Faraday Discuss. Chem. SOC., 1986,81, 95-106 A Single Spectroscopic Probe for the Determination of Both the Interfacial Solvent Properties and Electrostatic Surface Potential of Model Lipid Membranes Calum J. Drummond,* Franz Grieser and Thomas W. Healy Colloid and Surface Chemistry Group, Department of Physical Chemistry, The University of Melbourne, Parkville, Victoria 3052, Australia 2,6-Diphenyl-4-(2,4,6-triphenyl-l-pyridinio)phenoxide, &(30), has been investigated in order to ascertain its suitability as a probe for both the effective interfacial dielectric constant ( ceE) and the electrostatic surface potential (&,) of model lipid membranes in aqueous solution. This work establishes that the solvatochromic visible absorption band for b ( 3 0 ) can be used to provide a good estimate of the eeff for cationic micelles.It is also shown that the acid-base dissociation of E-,-(30) can be utilized to obtain a quantitative measure of the I/+, in the case of cationic micelles. There are problems and uncertainties associated with the use of E,,-(30) in aqueous solutions of other types of charged self-assembled surfactant aggregates, and these are discussed. The interfacial region between an aqueous solution and a self-assembled lipid phase possesses physicochemical properties which are, in general, dissimilar to both that of bulk water and the interior of the lipid self-assembled unit. The importance of the interfacial region to the function of biological membranes has long been recognized. Indeed, in order to fully understand phenomena such as the surface reactions which occur at or in biological membranes, the transport of species across membranes, and the adsorption of species onto membranes it is necessary to have an intimate knowledge of the nature of such interfacial microenvironments.Consequently, a wide range of spectroscopic probes have been employed to investigate the solvent properties (effective dielectric ~0nstantsl-l~ and m i ~ r ~ ~ i ~ ~ ~ ~ i t i e ~ ~ ~ ~ ~ ~ ~ - ~ ~ ) of the interfacial microenvironments and to determine the electrostatic surface potential^^'*^-^^ of model lipid membranes. Unfortunately, one of the spectroscopic methods currently employed which relies on the use of acid-base indicators to calculate the electrostatic surface potential generated at a charged i n t e r f a ~ e ~ - ~ ~ - ~ ~ also involves some ~ncertainty.~~ If the influence of specific molecular interactions and salt effects on the acid-base equilibrium of an interfacially located acid-base indicator can be neglected, the observed pK, (pKZbs) is dependent upon two principal factors, namely the lower effective dielectric constant ( E , ~ ) of the interfacial region and the electrostatic field at the surface.2 This dependence is shown in the following relationship: 2926-33,38339 where pKZb" is the apparent pK, value for the acid-base indicator at the charged surface, pKL is the apparent pK, value for the indicator at the interface if the surface potential (&) is zero and F, R and T are the Faraday constant, the universal gas constant and the absolute temperature, respectively.In the absence of any extra contribution from specific molecular interaction andlor salt effects to the acid-base equilibrium at the interface, the magnitude of the pKg value for an acid-base indicator is believed to be 9596 Probe for the Surface Region of Model Membranes directly related to the E , ~ value characterizing the interfacial microenvironment.2 Until now, since there are two unknowns in eqn (1) (pKa and t,bo), it has been the practice to assume, for micellar systems at least, that the pKa value can be equated with the pKZbs value determined for the acid-base indicator in non-ionic micelles of surfactants with poly( ethylene oxide) h e a d g r o ~ p s . ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ However, for a number of acid-base indicators this latter assumption is highly q ~ e s t i o n a b l e .~ ~ In addition, in the case of the acid-base indicators, such as 4-heptadecyl-7-hydroxycoumarin and 4-octadecyloxy-l- naphthoic acid, where it does appear to be an acceptable assumption to make for charged micelles, it is not clear whether it is also valid for charged vesicles.27 To avoid the uncertainty in the value of pKa and consequently Go, we have investi- gated the use of 2,6-diphenyl-4-(2,4,6-triphenyl-l -pyridinio)phenoxide, ET(30), in a dual role as a probe for both the effective interfacial dielectric constant and the electrostatic surface potential of model lipid membranes. The molecule ET( 30) possesses a phenolic oxygen atom capable of being protonated, with a pK, value in the middle region of the pH scale, and a strongly solvatochromic ( T + T") absorption band with intramolecular charge-transfer c h a r a ~ t e r .~ ~ Zachariasse et aL7 have previously utilized the solvato- chromic absorption band of ET(30) to estimate the ceff values for the interfacial microen- vironments of a number of micelles, vesicles and microemulsions. In the present work we have assumed that the E , ~ value obtained from the solvatochromic band of the ET(30) visible absorption spectra can provide the appropriate organic solvent- water mixture to best approximate or calibrate the solvent properties of the interfacial region of a charged surface. It is shown how the pK, value obtained for ET(30) in this organic solvent-water mixture can then be converted to a pKL value and used in eqn (1). We have applied this type of analysis to a number of aqueous solutions of micelles and vesicles and have demonstrated that this procedure with ET(30) can provide a quantitative measure of the electrostatic surface potential at a number of these charged interfaces.Experimental The sample of ET(30) was generously supplied by Prof. C. Reichardt, University of Marburg. The surfactants came from a variety of sources. Hexadecyltrimethylammonium bromide (CTAB), hexadecyltrimethylammonium chloride (CTAC), dodecyltrimethyl- ammonium bromide (DTAB), dodecyltrimethylammonium chloride (DTAC), dodecyl- benzene sulphonate (DBS) and hexadecylpyridinium bromide (CPB) were all purchased from Tokyo Kasei. Specially purified sodium dodecyl sulphate (SDS) and oleic acid and Brij-35 were obtained from B.D.H.Didodecyldimethylammonium bromide (DDDAB), dihexadecyldimethylammonium bromide (DHDAB) and dioctadecyl- dimethylammonium bromide (DODAB) were kindly given to us by Dr J. Brady and Prof. D. F. Evans, University of Minnesota. Prof. B. W. Ninham, Australian National University, generously provided us with dodecylethyldimethylammonium bromide (DEAB) and dodecylbutyldimethylammonium bromide (DBAB). The Nikko Chemical Co. was the source of n-dodecyloctaoxyethylene glycol monoether (C,2E8). Dihexadecylphosphate (DHP) was obtained from Sigma. L-(Y -Dimyristoylphos- phatidylcholine (DMPC) and L-(Y -dipalmitoylphosphatidylcholine (DPPC) were puriss grade from Fluka. All the solid cationic and anionic surfactants were recrystallized prior to use.The rest of the surfactants were used as received. The inorganic reagents, NaCl, NaBr, NaOH and HCl, were all analytical grade. Tetraethylammoniu'm chloride (TEAC) and tetramethylammonium chloride (TMAC) were purum grade from Fluka. These reagents were all used without further purification. The aqueous solutions were prepared with Millipore filtered water (conductivity < 1 x lop6 LR-' cm-' at 25 "C). The 174-dioxane was U.V. spectroscopic grade from Fluka and was passed through an aluminium oxide column (active neutral Brockmann grade 1C. J. Drummond, F. Grieser and T. W. Healy 97 from B.D.H.) immediately before use. This served the dual function of removing residual water and also any peroxides present in the 174-dioxane.Spectrosol-grade methanol and chloroform from Ajax Chemicals were used as received. The solid crystalline form of ET(30) was difficult to dissolve in pure water and in aqueous surfactant solution. Therefore, a small aliquot (< 1 wt YO of the total surfactant solution) of a stock solution (50: 50 wt O/O methanol : water) of E,(30) was added to the aqueous surfactant solutions so that the concentration of ET(30) was ca. 5~ lop5 mol dm-3. Unilamellar vesicle dispersions of DDDAB, DHDAB, DODAB, DMPC, DPPC and DHP were produced by following established procedure^.^.'^^^^ These procedures have been outlined in an earlier p~blication.~' The aliquot of ET(30) stock solution was always added upon the completion of the vesicle preparation. All experiments, unless otherwise stated, were performed at 25 "C.The u.v.-visible absorption spectra were measured on a Varian Cary model 210 spectrophotometer with 1 cm quartz absorption cells. The method of performing the pH titrations in surfactant solution has been given elsewhere,l2 with the exceptions that in the present study HCl (rather than H2S04) was employed with NaOH to adjust the pH, and the lower compartment of the double-junction reference electrode contained a 1 x mol dmp3 aqueous solution of TEAC. While performing the pH titrations care was always taken to ensure that the ionic strengths of the solutions were not significant altered. The pH-meter reading in a 1,4-dioxane-water mixture is not a direct measure of the hydrogen- ion activity in the solution. Hence, for these experiments the procedure of Van Uitert and Haasa was followed to enable the determination of stoichiometric hydrogen-ion concentrations from the pH-meter readings.The calibration factors obtained were in reasonable agreement with previously calculated values and will be reported in a subsequent p ~ b l i c a t i o n . ~ ~ In pure water the positively charged protonated form of ET(30) was found to be completely soluble at a concentration of 5 x lop5 mol dmp3, while the zwitterionic phenoxide form of E,(30) was found to have a solubility limit of ca. 2 x lop6 mol dmp3. Consequently, for the pH titration of ET(30) in pure water, the solution was first filtered at pH 10, pH 6 and then pH 10 again, in that order, to ensure the complete removal of any undissolved ET(30).An example of the change with pH in the u.v.-visible absorption spectrum of the solvatochromic band of ET(30) is shown in fig. 1. Not shown in fig. 1 is the isosbestic point, which occurs at a shorter wavelength (334nm in pure water). The spectra of fig. 1 are representative of the type of changes observed in the u.v.-visible absorption spectrum of ET(30) as a function of pH in pure water, 174-dioxane-water mixtures and every aquebus micellar and vesicular solution studied in this work with the exception of the SDS, DBS, DHP and oleate solutions. Fig. 2 illustrates the kind of spectra obtained for ET(30) when the pH was varied in SDS, DBS, DHP and oleate solutions. The pK, values for ET(30) in the various media investigated were determined from the spectra with the aid of eqn (2): where a is the percentage of E,(30) phenolic hydroxy groups that have been ionized.Values for a were calculated as a function of pH at the solvatochromic band maximum, A,,,, with a fixed concentration of ET(30). The maximum absorbance obtainable for ET(30) at A,,,, by varying the pH of the solution, was taken as the 100% ionization value and the minimum absorbance as the 0% ionization value. At least six different a values between zero and 100 were examined for each 174-dioxane-water mixture or self-assembled surfactant solution. This protracted procedure established that the PK:~" values for ET( 30) residing in the interfacial microenvironments of the lipid self-assembled98 Fig. 1. U.v.- Probe for the Surface Region of Model Membranes 0.L 0.3 8 2 2 0.2 3 rd 0.1 0 LOO 5 00 600 700 wavelength/nm -visible absorption spectrum of 7.2 x mol dmP3 E,,-(30) in a 50 wt % water mixture ( E = 35.85) as a function of pH at 25 "C. The structure of b ( 3 0 ) is also shown. wavelength/nm Fig. 2. U.v.-visible absorption spectrum of 5.0 x mol dmP3 b ( 3 0 ) in 0.069 mol dm-3 SDS solution as a function of pH at 25 "C. an aqueous phases were well defined. All the pK, values quoted in the present work are the average values calculated by using the results of the pH titrations and eqn (2). The magnitude of the error associated with each pK, value indicates the maximum deviation from the average value. Results Fig. 3 and 4 illustrate how the A, value for the solvatochromic band of ET(30) varies with the dielectric constant of a number of organic solvent-water mixtures and neat n-alcohols.The A,,, values in the isopropanol-water, ethanol-water, methanol-water,C. J. Drummond, F. Grieser and T. W. Healy 60 01 I I I I I I I 1 t t 1 dielectric constant 99 Fig. 3. A,,, values for the solvatochromic band of E-,-(30) in 0, isopropanol-water: M, ethanol- water; 0, methanol-water and 0, ethylene glycol-water mixtures as a function of the dielectric constants of the mixtures at 25 "C. dielectric constant Fig. 4. A, values for the solvatochromic band of b ( 3 0 ) in 0 , acetone-water mixtures; 9, 1,4-dioxane-water mixtures and 0 , neat n-alcohols as a function of the dielectric constants of the media at 25 "C.100 Probe for the Surface Region of Model Membranes dielectric constant Fig.5. ApKF (0) and ApKa (0) values for G(30) in 1,4-dioxane-water mixtures as a function of the dielectric constants of the mixtures at 25 "C. ethylene glycol-water, acetone-water and 1,4-dioxane-water mixtures were acquired from the works of Dimroth and Reichardt& and Kosower et al.47 The dielectric constants of the mixtures were interpolated from the data of Akerlof4* and Critchfield et al.49 The A,,, values in the neat n-alcohols were obtained from the compilation of Reichardt and Harbusch-GonertS0 and the dielectric constants for these neat solvents were taken from ref. (51) and (52). Also included in fig. 4 are our own results for 1,4-dioxane-water mixtures. We obtained a A,,, value of 454nm for ET(30) in pure water, which is in close agreement with the value of 453 nm found by Dimroth and R e i ~ h a r d t .~ ~ The changes in the pK, value for ET(30) in 1,4-dioxane-water mixtures relative to its pK, value in pure water (i.e. ApKr) are shown in fig. 5 as a function of the dielectric constant of the mixtures. The ApKF data points shown in fig. 5 refer to pure water and 10, 20, 30, 40, 50, 60, 70 and 80 wt '/o 1,4-dioxane-water mixtures. The values of Critchfield et al.49 for the dielectric constants of 1,4:dioxane-water mixtures were used for fig. 5. The pK, value found for ET(30) in pure water was 8.63k0.03. The explanation of how, in the absence of distortion of the acid-base equilibria by any specific molecular interactions and salt effects, the p K a values can be derived from the pKF values has already been clearly and comprehensively covered by Fernandez and Fromherz,2 and only a very brief account will be given here.The basic premise is that for a 1,4-dioxane-water mixture which has a dielectric constant identical to the E~~ characterizing a charged interface the difference between the pK," and pKH value is solely due to the medium effect on the activity coefficient of the proton, i.e. pK: = pKF - log my"+. The medium effect for the proton, cannot be measured. Hence we have followed the procedure of Fernandez and Fromherz2 and have used the values of the medium effect for HCl in 1,4-dioxane-water mixtures, ,y+, to approximate the ,yH+ values. In the present work, the log my* values were obtained from the change in the standard potential, E", for the cell Pt I H,(g), HCl in 1,4-dioxane-water mixture, AgCl I Ag usingC.J. Drummond, F. Grieser and T. W. Healy 101 Table 1. The A,,,, E , ~ , pK:, pK",' and (clo values obtained with b(30) in the aqueous self- assembled surfactant solutions investigated at 25 "C. Also included are the (clo values calculated using 4-heptadecyl-7-hydroxycoumarin, (clo( HHC) (see text for explanation) concentration A,,, $0 +o(HHC) surfactant /mmol dm-3 /nm Eeff PK', p K zbs /mV /mV C12E8 C12E8 CTAC CTAB DTAC DTAC" DTAB DTAB DTAB DTAB DTAB' DEAB DBAB CPBf SDSg SDSg SDSg SDSg DBSg Oleateg DHPg DMPC DPPC 10 204 50 50 50 50 50 65 162' 324d 50 50 50 50 50 69 173' 347d 50 50 2.5 5 5 540 542 532 534 524 537 528 528 530 532 542 530 532 533 493 493 493 493 490 505 503 545 517 30 29 34 33 39 31 36 36 35 34 29 35 34 33 56 56 56 56 58 49 51 27 43 9.61 9.61 9.60 9.60 9.57 9.61 9.59 9.59 9.59 9:60 9.61 9.59 9.60 9.60 9.32 9.32 9.32 9.32 9.27 9.45 9.41 9.61 9.53 9.06 f 0.04 9.31 f 0.21 6.93 f 0.06 7.22 f 0.01 7.39 f 0.03 8.82 f 0.03 7.59 f 0.04 7.67 f 0.02 7.93 f 0.02 8.07 f 0.04 9.30 f 0.06 7.79 f 0.03 7.78 f 0.03 7.08 f 0.06 10.72 f 0.03 10.70 f 0.03 10.68 f 0.06 10.48 f 0.06 10.29 f 0.03 10.91 f 0.05 10.96 f 0.08 10.55 f 0.02 10.48 f 0.10 0 0 +158 + 141 +129 +47 +118 +114 +98 +91 +18 +lo6 + 108 + 149 -83h -82h -80h -69h -60h -86h -92h -56h -56h 0 +154 +139 +127 +37 +116 +114 + 102 +85 +19 - - - - - - - - - - - -115h - 173h 4 mol dm-3 NaCl.2 wt '/o surfactant. ' 5 wt '/o surfactant. Experiment performed at 30 "C. 10 wt '/o surfactant. 4 mol dm-3 Spectra as a function of pH differed from that obtained (clo value is not the electrostatic NaBr.in pure water and 1,4-dioxane-water mixtures (see text for details). surface potential (see text for details). the r e l a t i ~ n s h i p ~ ~ where is the standard potential for the cell when pure water is the solvent and is equal to 0.222 34 V at 25 0C54 and ' E o is the standard potential of the cell in a 1,4-dioxane- water mixture. The ' E 0 values were taken from the compilation of Feakins and French,55 with the exception of the value in the 82 wt% percent 1,4-dioxane-water mixture which was taken from the work of Danyluk et aL56 Fig. 5 illustrates the change in the pKL value for ET(30) relative to its pK, value in pure water (i.e. ApK!J as a function of the dielectric constant of the 1,4-dioxane-water mixtures.Presented in table 1 are the A,,, and pKib" values determined for ET(30) in the self-assembled surfactant solutions investigated. The corresponding E , ~ values based on the 1,4-dioxane-water curve of fig. 4 are also given. E , ~ values based on other reference systems can also be easily determined with the aid of fig. 3 and 4. These E , ~ values are in reasonable agreement with those that have been determined by Zachariasse et al.,' with the exception of the DMPC and DPPC results. For both DMPC at 25 "C and DPPC at 50 "C they obtained an eeff value, based on the 174-dioxane-water reference system, of 14. We were unable to ascertain the reason for this discrepancy between the102 Probe for the Surface Region of Model Membranes two studies.The pKa values contained in table 1 were obtained from fig. 5 by using the eeff values for the particular self-assembled surfactant systems. These p K a values and the PK;~” values were then substituted into eqn (1) to gain the qb0 values. Not included in table 1 are the results for E,(30) in the unilamellar vesicle dispersions of DDDAB, DHDAB and DODAB, since at the surfactant concentrations employed in this study, (2.5 x lop3 and 5.0 x mol dm-3), both acidic and basic forms of ET(30) were found to partition only slightly, if at all, into the interfacial region of the vesicles, i.e. the same A, and pKZb” results were obtained for ET(30) in these vesicular solutions as were obtained in pure water. In micellar Brij-35 solutions it was not possible to reproduce the spectrum of ET(30) as a function of pH, as the absorption maximum of the phenoxide form of ET(30) disappeared rapidly with time.Once the maximum had vanished it was impossible to regain with the addition of NaOH. Similar behaviour was also observed when unpurified 1,4-dioxane was used as a solvent component and in micellar CI2E8 solution, but to a lesser extent and over a much longer time span in this case. We believe this behaviour is due to a reaction occurring between the phenoxide form of ET(30) and possibly peroxides, or other oxidants present in the sample of Brij-35. Similar behaviour for ET( 30) in unpurified cyclohexanol, 2,4-dimethylpentan- 1 -one, 1 -phenylethanol and 3- phenylpropan-1-01 has also been observed by Aslam et aL5’ Discussion The fundamental assumptions inherent in the type of treatment proposed in this study to determine +b0 values from the pKzd” values for ET(30) in aqueous solutions of charged self-assembled surfactant aggregates are that: (i) the ionizable phenolic hydroxy group of ET(30) resides, on average, in the plane of the charged headgroups of the self- assembled surfactant aggregates; (ii) the position of the solvatochromic band maximum for ET(30) in the self-assembled surfactant aggregates gives a true indication of the eeff of the interfacial region; (iii) the interfacial solvent properties characterized by the eeff value have the same influence on the pKa value for ET(30) as does a 1,4-dioxane-water mixture of equivalent dielectric constant; (iv) the differences between the pK, in pure water and the pKZbS values are solely due to the surface potential and the lower E , ~ at the charged interface; i.e.both acidic and basic forms of ET(30) have fully partitioned into the interfacial region and specific molecular interactions and/or interfacial salt effects do not influence the PK:~” values. The forthcoming discussion will primarily consist of an assessment of the validity of each of these assumptions when dealing with the different kinds of self-assembled surfactant aggregates. Cationic Micelles Two very different types of n.m.r. have established that the average location for the phenolic oxygen atom of ET(30) in cationic micelles is in the plane of the positively charged surfactant headgroups. Thus assumption (i) appears to be justified for cationic micelles.A large electrochromic component60961 to the A,,, values in self-assembled surfactant solutions would invalidate assumption (ii). However, the A, results contained in table 1, especially those for the CI2E8 micelles and DTAB and DTAC micelles with up to 4moldm-3 electrolyte, suggest that the shifts in A,,, are not the result of a large electrochromic response. Since there is a large effective counter-ion concentration present within the interfacial region of m i ~ e l l e s ~ ~ . ~ ~ assumption (ii) would be unjustified if there was a major com- ponent in the ET(30) A,,, values which was due to electrolyte interaction. We tried toC. J. Drummond, F. Grieser and T. W. Healy 103 gauge if there was any possibility of a large electrolyte induced A,,, shift by attempting to determine the A,,, of ET(30) in a number of aqueous electrolyte solutions.We found, however, that the phenoxide form of ET(30) precipitated on the addition of electrolyte, 10 mmol dm-3 NaCl and TMAC, and consequently we were unable to obtain any Amax values in this type of media. Nevertheless, the results for l-methyl-4-[(oxo- cyclohexadienylidene)ethylidene]- 1,4-dih~dropyridine,~~ a more water soluble molecule that possesses solvatochromic behaviour similar to that of ET(30), suggest that it is unlikely that there would be a large electrolyte effect on the A,,, values for ET(30). Furthermore, the E , ~ values determined for the interfacial microenvironments of the cationic micelles are also similar to estimates that have been obtained by employing other solvatochromic spectroscopic One way to test assumptions (iii) and (iv) would be to locate ET(30) in an interfacial microenvironment where the surface potential is zero or close to zero.In principle, this can be achieved by using either non-ionic micelles or by using a high electrolyte concentration to ‘screen’ the surface charge density of a charged interface. Unfortunately, owing to the problems associated with the use of ET(30) in non-ionic micelles comprised of surfactants with poly( ethylene oxide) headgroups, which have been mentioned in the Results section, it was not possible to obtain an accurate pKzbs value for ET(30) in CIZEs micelles. The pK:bs results for C&, given in table 1, also indicate that it is necessary to have a very high concentqation of C,,E, micelles in order to ensure that most of the protonated form of ET(30) has partitioned into the micellar phase.In addition, from the results for DTAC micelles in the presence of 4 mol dm-3 NaCl and DTAB micelles with 4 mol dmP3 NaBr it is evident that the surface charge density of these micelles is still not fully ‘screened’ even at this high electrolyte concentration. A range of added electrolyte concentrations, between 0 and 4 mol dmP3, were investigated for DTAC and DTAB micelles. There is a monotonic decrease in the t,bo as electrolyte is added for both DTAC and DTAB m i ~ e l l e s . ~ ~ A 6 mol dm-3 NaBr solution of DTAB micelles was also investigated but was found to be extremely viscous, and it was not possible to perform reliable pH titrations with ET(30) in this medium.Although pKgb” values for ET(30) in C& micelles and DTAB/4 mol dmP3 NaBr micelles do not prove that the magnitude of the pKa values are quantitatively correct they certainly suggest that the pKa values are at least within 0.3 pK, units of being correct. with 4-heptadecyl-7-hydroxycoumarin (HHC) t,bo values for man of the self-assembled aggregates of table 1 were calculated by assuming that the pK$ value for HHC in C&8 micelles could serve as the pKg value in eqn (1). Since it is now clearly evident that ceff is not the same for each of the different cationic micelles, we have used the ceff values obtained with ET(30) and the curve of pKg as a function of the dielectric constant of 1,4-dioxane-water mixtures for the 7-hydroxycoumarin chromophore, which was determined by Fernandez and Fromherz,* to correct our earlier estimates of the t,bo values.These corrected values are given as the t,bo (HHC) values in table 1. For cationic micelles the agreement between the t,bo values calculated with ET(30), and the t,bo values calculated with HHC is extremely good. Both the un-ionized and ionized forms of HHC fully partition into the micellar phase and both forms of the 7-hydroxycoumarin chromophore reside within the interfacial region of cationic m i ~ e l l e s . ~ ~ It has been shown27 that there is probably little, if any, influence from specific molecular interaction and interfacial salt effects on the acid-base equilibria of HHC at charged interfaces.For 1,4-dioxane-water mixtures, the pKg behaviour of the HHC with changing dielectric constant is not the same as the ET(30) pKL behaviour with changing dielectric constant. Therefore, the close agreement between the t,bo value calculated with ET(30) and the t,bo value calculated with HHC for a particular cationic micelle is a clear indication that assumptions (ii), (iii) and (iv) must be valid in the case of cationic micelles. In an earlier104 Probe for the Surface Region of Model Membranes Anionic Micelles The n.m.r. study of Plieninger and B a ~ m g a r t e l ’ ~ ~ ~ ~ has shown that in the case of SDS micelles the N+ centre of ET(30) is aligned on average in the plane of the sulphate headgroups and the ionizable hydroxy group is positioned some distance out from the plane of the anionic headgroups. This should also be the case for E,(30) in the anionic DBS and oleate micelles. Consequently the acid-base dissociation of ET( 30) in these anionic micelles will not be influenced by the electrostatic surface potential but by the electrostatic potential at the average position of ‘sit’ for the hydroxy group.As can be seen by comparing fig. 1 and 2, the behaviour of the visible absorption spectrum of ET(30) in anionic micelles as a function of pH is unlike that seen in pure water, 1,4-dioxane-water mixtures or cationic micelles. This is probably a result of specific molecular interaction between the N’ centre of the ET(30) molecule and an anionic headgroup which affects the optical properties of the ET(30) molecule.Because of this difference in the optical properties of ET(30) in anionic micelles we are uncertain whether or not the E,(30) A,,, values in these micelles can be compared with the A,,, results of fig. 3 and 4 to obtain eeff values for the interfacial microenvironments. Table 1 contains E , ~ values based on the assumption that this type of comparison can be made. Note that the E , ~ values are consistent with the known location of the probe, i.e. higher ceff values would be expected for ET(30) in anionic micelles than for ET(30) in cationic micelles because ET(30) is on average located further out from the interface in the former case. Interestingly, Plienir~ger~~ observed that the addition of a long alkoxy chain to the ET(30) molecule, at the opposite end of the molecule to the phenolic oxygen atom, altered the A,,, values obtained with this probe in SDS micelles in a manner that was consistent with the long alkoxy chain ‘dragging’ the chromophore in closer to the micelle interior.For CTAB micelles P l i e ~ ~ i n g e r ~ ~ found very little difference between the results obtained with ET(30) and the alkoxy ET(30). The magnitude of the t,bo values, which were calculated by assuming that the eeff values given in table 1 are correct, are also consistent with the ET( 30) molecule ‘sensing’ an electrostatic potential out from the plane of the headgroups in the case of anionic micelles. It should be emphasized, however, that the different optical properties found for ET(30) in anionic micelles introduces a large element of uncertainty into the analysis of.these systems. Vesicles No studies on the average location of ET(30) in vesicles have been reported. As discussed in the Results section, at the concentration of surfactant employed in this investigation, ET(30) does not appear to partition into the cationic DDDAB, DHDAB and DODAB vesicles to any great extent. The spectrum of ET(30) as a function of pH in the anionic DHP vesicles is similar to that seen for ET(30) in the anionic micelles. Therefore, it is highly likely that the N+ centre of ET(30) is aligned on average in the plane of the phosphate headgroups with the ionizable hydroxy group situated on average at a position out from the plane of the headgroups. Because of the optical properties of ET(30) in DHP vesicles being different to those of ET(30) in 174-dioxane-water mixtures, the same uncertainty is involved in the calculated eeff and t,bo values, table 1, as has been discussed for the case of the anionic micelles.The optical properties of E,(30) in DMPC and DPPC vesicles suggest’that there is, on average, little or no specific molecular interaction between the Nt centre of ET(30) and the negative phosphate part of the zwitterionic phosphatidylcholine headgroups. Although it is probably situated, on average, somewhere in the vicinity of the glycerol backbone region of the vesicles, the exact location of ET(30) in DMPC and DPPC vesicles is unknown. Consequently, at this stage it is not possible to assess whether the different eeff values found for the DMPC and DPPC vesicles at 25 “C are due to these types of vesicles possessing different interfacial properties above their gel-liquid-crystal-C.J. Drummond, F. Grieser and T. W. Healy 105 line phase transition temperature, T,, to what they do below their T, (DMPC, T, = 24 "C and DPPC, T, = 41 "C) or to the ET(30) molecule being located at a different position in the two vesicles. The $o results for the DPPC and DMPC vesicles indicate that the ET( 30) molecule 'senses' a local negative electrostatic potential, which is consistent with ET(30) residing on average in the glycerol backbone region of these vesicles. Work with HHC has indicated2' that a local negative electrostatic potential exists in this region. The differences between the & (HHC) values and the E,(30) values (table 1) can be attributed to the two chromophores having different average sites of residence in the glycerol backbone region.The finding that local electrostatic potentials exist in the interfacial microenvironments of vesicles with zwitterionic phosphatidylcholine head- groups, and that they influence the acid-base dissociation of molecules, is interesting in view of the fact that electrophoretic mobility measurements66 indicate that the net electrostatic surface potential for these vesicles is zero for the pH range studied in this paper. Conclusions It has been established that ET (30) can be employed to determine both the effective interfacial dielectric constant and the electrostatic surface potential of a cationic micelle. The optical properties of ET(30) in anionic micelles and vesicles have been found to be dissimilar to those of ET( 30) in pure water, 174-dioxane-water mixtures, cationic micelles and vesicles with zwitterionic phosphatidylcholine headgroups.As a consequence of this we are uncertain whether or not the A,,, values obtained for ET(30) in anionic micelles and vesicles can be compared with the reference 174-dioxane-water mixtures to determine tzeff values. However, if it is assumed that the comparison can be made, the ceff and values calculated are consistent with the known average location and orientation of the ET( 30) molecule. At 25 "C the ceff values determined for the DMPC and DPPC vesicles are not the same. However, the average location of ET(30) in these vesicles is unknown.Therefore, it is not clear whether this difference is a result of the ET(30) molecule being situated at a different average site of residence in the two types of vesicles or due to the ceff value at the one site being different. The I,!J~ results for the DMPC and DPPC vesicles clearly indicate the presence of a local negative electrostatic potential within the inter- facial microenvironment of these zwitterionic vesicles. We are currently investigating other kinds of solvatochromic acid-base indicators in the hope that some of the problems that have been associated with the use of ET(30) in anionic micelles and in all types of vesicles may be resolved. We thank Prof. C. Reichardt, University of Marburg, for his generous gift of the E,(30). We thank Dr J Brady and Prof.D. F. Evans, University of Minnesota, and Prof. B. W. Ninham, Australian National University, for supplying some of the surfactants. We also thank Prof. H. Baumgartel, Freie University of Berlin, for a copy of ref. (59). This work was supported through a Program Grant provided by the Australian Research Grants Scheme. C.J.D. is the recipient of a Commonwealth Postgraduate Research Award. References 1 E. H. Cordes and C. Gitler, Progr, Biorg. Chem., 1973, 2, 1 . 2 M. S. Fernandez and P. Fromherz, J. Phys. Chem., 1977, 81, 1755. 3 K. Kalyanasundaram and J. K. Thomas, J. Phys. Chem., 1977, 81, 2176. 4 P. Mukerjee, J. R. Cardinal and N. R. Desai, in Micellization, Solubilization and Microemulsions, ed. 5 K. Y. Law, Photochem. Photobiol., 1981, 33, 799.6 S. Lukac, J. Am. Chem. Soc., 1984, 106, 4386. 7 K. A. Zachariasse, N. Van Phuc and B. Kozankiewicz, J. Phys. Chem., 1981, 85, 2676. 8 K. Kano, H. Goto and T. Ogawa, Chem. Lett., 1981, 653. K. L. Mittal (Plenum Press, New York, 1977), vol. 1, p. 241.106 Probe for the Surface Region of Model Membranes 9 C. Ramachandran, R. A. Pyter and P. Mukerjee, J. Phys. Chem, 1982,86, 3198. 10 T. Handa, C. Ichihashi, I. Yamamoto and M. Nakagaki, Bull. Chem. SOC. Jpn, 1983, 56, 2548. 11 S. Kaneshina, H. Kamaya and I. Ueda, Biochim. Biophys. Acta, 1984,777, 75. 12 C. J. Drummond, G. G. Warr, F. Grieser, B. W. Ninham and D. F. Evans, J. Phys. Chem., 1985,89,2103. 13 F. Dion, F. Bolduc and I. Gruda, J. Colloid Interface Sci., 1985, 106, 465. 14 Y. Kimura and A. Ikegami, J.Membr. Biol., 1985,85, 225. 15 M. Shinitzky, A. C. Dianoux, C. Gitler and G. Weber, Biochemistry, 1971, 10, 2106. 16 M. Gratzel and J. K. Thomas, J. Am. Chem. SOC., 1973,95, 6885. 17 K. Zachariasse, Chem. Phys. Lett., 1978,57, 429. 18 K. A. Zachariasse, W. Kuhnle and A. Weller, Chem. Phys. Lett., 1980, 73, 6. 19 N. J. Turro, M. Aikawa and A. Yekta, J. Am. Chem. Soc., 1979, 101, 772. 20 M. Almgren, F. Grieser and J. K. Thomas, J. Am. Chem. Soc., 1980, 102, 3188. 21 M. Van der Auweraer, C. Dederen, C. Palmans-Windels and F. C. De Schryver, J. Am. Chem. SOC, 22 F. Grieser, M. Lay and P. J. Thistlethwaite, J. Phys. Chem., 1985, 89, 2065. 23 B. R. Suddaby, P. E. Brown, J. C. Russell and D. G. Whitten, J. Am. Chem. SOC., 1985, 107, 5609. 24 K. R. Thulborn, L. M. Tilley, W.H. Sawyer and F. E. Treloar, Biochim. Biophys. Acta, 1979, 558, 166. 25 E. Blatt, K. P. Ghiggino and W. H. Sawyer, J. Phys. Chem., 1982, 86, 4461. 26 S. Lukac, J. Phys. Chem., 1983, 87, 5045. 27 C. J. Drummond and F. Grieser, to be published. 28 B. Lovelock, F. Grieser and T. W. Healy, J. Phys. Chem., 1985, 89, 501. 29 P. Mukerjee and K. Banerjee, J. Phys. Chem., 1964, 68, 3567. 30 M. Montal and C. Gitler, J. Bioenerg., 1973, 4, 363. 31 J. V. Moller and U. Kragh-Hansen, Biochemistry, 1975, 14, 2317. 32 N. Funasaki, Nippon Kagaku Kaishi, 1976, 5, 722. 33 T. Mashimo, I. Ueda, D. D. Shieh, H. Kamaya and H. Eyring, Proc. Nut1 Acad. Sci. USA, 1979,76,5114. 34 S. McLaughlin, Curr. Top. Membr. Transp., 1977, 9, 71. 35 J. D. Castle and W. L. Hubbell, Biochemistry, 1976, 15, 4818.36 M. Nakagaki, I. Katoh and T. Handa, Biochemistry, 1981, 20, 2208. 37 B. Ehrenberg and Y. Berezin, Biophys. J., 1984, 45, 663. 38 G. S. Hartley and J. W. Roe, Trans. Faraday SOC., 1940, 36, 101. 39 J. T. Davies, Adv. Catal., 1954, 6, 56. 40 C. M. Harris and B. K. Selinger, Z. Phys. Chem. N.F., 1983, 134, 65. 41 J. Garcia-Soto and M. S. Fernandez, Biochim. Biophys. Acta, 1983, 731, 275. 42 C. Reichardt, Solvent Efects in Organic Chemistry (Verlag Chemie, Wenheim, 1979), chap. 7. 43 Y-M. Tricot, D. N. Furlong, W. H. F. Sasse, P. Daivis, I. Snook and W. Van Megen, J. Colloid Interface 44 L. G. Van Uitert and C. G. Haas, J. Am. Chem. SOC., 1953, 75, 541. 45 C. J. Drummond, F. Grieser and T. W. Healy, in preparation. 46 K. Dimroth and C. Reichardt, Z. Anal. Chem., 1966, 215, 344. 47 E. M. Kosower, H. Dodiuk, K. Tanizawa, M. Ottolenghi and N. Orbach, 1. Am. Chem. SOC., 1975,97, 48 G. Akerlof, J. Am. Chem. SOC., 1932, 54, 4125. 49 F. W. Critchfield, J. A. Gibson and J. L. Hall, J. Am. Chem. Soc., 1953, 75, 1991. 50. C. Reichardt and E. Harbusch-Gornert, Liebigs Ann. Chem., 1983, 721. 51 CRC Handbook of Chemistry and Physics, ed. R. C. Weast (CRC Press, Boca Raton, Florida, 60th edn, 1981). 52 International Critical Tables of Numerical Data, Physics, Chemistry and Technology, ed. E. W. Washburn (McGraw-Hill Book Company, Inc., New York, 1928). 53 R. G. Bates, Determination ofpH: Theory and Practice, (John Wiley, New York, 2nd edn, 1973), p. 270. 54 R. G. Bates, Determinution ofpH: Theory and Practice, (John Wiley, New York, 2nd edn, 1973), p. 334. 55 D. Feakins and C. M. French, J. Chem. SOC., 1957, 2581. 56 S. S. Danyluk, H. Taniguchi and G. J. Janz, J. Am. Chem. SOC., 1957, 61, 1679. 57 M. H. Aslam, G. Collier and J. Shorter, J. Chem. SOC., Perkin Trans. 2, 1981, 1572. 58 P. Plieninger and H. Baumgartel, Liebigs Ann. Chem., 1983, 860. 59 P. Plieninger, Ph.D. Dissertation (Freie University of Berlin, 1981). 60 J. R. Platt, J. Chem. Phys., 1961, 34, 862. 61 W. Liptay, Ber. Bunsenges. Phys. Chem., 1976, 80, 207. 62 P. Mukerjee, J. Phys. Chem., 1962, 66, 943. 63 F. M. Menger, H. Yoshinga, K. S. Venkatasubban and A. R. Das, J. Org. Chem., 1981,46, 415. 64 S. J. Davidson and W. P. Jencks, J. Am. Chem. SOC., 1969, 91, 225. 65 A. Haase, Ph.D. Dissertation (University of Giessen, 1980). 66 D. Papahadjopoulus, Biochim. Biophys. Acta, 1968, 163, 240. 1982, 104, 1800. Sci., 1984, 97, 380. 2167. Received 12th December, 1985