecule of adsorbate between organic molecule and solvent.The following relationships hold :30 (co - c,) = 2.3 RTr,(a2 log pjaE2) (1 / C , - 1 /Co) = 2.3 R T rS(a2 log P/aa2) ( 5 4 (5b) for the two parallel capacitor model, and : for the two series capacitor model. Since in the range of AYq5 0.2 to -0.1 V the mean value of Co is - 30, eqn (5a) predicts 3.8 V-2 for the slope of the steeper straight line in fig. 6(b). This value fits satisfactorily to the experimental points which, however, are more likely to follow a curved line. Conversely, at potentials more negative than the adsorption maximum the mean value of Co is 20 pF cm-2.Eqn (5a) predicts in this case a slope of 1.7 V-2 whereas that experimentally observed is 1.94 V-’. Accordingly, eqn (5b) predicts for the curvature of the (A&, charge) relationship at strongly negative charges the value of 0.0073 p C 2 cm4, while the observed value is 0.0076 P C - ~ cm4. At charges more positive than a,,, the predicted mean slope would be 0.0108 whereas in fact it may be at most even lower than the former. The charge at maximum adsorption is -2.8 at low 8 shifting to about -3.7 at high coverage. Consideration of molecular polarizability predict * a charge of maximum adsorption of - 1.7. A normal component of fixed dipoles with the positive end towards the electrode can explain the observed value of om,,.Molecules are thus thought to lie flat on the surface with the two OH groups slightly turned towards the solution. This view is supported by the small positive value of 0.07 to 0.08 contributed by adsorbate molecules to the adsorption potential shift at CT = 0 amounting (fig. 3) to 0.14 to 0.15 V, 0.07 V of which are due to displacement of oriented water molecules. 39 BD adsorption according to a Langmuir isotherm is understandable in terms of the complex nature of the interaction parameter A in eqn (1). In adsorption from solutions particle-particle interactions occur via solvent molecules occupying other- wise free sites. Thus, it is possible to write : where Go (negative quantities) are interaction free energies and s and P stand for solvent and particle, respectively.Since the adsorption layer actually is a two- component fluid, A 2 0 corresponds to complete miscibility on the surface, and A < 0 to partial miscibility. The difference between A = 0 and A > 0 may be described in terms of a disordered or an ordered structure of the interfacial layer, respectively. In the limit, the phenomenon of surface condensation40 may be observed : this is the experimental manifestation of complete immiscibility. In terms of eqn (6), constancy of A over all the charge range means that possible variations in the various terms compensate almost completely. When the charge becomes more positive than -4 pC cm-2, corresponding to about the position of zero net dipole orientation for water molecules,36 the position of the latter tends to favour stronger interaction with the adsorbate due to the opposite orientation of the molecular dipole^.^ A is thus expected to shift from zero to increasingly positive values.A decreasing value of rs, as actually found experimentally, implies that fewer molecules can be accommodated on the surface at saturation ; this, qualitatively, corresponds to the onset of some If BD binds surrounding H,O molecules more strongly than they are bound to other solvent molecules, the former become a part of the adsorbate and are no longer replaced by other BD molecules. A = -(2GLp-Gg-p-GLs)/RT (6)92 ADSORPTION AT Hg-SOLUTION INTERFACE This effect is expected to increase as the charge becomes increasingly positive. The higher value of AGld at positive charges may be understood consistently.The authors are grateful to Dr. R. Parsons for the use of his computer programs. Financial support by the National Research Council (C.N.R., Rome) is gratefully acknowledged. R. Parsons, Rev. Pure Appl. Chem., 1968,7, 91. S . Trasatti, J. Electroanulyt. Chem., 1974, 53, 335. R. Parsons, Proc. Roy. SOC. A, 1961,261,79. S. K. Rangarajan, J. Electroanalyt. Chem., 1973, 45, 279. S. K. Rangarajan, J. Electruunalyt. Chem., 1973, 45, 283. B. A. Abd-El-Nabey and S. Trasatti, J.C.S. Faraday I, 1975,71, 1230. A. De Battisti, V. Faggiano and S. Trasatti, J. Electruunalyt. Chem., 1976, 73, 327. lo E. Dutkiewicz, J. D. Garnish and R. Parsons, J. Electruanulyt. Chem., 1968, 16, 505. l1 B. B. Damaskin, Elektrukhim., 1975, 11, 428.l 2 A. N. Frumkin, B. B. Damaskin and A. A. Survila, J. Electronalyt. Chem., 1968, 16, 493. l3 S. Trasatti, J. Electruunalyt. Chem., 1970,28, 257. l4 J. D. Garnish, Ph.D. Thesis (University of Bristol, 1966). l 5 J. Lawrence, R. Parsons and R. Payne, J. Electruanalyt. Chem., 1968, 16, 193. l6 S. Trasatti, J. Electruanalyt. Chem., 1971, 31, 17. R. Payne, J. Electruanalyt. Chem., 1975, 60, 183. l8 A. De Battisti, R. Amadelli and S. Trasatti, J Culluid Interface Sci., 1978, in press. l9 D. M. Mohilner, L. W. Browman, S. J. Freeland and H. Nakadomari, J. Electruchem. SOC., 2o D. M. Mohilner and H. Nakadomari, J. Electruanalyt. Chem., 1975, 65, 843. 22 A. De Battisti and S. Trasatti, J. Electruanalyt. Chem., 1974, 54, 1. 2 3 B. A. Abd-El-Nabey, A. De Battisti and S. Trasatti, J. Electruanalyt. Chem., 1974, 56, 101. 24 N. B. Grigoryev, S. A. Fateev and I. A. Bagotskaya, Elektrukhim., 1972, 8, 583. 25 K. Doblhofer and D. M. Mohilner, J. Phys. Chem., 1971, 75, 1968. 26 A. De Battisti, B. A. Abd-El-Nabey and S. Trasatti, J.C.S. Faraday I, 1976, 72,2076. 4B. B. Damaskin, Electrokhim., 1969, 5, 771. ’ A. De Battisti and S. Trasatti, J. Electruunalyt. Chem., 1973, 48, 213. 1973, 120,1658. J. E. B. Randles, B. Behr and Z. Borkowska, J. Electroanulyt. Chem., 1975, 65, 775. A. N. Frumkin and B. B. Damaskin, Modern Aspects ofElectrochemistry, ed. J. O’M. Bockris and B. E. Conway (Butterworth, London, 1964), vol. 3, p. 149. 28 R. Parsons, Modern Aspects of Electrochemistry, ed. J. O’M. Bockris (Butterworth, London, 1954), p. 103. 29 Interatomic Distances. Supplement (The Chemical Society, London, 1965). 30 R. Parsons, J. Electruanalyt. Chem., 1963, 5, 397. 31 R. Parsons, Trans. Faruday SOC., 1959,55,999. 32 J. Wumbel, F. Jona and E. P. Scherrer, Helv. Phys. Acta, 1959, 39, 412. 33 C. P. Smyth and W. S. Wells, J. Amer. Chem. SOC., 1932, 54, 2261. 34 D. W. Davidson and R. H. Cole, J. Chem. Plzys., 1951, 19, 1484. 35 J. B. Hasted, Water. A Comprehensive Treutise, ed. F. Franks (Plenum Press, New York- 36 S. Trasatti, J. Electrounalyt. Chem., 1975, 64, 128. 37 B. B. Damaskin and A. N. Frumkin, J. Electrunulyt. Chem., 1972, 34, 191. 38 S. Trasatti, Ext. Abstrs., 27th ISE Meeting (Zurich, 1976), p. 207. 39 A. De Battisti and S. Trasatti, Croat. Chem. Actu, 1976, 48, 607. 40 S. L. Dyatkina, B. B. Damaskin, N. V, Fedorovich, E. V. Stenina and V. A. Yusupova, London, 1972), vol. 1, p. 255. Elektrukhim., 1973,9, 1283. (PAPER 7/735)