Negative Dipole Potentials of Uncharged Langmuir Monolayers Due

For a more comprehensive list of citations to this article, users are ... Peter Karageorgiev , Jordan G. Petrov , Hubert Motschmann , and Helmuth Moeh...
1 downloads 0 Views 407KB Size
Download PDF
14102

J. Phys. Chem. B 2005, 109, 14102-14111

Negative Dipole Potentials of Uncharged Langmuir Monolayers Due to Fluorination of the Hydrophilic Heads Jordan G. Petrov,* Tonya D. Andreeva, Dirk G. Kurth, and Helmuth Mo1 hwald Max-Planck Institute of Colloids and Interfaces, Am Mu¨hlenberg 1, D-14476 Golm/Potsdam, Germany ReceiVed: March 23, 2005

The dipole potential, affecting the structure, functions, and interactions of biomembranes, lipid bilayers, and Langmuir monolayers, is positive toward the hydrocarbon moieties. We show that uncharged Langmuir monolayers of docosyl trifluoroethyl ether (DFEE) exhibit large negative dipole potentials, while the nonfluorinated docosyl ethyl ether (DEE) forms films with positive dipole potentials. Comparison of the ∆V values for these ethers with those of the previously studied37-39 monolayers of trifluoroethyl ester (TFEB) and ethyl ester of behenic acid (EB) shows that the reversal of the sign of ∆V causes the same change ∆(∆V) ) -706 ( 16 mV due to fluorination of heads. The ∆V values of both TFEB and EB films differ by -122 ( 16 mV from those of DFEE and DEE monolayers, respectively, with the same density. Such quantitative coincidence points to a common mechanism of reversal of the sign of the dipole potential for the ether and ester films despite the different structure of their heads. The mechanical properties and phase behaviors of these monolayers show that both fluorinated heads are less hydrated, suggesting that the change of the sign of ∆V could, at least partially, be related to different hydration water structure. The same negative contribution of the carbonyl bond in both TFEB and EB films contrasts with the generally accepted positive contribution of the Cδ+dOδ- bond in condensed Langmuir monolayers of fatty acids, their alcohol esters, glycerides, and phospholipids but concurs with the theoretical analysis of ∆V of stearic acid monolayers.42 Both results question the literature values of the molecular dipole moments of these substances calculated via summation of bonds and atomic group contributions. Mixed monolayers of DFEE and DEE show smooth monotonic variation of ∆V from +450 to -235 mV, indicating a way for adjustment of the sign and magnitude of the dipole potential at the membrane-water boundary and regulation of such membrane behaviors as binding and translocation rate of hydrophobic ions and ion-carriers, adsorption and penetration of amphiphilic peptides, polarization of hydration water, and short-range repulsion. The interaction of the hydrophobic ions tetraphenylboron TPhBand tetraphenylphosphonium TPhP+ with DFEE and DEE monolayers qualitatively follows the theory of binding of such ions to lipid bilayers,4 but the shifts ∆(∆V) from the values obtained on water are much smaller than those for DPPC monolayers. This difference seems to be due to the solid (polycrystalline) character of the DFEE and DEE films that hampers the penetration of TPhB- and TPhP+ in the monolayers and reduces the attractive interaction with the hydrophobic moiety. This conclusion orients the future synthesis of amphiphiles with fluorinated heads to those which could form liquid-expanded Langmuir monolayers.

Introduction The Langmuir monolayers of amphiphilic substances at the air-water interface are simple models of biomembranes. The dipole potential ∆V resulting from the hydrated polar groups at the membrane-water boundary is an important membrane characteristic that can be studied on such films [in the monolayer literature ∆V is called surface potential, but here we use the term dipole potential proposed by Brockman in order to stress that it arises from monolayer dipoles and water polarization].1-3 This potential determines the anion-cation specificity of the translocation kinetics and the binding of hydrophobic ions,4 modifies the interaction between membranes and amphiphilic peptides,5-9 affects the activity of phospholipase A2 in monolayers,10 and mediates the action of general anesthetics11,12 and the fusion of viral and cellular membranes.13 There is a number of reports that the dipoles of the headgroups control the polarization of the hydration water and the short-range repulsion * Corresponding author. Present address: Institute of Biophysics of the Bulgarian Academy of Sciences, 1 Acad. G. Bonchev Str., Block 21, 1113 Sofia, Bulgaria. E-mail: [email protected].

between bilayers or other hydrophilic surfaces.14-20 Hydration of the heads and dipole-dipole interactions are important determinants of the lateral structure of the membranes.21 This relationship is well documented for Langmuir monolayers, where the dipole-dipole repulsion and line tension regulate the size and the shape of the phase domains.22,23 The significance of the dipole potential for the membrane functions structure and interactions dictates the necessity of its regulation. This has been done adsorbing molecules with large dipole moments at the cell membrane surface or incorporating such molecules into lipid bilayers.24,25 Substantial reduction of the dipole potential was achieved with phloretin, while 6-ketocholestanol strongly increased its magnitude. Dramatic changes of the membrane permeability resulted from such treatment. Adsorption of phloretin on red blood cells decreasing their positive dipole potential24 caused a 105-fold increase of the transmembrane conductance of the positively charged ion-carrier K+-nonactin, and decreased the Cl- conductance by a factor of 102-103. The permeability of human erythrocyte membranes by urea and glucose was also drastically changed by phloretin and its analogues.26,27

10.1021/jp0515028 CCC: $30.25 © 2005 American Chemical Society Published on Web 07/01/2005

Negative Dipole Potentials Flewelling and Hubbell4 proposed remote modulation of the dipole potential of bilayer-water boundary via synthetic incorporation of “small polar groups with large dipole momentss e.g. CF3 groups or nitroxides ... along the fatty acid chains”. This was done by Voglino et al.,28,29 who introduced Brδ-Cδ+ or Nδ+-Oδ- dipoles at different positions of the acyl chains of phosphatidylcholines. Monolayers with Nδ+-Oδ- dipoles in 7th position showed ∆V ) +370 mV, which corresponds to 11% reduction of ∆V ) +415 mV of the egg PC or 36% decrease of ∆V ) +575 mV for DPPC films. As far as we are aware, no reversal of the sign of the positive dipole potential of lipid bilayers and cell membranes has been achieved with phloretin and other strongly polar additives. It is known, however, that fluorination of the terminal CH3 group, or part of the hydrocarbon chain, makes the dipole potential of Langmuir monolayers at air-water and oil-water interfaces negative.30,31,32 McIntosh et al.33 exploited this experience and studied phospholipid bilayers with partially fluorinated alkyl chains, motivated by their advantages as in vivo drug carriers with extended blood circulation time.34 It was found that the binding of hydrophobic ions qualitatively follows the theory of Flewelling and Hubbell;4 adsorption of the hydrophobic anion tetraphenylboron (TPB-) at monolayers of a fluoroalkylated PC (∆V ) -485 mV) caused much smaller ∆V shifts than those observed for egg PC films with ∆V ) +415 mV. However, the osmotic pressure-distance data for fluoroalkylated PC and DPPC bilayers in the subgel phase, where repulsion is mainly due to hydration forces, did not differ significantly, and the magnitude of the hydration pressure in the fluoroalkylated systems was not proportional to (∆V)2 as previously found for nonfluorinated phospholipids.15 The negative dipole potential of the above fluoroalkylated monolayers originates from the terminal F3C- dipole and the -CF2-CH2- dipole in the partially fluorinated chain. Such remote dipoles affect the monolayer-water boundary if located at distances less than 15 Å from it.35,36 To achieve maximum modulation of the electrostatic characteristics of this region we introduced a -CF3 dipole in the hydrophilic head of a long chain ethyl ester. The first representative of the new class of amphiphiles with fluorinated heads, the trifluoroethyl behenate (TFEB) RCOOCH2CF3 formed condensed monolayers with large negative dipole potential of -355 mV, dramatically contrasting to the positive dipole potential of +345 mV of the Langmuir films of the nonfluorinated ethyl behenate (EB) RCOOCH2CH3.37-39 Thus, fluorination of the CH3 terminus of the ethyl ester head reversed the sign of the dipole potential and changed its magnitude by 203%. Here we report the dipole potentials of pure and mixed Langmuir monolayers of the second representative of the new class of amphiphiles, the docosyl trifluoroethyl ether (DFEE) ROCH2CF3sFigure 1, left. Its main advantage is the possibly simplest fluorinated headgroup, whose allowed conformations and hydration water structure are much less complicated as by the trifluoroethyl ester TFEB. The goal of this synthesis was that the simplified DFEE monolayer-water boundary would help to better understand the structural origin of the dipole potential, which is still poorly understood.2,3,35 The dipole potential of DFEE monolayers was compared with ∆V of the nonfluorinated analogue, the docosyl ethyl ether (DEE) ROCH2CH3 to quantify the effect of CF3 substitution in the headssFigure 1, right. On the other side, we compared the ∆V values of the DFEE and DEE monolayers, which do not possess carbonyl bonds, with those of the TFEB-EB couple as done previously for phospholipids with ether and ester

J. Phys. Chem. B, Vol. 109, No. 29, 2005 14103

Figure 1. Molecular models of docosyl trifluoroethyl ether (left) and docosyl ethyl ether (right) obtained via MOPAC.

linkages of the hydrocarbon chains.45,54 Data from our previous investigations of the molecular structure of these monolayers,39,40 obtained via in situ X-ray diffraction at grazing incidence (GIXD), were used to more adequately interpret the surface potentials. To the best of our knowledge DFEE and TFEB constitute the only amphiphiles giving Langmuir films with negative dipole potentials arising from their hydrophilic heads. This peculiarity of the nature was realized years ago by Haydon and Hladky.41 In a critical review on the ion transport across thin lipid films they wrote: “It is interesting that many, if not all, of the dipole potentials reported for liquid interfaces are positive; the positive end of the dipole is oriented towards the non-polar phase. This in not surprising in view of the structure of the surface active lipids, indeed, it is not easy to think of a lipid molecule, which might have a dipole of the opposite sign.” In this context the monolayers of our DFEE and TFEB amphiphiles exhibit unusual electrostatic properties. Moreover, the results found for these substances are not universal. Our recent (still unpublished) data for the third synthesized amphiphile with fluorinated head, RCONHCH2CF3, showed that this new class of substances is much more complicated but also challenging; here the headgroup fluorination does not reverse the ∆V values but increases the positive dipole potentials of RCONHCH2CH3 films by more than 300%. This paper presents some central results on the dipole potentials of DFEE and DEE monolayers. Most of these data will be extended in future detailed investigations whose final goal is to understand the origin of the dipole potential and the mechanism of the effects caused by fluorination of the hydrophilic heads. Experimental Section Synthesis, Purification, and Characterization of the Amphiphiles. Docosyl Trifluoroethyl Ether: C24H47F3O, MW ) 408.62. Trifluoroethanol (5 g, 0.05 mol) was slowly added to sodium (0.2 g, 0.008 mol) under inert gas. After dissolution of sodium, docosyl bromide (1.9 g, 0.005 mol) in dioxane and a grain of dry KJ were added. The mixture was stirred under reflux

14104 J. Phys. Chem. B, Vol. 109, No. 29, 2005 until thin-layer chromatography indicated complete conversion of docosyl bromide. After cooling to room temperature, the mixture was filtered and the solvent was removed under vacuum. The solid was dissolved in ether. The organic phase was washed with water, dried over magnesium sulfate, and filtered, and the solvent was removed under vacuum. The crude solid was purified by flash chromatography on silica gel (Aldrich, Germany) with hexane:diethyl ether 1:1. Isolated yield: 1 g (48%). Melting point: 47-49 °C. Anal. Calcd for C24H47OF3: C, 70.54; H, 11.59; F, 13.95. Found: C, 70.59; H, 11.49; F, 13.92. Purity by GC and HPLC: 99%. 1H NMR (500 MHz, CDCl , 25 °C, TMS): δ ) 0.88 (triplet, 3 J(H,H) ) 7 Hz, 3H; CH3), 1.25 (multiplet, 38H; CH2), 1.60 (quintet, J(H,H) ) 6.8 Hz, 2H; OCH2CH2), 3.59 (triplet, J(H,H) ) 6.7 Hz, 2H; OCH2CH2), 3.79 (quartet, 3J(H,F) ) 8.8 Hz, 2H; CH2CF3). 13C NMR (125 MHz, CDCl , 25 °C, TMS): δ ) 14.1 (CH ), 3 3 22.7 (CH3CH2), 25.8 (CH2CH2O), 29.3-29.7 (CH2), 32.0 (CH3CH2CH2), 68.3 (quartet, 2J(C,F) ) 34 Hz; CH2CF3), 73.0 (CH2CH2O), 124.1 (quartet, 1J(C,F) ) 279 Hz; CF3). 19F NMR (280 MHz, CDCl , 25 °C, TMS): δ ) -74.7 3 (triplet, 3J(F,H) ) 9.2 Hz). MS (FAB): m/z [M - H]+ 407. IR (KBr): ν[cm-1] ) 2917 (νas CH2), 2848 (νs CH2), 1474 (δsc CH2), 1460 (δas CH3), 1276, 1184, 968, 730, 720, 671. Docosyl Ethyl Ether: C24H50O, MW ) 354.65. This compound was prepared in analogy to docosyltrifluoroethyl ether. The crude solid was purified by flash chromatography followed by crystallization from ethanol. Isolated yield: 1.2 g (37%). Melting point: 44 °C. Anal. Calcd for C24H50O: C, 81.28; H, 14.21. Found: C, 81.44; H, 14.63. Purity by HPLC: 99%. 1H NMR (300 MHz, CDCl , 25 °C, TMS): δ ) 0.88 (triplet, 3 J(H,H) ) 7 Hz, 3H; CH3), 1.20 (triplet, J(H,H) ) 7 Hz, 3H; O-CH2CH3), 1.25 (multiplet, 38H; CH2), 1.56 (multiplet, 2H; CH2CH2O), 3.99 (triplet, J(H,H) ) 6.7 Hz, 2H; CH2CH2O), 3.47 (quartet, J(H,H) ) 7.1 Hz, 2H; O-CH2CH3). 13C NMR (125 MHz, CDCl , 25 °C, TMS): δ ) 14.1 (CH ), 3 3 15.3 (OCH2CH3), 22.7 (CH3CH2), 26.2 (CH2CH2O), 29.7 (CH2), 31.9 (CH3CH2CH2), 66.0 (OCH2CH3), 70.8 (CH2O). MS (FAB): m/z [M]+ 355. IR (KBr): ν[cm-1] ) 2917 (νas CH2), 2851(νs CH2), 1471 (δsc CH2), 1460 (δas CH3), 1378 (δs CH3), 1142, 1121 (νas C-O-C), 733, 718. Experimental Methods. The dipole potential of the monolayers was measured via the vibrating capacitor method with accuracy 5 mV. The measurements were performed on a Langmuir film balance with a Teflon trough and Wilhelmy dynamometric system registering the surface pressure with accuracy 0.2 mN/m, and the molecular area within 0.2 Å2. Brewster angle microscopy (BAM-2, NFT, Goettingen) was used to visualize and compare the variation of the monolayer morphology during compression. The time and the velocity of compression enable addressing each image to particular molecular area. Aliquots of 1 mM chloroform solutions of DFEE and DEE were spread on Millipore-Q water or 10-4 M solutions of hydrophobic ions. After 10 min, allowed for evaporation of the solvent, the monolayers were compressed at velocity of 2.2 Å2/(molecule.min). All measurements were performed at 20°(0.2 °C maintained by a temperature control system. Results Surface Pressure)Molecular Area and Dipole Potential) Molecular Area Isotherms. Figure 2a presents ∆V/A and π/A

Petrov et al.

Figure 2. Isotherms surface pressure-average molecular area π/A, and dipole potential-average molecular area ∆V/A of DFEE (a) and DEE monolayers (b) on water.

dependencies for docosyl trifluoroethyl ether showing that the DFEE monolayer exhibits a negative dipole potential. Steep variation of ∆V starts below ∼24 Å2/molecule reaching a plateau ∆Vplt ) -250 mV at the collapse of the film. The π/A isotherm is steep, without plateaus and kinks, a trend that is typical for solid monolayers with closely packed upright hydrocarbon chains. This molecular arrangement was proven in our previous GIXD study of the same system,40 which showed that even the solid 2D islands existing in the initial π/A plateau are organized in S phase with closely packed upright molecules. Therefore, up to the collapse surface pressure the DFEE monolayer undergoes only the first-order G-S phase transition. The ∆V shoulder at ∼22 Å2 is irreproducible both in form and average molecular area; it appears when the floating islands are pushed together and further compression causes increasing of surface pressure (see Figure 3). The collapse of the monolayer occurs at πcol ) 32.5 mN/m and Acol ) 18.5 Å2, the inflection point, at which d2π/dA2 reverses sign and the monolayer loses stability, is located at πinf ) 28.5 mN/m and Ainf ) 19.1 Å2. Repeated measurements show a scatter of the πcol and πinf values within 2-3 mN/m; those of the average molecular areas Acol and Ainf coincide within the experimental error of 0.2 Å2. All scatter limits demonstrate that the π/A isotherms are very well reproducible at high surface pressure, where the heterogeneous films become sintered. The value of Ainf, measured in the Langmuir trough, coincides with the GIXD molecular area AXY ) 19.1 Å2 obtained earlier40 showing that at πinf the DFEE film becomes homogeneous and achieves maximum density. Therefore, the value of ∆Vinf ) -235 mV, which is also well reproducible (see Table 1) is characteristic for the compact DFEE monolayer. Figure 2b presents the ∆V/A and π/A isotherms for docosyl ethyl ether. Here the dipole potential is positive; it sharply increases below ∼27 Å2 and reaches a maximum of +470 mV

Negative Dipole Potentials

J. Phys. Chem. B, Vol. 109, No. 29, 2005 14105

Figure 3. Brewster angle microscopy images of DFEE (left column) and DEE (right column) at different stages of compressionscf. Figure 2. DFEE: (a) 0.3 mN/m, 36.8 Å2; (b) 0.3 mN/m, 33.7 Å2; (c) 0.3 mN/m, 27.9 Å2; (d) 6.6 mN/m, 19.6 Å2; (e) 20.5 mN/m, 19.1 Å2. DEE: (a′) 0.1 mN/m, 38.6 Å2; (b′) 0.1 mN/m, 26.9 Å2; (c′) 0.1 mN/m, 24.4 Å2; (d′) 4.3 mN/m, 21.3 Å2; (e′) 34.9 mN/m, 19.5 Å2.

TABLE 1: Dipole Potentials ∆Vm,inf of Compact Monolayers of the Present and Previously Studied37 Amphiphiles with Fluorinated and Nonfluorinated Hydrophilic Heads and the Same Upright Chains at the Inflection Points of Their π/A Isotherms Calculated from Eq 2 and the GIXD Data39,50 for the Actual Molecular Area AXY substance

∆Vinf (mV)

AXY (Å2)

AXY/Ainf/

∆Vm,inf (mV)

DFEE, R-OCH2CF3 DEE, R-OCH2CH3 TFEB, R-COOCH2CF3 EB, R-COOCH2CH3

-236 ( 8 +455 ( 5 -355 ( 10 +345 ( 10

19.1 19.1 18.9 18.8

1.00 0.98 0.97 0.99

-236 ( 8 +464 ( 5 -364 ( 10 +348 ( 10

at the initial increase of the surface pressure. In the low-pressure part of the π/A isotherm ∆V slightly decreases, and after the π/A kink remains constant until the collapse of the monolayer, i.e., ∆Vinf ) ∆Vcol ) +450 mV. A phase transition between the closely packed tilted L2′ phase and S-phase consisting of closely packed upright molecules, occurs in the low-pressure section of the π/A isotherm. Above the kink the upright DEE molecules occupy the same real molecular area AXY ) 19.1 Å2 as found for the DFEE film,40 despite the larger volume of the CF3 groupssee Figure 1. The collapse surface pressure, πcol )

52.3 ( 2.4 mN/m, and the inflection surface pressure, πinf ) 47.3 ( 1.9 mN/m, are significantly higher, demonstrating that the DEE monolayer is more stable against compression than the DFEE film. The average molecular area at the inflection point Ainf ) 19.5 ( 0.2 Å2 slightly exceeds the GIXD area AXY, indicating a presence of voids in the DEE film even at πinf. The values of ∆Vinf in Figure 2 show, that fluorination of the hydrophilic head of DEE reverses the sign of the dipole potential and causes a change of -685 mV or 152% of ∆Vinf of the DEE film. It reduces the collapse surface pressure and monolayer stability and changes the monolayer phase behavior at low surface pressure. Morphology of DFEE and DEE Monolayers. Figure 3 visualizes the variation of the morphology of DFEE and DEE monolayers with decreasing of the average molecular area A. The left column corresponds to the DFEE film and the right one to the DEE monolayer. When spread at A ) 70 Å2, both substances form incompact islands and “archipelago structures” with irregular shape, which are pushed together by the moving barrier (Figure 3, parts a, b, and c and parts a′, b′, and c′). Dark closed areas, containing 2D microcrystals and invisible gaseous phase occupy significant part of the islands and archipelagos

14106 J. Phys. Chem. B, Vol. 109, No. 29, 2005

Petrov et al.

in the initial plateaus of the π/A isotherms (cf. Figure 2). Such “lakes” can be occasionally seen also at higher surface pressure (Figures 3d′) but they disappear at further compression (A = AXY at πinf) and give homogeneous films (Figure 3, parts d, e, and e′). The fact that they are closed under compression shows that the compact DFEE and DEE monolayers consist of 2D microcrystals retaining their autonomy and mobility. The DFEE images are optically isotropic from zero to the collapse surface pressure demonstrating an absence of tilting transitions in the fluorinated monolayer. At low surface pressure the images of the DEE film are anisotropic, but the anisotropy disappears above 10 mN/m. These behaviors have been directly proven in our previous GIXD study,40 which showed that at 0 < π < πcol the compact parts of the DFEE film are organized in the upright solid S phase, while the DEE monolayer undergoes a tilting L2′-S transition that ends at 10 mN/m. Variation under Compression of the Dipole Potential and the Apparent Molecular Dipole Moment in the Compact Parts of the DFEE and DEE Monolayers. The ∆V/A and π/A isotherms in Figure 2 show that the main change of the dipole potential occurs at zero surface pressure. Comparison of the ∆V/A dependencies with the BAM video films suggests that the sharp change of ∆V occurs when the islands and the archipelagos come close together and the motion of the monolayers in the laser beam spot abruptly slows down. The GIXD areas per molecule Axy obtained previously40 enable determination of the partial area occupied by the compact film under such conditions. At the initial increase of the surface pressure, where Ai ) 19.7 Å2 for DFEE and 22.2 Å2 for DEE, both monolayers approach homogeneity; the compact DFEE monolayer occupies a partial area Axy/Ai ) 97.0%, and the compact DEE film shows Axy/Ai ) 93.7%. The dipole potential of the sintered but still heterogeneous monolayers at A < Ai consists of two components, ∆Vm, due to the compact film with partial area Axy/A, and ∆Vv, due to the gaseous phase in the voids occupying partial area (A - Axy)/A:

∆V ) ∆Vm

(

)

AXY A - AXY + ∆VV A A

(1)

The second right-hand side term is negligible, because ∆Vm . ∆Vv and (Axy/A) . (A - Axy)/A. Therefore, the variation of ∆Vm with decreasing of the average molecular area A, measured in the Langmuir trough, reads:

A ∆Vm ) ∆V AXY

(2)

A simple expression exists for compact uncharged monolayers of polarizable molecules, whose dipole moment in vacuum µ is directed at an angle Θ to the subphase normal:42

∆Vm )

µ cos Θ 0Am

(3)

Here 0 is the permittivity of vacuum, and  is the effective relative permittivity of the monolayer accounting for the intraand intermolecular dipole-dipole interactions in the film. The same expression holds if µ is the molecular dipole moment component along the hydrocarbon chain and Θ is the tilt angle of the chain from the subphase normal. Our GIXD data40 for AXY ) Am and Θ enable calculation of the dipole potential ∆Vm and the apparent molecular dipole moment component along the chain µ/ for the compact parts of the monolayers independent of the presence of voids. Figure

Figure 4. π/A and ∆V/A isotherms with extended abscissa scale (closed points). The dipole potential ∆Vm of compact monolayers of DFEE (top) and DEE (bottom) and the apparent molecular dipole moment component along the chain µ/ are shown by open points. Section 1 corresponds to the initial π/A plateaus, in section 2 surface pressure rises to the inflection value πinf, and section 3 represents the collapse of the monolayer.

4 presents ∆V, ∆Vm, and µ/, calculated from eqs 2 and 3, vs A, for partial areas of the compact films above 83% for DFEE and 90% for DEE. In region 2 (top), where π increases from zero to πinf, the DFEE molecules are upright and closely packed so that AXY ) constant and cos Θ ) 1.40 Nevertheless, the absolute values of ∆Vm and µ/ increase (become more negative). Such variation could result from an increase of the negative dipole moment component along the vertical chain µ or/and from a decrease of the positive effective permittivity  of the monolayer. Since compression depolarizes the monolayer and increases the value of ,35,42 the increase of |∆Vm| and |µ/ | should result from increase of the molecular dipole moment component along the upright chains µ. Therefore, the variation of ∆Vm and µ/ in region 2 originates from changes of the conformation or/and polarization of the hydration water of the trifluoroethyl ether heads. Region 2b in Figure 4 (bottom) corresponds to upright closely packed DEE molecules with constant area per molecule AXY ) 19.1 Å2. Despite the same values of AXY for the DEE and DFEE monolayers,40 here ∆Vm and µ/ are positive and decrease under compression. Such variation could result either from changes of the headgroup conformation or/and polarization of hydration water and an increase of the permittivity  of the DEE monolayer, but the two contributions cannot be distinguished in this case. Comparison of the Dipole Potentials of the Monolayers with Fluorinated and Nonfluorinated Heads. Table 1 compares the dipole potentials ∆Vm,inf at the π/A inflection for compact monolayers of DFEE, DEE, TFEB and EB calculated from eq 2. The values of ∆Vinf are averaged from at least five measurements, and the partial area occupied by the compact films AXY/A is obtained from the GIXD molecule areas determined previously.39,40 Table 1 suggests the following conclusions:

Negative Dipole Potentials •Substitution of the terminal CH3 groups in the ethyl ether and ethyl ester heads by CF3 groups reverses the sign of the dipole potential of their condensed Langmuir monolayers and dramatically changes its magnitude. •The changes of ∆V due to fluorination of the ethyl ether heads, ∆VDFEE - ∆VDEE ) -700 ( 13 mV, and ethyl ester heads, ∆VTFEB - ∆VEB ) -712 ( 20 mV, are practically the same. This quantitative agreement suggests the same mechanism of reversal of the sign of dipole potential of the fluorinated ether and ester monolayers. •The differences between ∆V of the nonfluorinated ester and ether films, ∆VEB - ∆VDEE ) -116 ( 15 mV, and between their fluorinated analogues, ∆VTFEB - ∆VDFEE ) -128 ( 18 mV, are also the same within the scatter limits. This fact also points to a common mechanism of reversal of the sign of ∆V. •The values of ∆VEB - ∆VDEE and ∆VTFEB - ∆VDFEE show a negative contribution of the Cδ+dOδ- bond to the dipole potential of both ethyl ester and trifluoroethyl ester monolayers. This conclusion opposes the general opinion that Cδ+dOδyields a positive contribution to ∆V of fatty acids and their esters, glycerides, and phospholipids.43-49 •A positive Cδ+dOδ- contribution of ∼+100 mV to the dipole potentials of phospholipid monolayers was found by comparing the ∆V values of PC monolayers and bilayers with ester and ether links of the hydrocarbon chains.45,54 It is intriguing but unclear why the same comparison of ∆V for the EB-DEE and TFEB-DFEE couples, gives a Cδ+dOδ- contribution of the same magnitude, but opposite (negative) sign. Mixed Monolayers of DFEE and DEE. Mixed solutions of DFEE and DEE in chloroform with a total concentration of 1 mM were spread on water and π/A and ∆V/A isotherms of the monolayers were recorded after evaporation of the solvent. Both dependencies vary systematically with the molar fraction of the fluorinated ether XDFEE. Figure 5a shows that ∆Vinf (averaged from at least five isotherms) changes smoothly from +455 ( 5 mV to -236 ( 8 mV when XDFEE increases from 0 to 1. The negative deviation of the ∆Vinf/XDFEE dependence from the additive law presented by the dashed line indicates attraction between the DFEE and DEE molecules causing depolarization of the mixed films50 This could be due to the opposite dipoles of the fluorinated and nonfluorinated heads and/or to the van der Waals attraction between the long chains. The good reproducibility of the ∆Vinf values, whose scatter is comparable with the experimental error, the smooth variation of ∆Vinf with XDFEE, and the negative deviation from additivity, implies miscibility of DFEE and DEE at the air-water interface. According to the Gibbs phase rule applied for Langmuir monolayers51 πinf of mixed films should remain constant if their components form separate phases. Figure 5b shows that this is not the case; πinf varies with XDFEE, implying miscibility of DFEE and DEE. The same holds for the surface pressure of the L2′-S phase transition πtr, which decreases with increasing XDFEEsFigure 5c. Although this variation can be followed only up to XDFEE ) 0.5, the change of this rigorous thermodynamic quantity gives a convincing support to the conclusion about the molecular miscibility of DFEE and DEE in Langmuir films. The miscibility of DFEE and DEE molecules at the airwater interface can be understood on molecular level on the basis of our GIXD study.40 It showed that at high surface pressure the 2D-lattices of DFEE and DEE monolayers have the same unit cell parameters and the same molecular area AXY. Therefore, replacement of DEE by DFEE molecules should not change the isomorphic S-phases giving the same molecular arrangement in their mixtures.

J. Phys. Chem. B, Vol. 109, No. 29, 2005 14107

Figure 5. Mixed DFEE/DEE monolayers. (a) Dependence of the dipole potential ∆Vinf at the π/A inflection point vs the molar fraction XDFEE of the fluorinated ether. (b) Dependence of the inflection surface pressure πinf on XDFEE. (c) Dependence of the surface pressure of the L2′-S transition πtr on XDFEE. According to the Gibbs phase rule applied to monolayers51 a variation of πinf and πtr with XDFEE indicates molecular mixing.

Binding of Hydrophobic Ions to the DFEE and DEE Monolayers. Since the free energy of the ion-membrane interaction is proportional to the dipole moment of the lipid molecules,4 a reversal of its sign should reverse the well-known specificity of binding of hydrophobic anions and cations to phospholipid bilayers and monolayers.25,52 This prediction was checked for the DFEE and DEE monolayers via comparison of their ∆V/A and π/A isotherms on water and 10-4 M solutions of tetraphenylphosphonium cation (TPP+) and tetraphenylboron anion (TPB-). The potential of the TPP+ and TPB- subsolutions was set to 0 before spreading the monolayers in order to subtract the adsorption of TPP+ and TPB- at the pure air-water interface. Figure 6 (top) shows the π/A and ∆V/A isotherms of DFEE monolayers on pure water, TPP+ and TPB- solutions. The maximum dipole potential on the TPP+ substrate, ∆Vmax(TPP+) ) -200 mV, is shifted by +40 mV from ∆Vmax(w) ) -240 mV on water, due to binding of the cation to the monolayer. On the TPB- substrate, ∆Vmax(TPB-) ) -255 mV; here the binding of the anion yields only a -15 mV shift from ∆Vmax(w). The preferential binding of the cations to monolayers with

14108 J. Phys. Chem. B, Vol. 109, No. 29, 2005

Figure 6. Effect of hydrophobic ions, tetraphenylboron (-) and tetraphenylphosphonium (+), dissolved in the subphase at 10-4 M on π/A and ∆V/A isotherms of DFEE and DEE monolayers. Comparison with pure water subphase (w). The potentials of the subsolutions were set to 0 before spreading the films (see the text).

negative dipole potentials qualitatively follows the theory,4 but the observed ∆(∆Vmax) shifts are very small. Figure 6 (bottom) shows the π/A and ∆V/A isotherms of the DEE films. The corresponding dipole potentials are ∆Vmax(w) ) +455 mV, ∆Vmax(TPP+) ) +440 mV, and ∆Vmax(TPB-) ) +430 mV. As predicted by the theory,4 the cation binds to a lesser extent than the anion to monolayers with positive dipole potentials, but the ∆(∆Vmax) shifts of -25 mV for TPB- and -15 mV for TPP+ are again very small as for the DFEE monolayer. The above data contrast the strong modulation of the ∆V/A and π/A isotherms of DPPC monolayers by the binding of TPP+ and TPB- ionssFigure 7. The values of ∆V on water and 10-4 M solutions of TPB- and TPP+ at 42.0 Å2 (the dotted line) are ∆V(w) ) +600 mV, ∆V(TPP+) ) +520 mV, and ∆V(TPB-) ) +255 mV, respectively. Binding of TPB- to the phospholipid film changes ∆V(w) by -345 mV, while TPP+ has a much smaller effect of -80 mV. Both shifts have the same signs as those for the DEE monolayer, but the ∆(∆V) values for DPPC are 14 times (TPB-) and 5 times (TPP+) larger, respectively. Another difference between Figures 6 and 7 is that the binding of TPB- and TPP+ ions does not change the π/A isotherms of the DFEE and DEE monolayers, while the isotherm of the DPPC film on TPB- subsolution is significantly changed. The liquidexpanded state of the DPPC film is strongly promoted by the binding of TPB-, while the steep high-pressure part remains practically the same. The binding of TPP+ shifts the steep π/A part to larger areas, but does not influence the liquid-expanded and transition regions. Discussion Fluorination of the Heads and Reversal of the Sign of the Dipole Potential. The central message of this paper is that

Petrov et al.

Figure 7. Effect of 10-4 M tetraphenylboron (-) and tetraphenylphosphonium (+) subsolutions on the π/A and ∆V/A isotherms of DPPC monolayers compared with water subphase (w).

substitution of the CH3 terminal of the hydrophilic heads of long-chain ethyl ethers and ethyl esters by CF3 groups reverses the sign and dramatically changes the magnitude of the dipole potential of their condensed monolayers on water. It is wellknown that fluorination of the ω-methyl group of long chain amphiphiles yields Langmuir films with strongly negative dipole potentials.31,32,46,49,53 This occurs because of the replacement of the (Hδ+)3-Cδ- dipole by the (Fδ-)3-Cδ+ dipole whose negative end points toward the air. In the trifluoroethyl ether or trifluoroethyl ester heads the (Fδ-)3-Cδ+ dipoles are hydrated and their orientation is unknown. If the CF3-groups in the DFEE and TFEB films are accommodated beneath the long hydrocarbon chain, as Alexander and Schulman43 and Fort and Alexander48 suggest for condensed ethyl ester and ethyl ether monolayers (Figure 8) the (Fδ-)3-Cδ+ dipole would yield a positive contribution to ∆V and positive dipole potentials. Therefore, fluorination of the CH3 group in the ethyl ether and ethyl ester heads either changes their conformation or reverses the orientation of some of the water dipoles in their hydration shells. Our conclusion about the negative contribution of the Cδ+dOδ- dipoles to ∆V of both the ethyl- and trifluoroethyl ester monolayers contrasts from the generally accepted headgroup conformations of uncharged long chain fatty acids and ethyl esters. Many of these conformations were constructed to match the measured dipole moments with those calculated via summation of the bond or group dipole momentssFigure 8. However, the electrostatic theory of Taylor and Bayes42,50 shows that the dipole-dipole interactions in the condensed stearic acid films on water induce a large negative component of the dipole moment of the horizontal carbonyl bond, oriented along the upright hydrocarbon chain. Therefore, the present state of the theory and our experimental data question the published

Negative Dipole Potentials

J. Phys. Chem. B, Vol. 109, No. 29, 2005 14109 monolayers from positive to negative values (Figure 5) suggests a method for adjustment of the sign and magnitude of the dipole potential in lipid monolayers, bilayers, and biomembranes. Besides that, fluorination of the hydrophilic heads reduces the hydration of the monolayer-water boundary. Both effects strongly modulate the interaction between lipid membranes and amphiphilic peptides and proteins so that such mixtures might enable controlled variation of this interaction. Langmuir films with phase separated components,58 especially such with fluorinated vs nonfluorinated heads, would be even more interesting, because they would exhibit a local modulation of the dipole potential and hydration of the membrane-water boundary. Binding of Hydrophobic Ions to Monolayers with Fluorinated Hydrophilic Heads. The free energy of interaction between a membrane and a hydrophobic ion in the aqueous environment consists of Born image ∆GB-I, neutral ∆GN, and dipole ∆GD components:

∆Gint ) ∆GB-I + ∆GN + ∆GD Figure 8. Conformation of the ethyl ether heads, according to Ford and Alexander,48 and ethyl ester head, accepted by Alexander and Schulman,43,48 for condensed Langmuir monolayers at the air-water interface. The calculated dipole moments are obtained via summation of the bonds dipole moments.

headgroup conformations and dipole moments obtained via summation of the contributions of atomic groups and polar bonds in condensed monolayers.43-49 Because of the hydrophobic character of the CH3 and especially of the CF3 groups in the heads (see our maps of molecular lipophilicity of TFEB and EB in vacuum38), these terminals would “try” to get out of the water substrate. Such orientation has been proposed by Adam55 for esters with alkyl chains in the heads having four or more C atoms, because they occupy molecular areas of two upright hydrocarbon tails. Table 1 shows that the real molecular areas AXY of TFEB (18.8 Å2) and EB (18.9 Å2) exceed the hydrocarbon chain cross-section 18.2 Å2 in anhydrous ethyl behenate crystals,56 but these differences are only 3.3% and 3.8%, respectively. The GIXD areas per molecule of DFEE and DEE at high surface pressure are the same, AXY ) 19.0 ( 0.1 Å2, but again larger than the chain cross-section 18.2 Å2. The larger molecular cross-section in the S-phase of the monolayers on water vs the molecular cross-section in anhydrous bulk crystals, imply that the hydrated heads determine the closest packing in the Langmuir films. Since the van der Waals radius of the CF3 is larger than the CH3 radius by 15%,57 the same molecular areas of the closely packed upright fluorinated and nonfluorinated molecules suggest thinner hydration shells of the DFEE and TFEB heads. The lower πcol and πinf, demonstrating decreased stability of the fluorinated monolayers against compression, support this conclusion, because they indicate weaker fluorinated heads-water adhesion. As commented by us previously,40 the existence of the tilted L2′ phase only in the DEE and EB films could result from a larger cross section of the nonfluorinated heads, causing a tilt of the chains to achieve their closest packing. All these facts suggest that the reversal of the sign of the dipole potential caused by the fluorination of the DEE and EB heads might originate from structure of the hydration shells. It is known that if only 2% of the water dipoles at the hydrocarbon/water interface were oriented perpendicular to it they would change the dipole potential by 100 mV.41 Mixed Monolayers with Fluorinated and Nonfluorinated Heads. The smooth variation of ∆V of the mixed DFEE + DEE

(4)

Flewelling and Hubblle4 presented the Born image energy ∆GB-I as

∆GB-I(x) )

q2 r r x2 1 - - 1.2 2rhf 2x d d

[

( )( ) ]

(5)

Here, q is the charge, r the radius, and x the penetration depth of the ion in a membrane of thickness d, and hf is the average permittivity of the region in which the hydrophobic ion binds. ∆GB-I is proportional to q2 and positive for both TPP+ and TPB- ions. However, when x < r/2, ∆GB-I becomes negative, and the Born image interactions facilitate the binding of the hydrophobic ions to the membrane. The neutral energy ∆GN results from the hydrophobic attraction between TPB- and TPP+ and the hydrocarbon moieties of the bilayers or monolayers:

∆GN(x) )

∆G0N 1 + 10-2x/r

(6)

∆G0N is the free energy of transfer of a hydrophobic ion from an aqueous to a hydrophobic surrounding. It is negative and varies between -4 and -8 kcal/mol4,25 so that ∆GN is always negative and helps the binding of both TPB- and TPP+ to the membrane. The theory4 relates the specificity of the interaction of the anion TPB- and the cation TPP+ with phospholipid bilayers to the dipole component of the free energy, ∆GD:

∆GD(x) )

qµ L3 a (x) 2

(7)

Here µ is the dipole moment of the lipid molecules considered as point dipoles, a is the distance between these dipoles, and L3 is a 3D-lattice function summed-up over the 2D-dipole layer. The dielectric function (x) characterizes the transition region of thickness h, between water and the hydrocarbon core of the membrane. Our GIXD study40 showed that at high surface pressure the DFEE and DEE molecules arrange in an orthorhombic lattice with the same unit-cell parameters so the a ) (AXY)1/2 and the lattice function L3 should be the same for DFEE and DEE monolayer. The fluorinated heads of DFEE poses negative dipole moments µ so that qµ and ∆GD are negative for the cation TPP+ and positive for the anion TPB-. This difference facilitates the binding of the cation and confines the

14110 J. Phys. Chem. B, Vol. 109, No. 29, 2005 binding of the anion. For the DEE film µ is positive, and the binding of the TPB- anion is promoted, because qµ and ∆GD are negative, while the positive qµ and ∆GD for the TPP+ restrict the binding of the cation to the DEE monolayers. Our results for ∆V of DFEE and DEE on TPB- and TPP+ subsolutions qualitatively confirm the prediction of the theory4 that the reversal of the sign of the dipole moment of the heads reverses the preference in the binding of the hydrophobic anions and cations. However, the binding of both TPB- and TPP+ is extremely small. The much stronger effect of TPB- and TPP+ on ∆V of DPPC monolayers (Figure 7) seems to have a nonelectrostatic origin, because large negative shifts of the dipole potential are observed for both the anion and the cation. This fact suggests strong attractive interactions of both TPB- and TPP+ with the DPPC monolayers, probably due to a deeper penetration in the DPPC film and more significant neutral interaction energy ∆GN. Previous studies have shown that, at concentrations below 10 mol % with respect to the lipid, these ions bind in the alkyl chains region of egg-PC bilayers.59 Such penetration yields a positive Born image energy, but the negative ∆GN component increases much stronger with increasing x/r as compared to ∆GB-I (cf. eqs 5 and 6). Because of the solid (polycrystalline) character of the DFEE and DEE monolayers the penetration depth x/r of the TPB- and TPP+ ions should be negligible or very small. Small values of x/r give a negative ∆GB-I term, but the main attractive component of the interaction energy ∆GN is also very small. Therefore, the small ∆V-shifts for the DFEE and DEE films could be due to limited penetration of the TPB- and TPP+ ions binding either below the microcrystals or at their boundaries. Elena et al.59 found that at concentrations above 20 mol % the hydrophobic anion TPB- forms a complex with the positively charged choline group of PC bilayers. Such a specific binding could also explain the difference between the DEE and DPPC monolayers. For DPPC films on 10-4 M TPB- solutions one finds at 42 Å2/molecule a molar ratio TPB-:DPPC ) 250: 1, which exceeds the above limit of 20 molar % by 4 orders of magnitude. Biological Significance of the Present Results and Future Investigations. A negative dipole potential due to the polar heads was reported by Adam et al.60 for epi-cholestanol films on water, while its isomer cholestanol, having just one OH group with opposite orientation, forms monolayers with positive ∆V values. It is intriguing that only the cholestanol is present in living systems with normal functions, the appearance of epicholestanol accompanies some brain deaths. Many other biorelevant amphiphiles, such as cholesterol, mono-, di- and triglycerides, all zwitterionic phospholipids, and the charged ones on subsolutions with high ionic strength, also form Langmuir films with positive dipole potentials.61 Undissociated fatty acids, their methyl, ethyl, and higher esters, amines, alcohols, ethers, ketones, nitrils, and many other uncharged amphiphiles give condensed monolayers with positive dipole potentials too.43,48,49,62,63 In a systematic study, Smaby and Brockman64 determined the dipole potentials of monolayers of 38 biorelevant lipids of 19 chemical classes and also found only positive ∆V values. The dipole potential of lipid bilayers is also positive;3,25 fully saturated phosphatidylcholine membranes have dipole potentials between +220 and +280 mV. In this context, the monolayers with fluorinated hydrophilic heads represent a new class of membranes with unusual, if not exceptional properties. The most interesting extension of our investigations is the synthesis of phospholipids with fluorinated headgroups. This

Petrov et al. could be done via substitution of the CH3 terminals of the choline group by CF3CH2 terminals. Another possibility is to follow the approach of McDonald et al.65 who synthesized and thoroughly studied a new class of cationic alkyl phosphocholine triesters having a CH3CH2 group attached to the phosphate oxygen. Attaching a CF3CH2 group at the same position would give fluorinated analogues of the above cationic phospholipids. The phospholipids with fluorinated heads can be studied both as liposome suspensions in bulk and as monolayers on water. Their structural similarity to the alkyl phosphocholine triesters suggests a number of interesting properties such as the absence of intermolecular hydrogen bonding and easy release of single molecules from liposomes, formation of DNA-lipid complexes, etc.65 On the other side, the fluorinated heads should modulate the interaction between the phospholipid membranes and amphiphilic peptides and proteins. The fluorinated heads containing CF3CH2O groups should change the secondary structures of the macromolecules at the membrane-water boundary in a manner similar to the changes of peptides and proteins in CF3CH2OH/H2O solutions.66,67 Conclusions 1. Substitution of the terminal CH3 group in the hydrophilic head of docosyl ethyl ether (DEE) by the CF3 group reverses the positive sign of the dipole potential of DEE monolayers and changes its magnitude by 152%. The monolayers of the fluorinated docosyl trifluoroethyl ether (DFEE) show reduced collapse surface pressure and absence of a tilted phase at low surface pressure as compared to the nonfluorinated DEE film. The same changes of the sign of ∆V, mechanical properties, and phase behavior of Langmuir monolayers of the ethyl ester of behenic acid (EB) were found37-39 for films of its fluorinated analogue trifluoroethyl behenate (TFEB). 2. The change of ∆V due to fluorination of the ethyl ether heads, ∆VDFEE - ∆VDEE ) -700 ( 13 mV, is practically the same with ∆VTFEB - ∆VEB ) -712 ( 20 mV for the previously studied TFEB-EB pair. The difference between ∆V of the nonfluorinated ester and ether films, ∆VEB - ∆VDEE ) -116 ( 15 mV, and between their fluorinated analogues, ∆VTFEB ∆VDFEE ) -128 ( 18 mV, is also the same within the scatter limits. This quantitative coincidence suggests the same mechanism of reversal of the sign of the dipole potential for both fluorinated ether and ester monolayers. 3. Comparison of the GIXD molecular areas of the above amphiphiles at highest monolayer density with the molecular cross section in anhydrous bulk crystals shows that the packing of all four films is determined by their hydrophilic heads. Since the CF3 group is larger than the CH3 group, the fluorinated heads are less hydrated than the nonfluorinated ones. The existence of a tilted L2′ phase in DEE and EB monolayers and its absence in the DFEE and TFEE films supports this conclusion.40 The reduction of πcol due to fluorination of the heads points to weaker head-water adhesion. The increase of the negative dipole potential ∆Vm and apparent molecular dipole moment along the chain µ/ of the compact DFEE film (region 2 in Figure 4a) originates from changes of the conformation or/and polarization of the hydration water of the trifluoroethyl ether heads. All these facts suggest significant changes of hydration water due to fluorination, but do not exclude different conformation of the polar heads, because hydration and conformation of the polar groups are interrelated. 4. The differences ∆VEB - ∆VDEE and ∆VTFEB - ∆VDFEE show a negative contribution of the Cδ+dOδ- group to the dipole potential of both ethyl ester and trifluoroethyl ester

Negative Dipole Potentials monolayers. This conclusion contradicts the general opinion that Cδ+dOδ- yields a positive contribution to the dipole potential of condensed monolayers of fatty acids and their n-alcohol esters, glycerides, and phospholipids.43-49 However, it concurs the results of the electrostatic theory of Taylor and Bayes35,42,50 for condensed uncharged monolayers, which shows that the dipole-dipole interactions in stearic acid films induce a strong negative component of the horizontal carbonyl bonds oriented along the vertical hydrocarbon chains. Therefore, the present stage of the theory and our result question the literature data for molecular dipole moments in Langmuir films calculated via summation of the dipoles of atomic groups and polar chemical bonds. 5. The smooth monotonic variation of the dipole potential of mixed DFEE+DEE monolayers from +450 to -235 mV suggests a way for adjustment of the sign and magnitude of the dipole potential at the membrane-water boundary and regulation of such membrane behaviors as binding and translocation kinetics of hydrophobic ions and ion-carriers, adsorption and penetration of amphiphilic peptides and proteins, and polarization of hydration water determining the short-range repulsion between hydrophilic surfaces. 6. The ∆(∆V) shifts of the dipole potential, caused by adsorption of the hydrophobic tetraphenylboron TPhB- and tetraphenylphosphonium TPhP+ ions at DFEE and DEE monolayers, qualitatively follow the theory of Flewelling and Hubbell4 for binding of such ions to lipid bilayers, but are much smaller than ∆(∆V) for DPPC monolayers. This difference seems to be due to the hampered penetration of TPhB- and TPhP+ in the solid DFEE and DEE monolayers, which reduces the attractive interaction energy with the hydrophobic moieties. This conclusion orients the future synthesis to amphiphiles with fluorinated heads which could form liquid-expanded Langmuir monolayers. Acknowledgment. J.G.P. gratefully acknowledges the Alexander von Humboldt Foundation for the 3 months research stipend in 2003, which enabled this study. T.D.A. is indebted to the Max-Planck Society for the three months promotion stipend at Max-Planck Institute of Colloids and Interfaces in Golm, Germany. We thank Prof. V. Shapovalov for supplying his apparatus for a part of the ∆V measurements. References and Notes (1) Brockman, H. Chem. Phys. Lipids 1994, 73, 57. (2) Brockman, H. Curr. Opin. Struct. Biol. 1999, 9, 438. (3) Clarke, R. J. AdV. Colloid Interface Sci. 2001, 89-90, 263. (4) Flewelling, R. F.; Hubbell, W. L. Biophys. J. 1986, 49, 541. (5) Cafiso, D. S. Curr. Opin. Struct. Biol. 1991, 1, 185. (6) Cafiso, D. S. Influence of charges and dipoles on macromolecular adsorption and permeability. In Permeability and stability of lipid bilayers; DiSalvo, E. A., Simon, S. A., Eds.; CRC Press: Boca Raton, FL, 1995. (7) Cladera, J.; O’Shea, P. Biophys. J. 1998, 74, 2434. (8) McIntosh, T. J.; Vidal, A.; Simon, S. A. Biochem. Soc. Trans. 2001, 29, 594. (9) Allende, D.; Vidal, A.; Simon, S. A.; McIntosh, T. J. Chem. Phys. Lipids 2003, 122, 65. (10) Maggio, B. J. Lipid Res. 1999, 40, 930. (11) Qin, Zh.; Szabo, G.; Cafiso, D. S. Biochemistry 1995, 34, 5536. (12) Cafiso, D. S. Toxicology Lett. 1998, 100-101, 431. (13) Cladera, J.; Martin, I.; Ruysschaert, J. M.; O’Shea, P. J. Biol. Chem. 1999, 274, 29951. (14) Belaya, M. L.; Feigel’man, M. V.; Levadnyii, V. G. Langmuir 1987, 3, 648. (15) Simon, S. A.; McIntosh, T. J. Proc. Natl. Acad. U.S.A. 1989, 86, 9263. (16) Simon, S. A.; McIntosh, T. J.; Magid, A. D.; Needham, D. Biophys. J. 1992, 786. (17) McIntosh, T. J.; Simon, S. A.; Needheam, D.; Huang, C. Biochemistry 1992, 31, 2020. (18) McIntosh, T. J.; Simon, S. A. Biochemistry 1993, 32, 8374.

J. Phys. Chem. B, Vol. 109, No. 29, 2005 14111 (19) McIntosh, T. J.; Simon, S. A. Annu. ReV. Biophys. Biomol. Struct. 1994, 23, 27. (20) Jandrasiak, G. L.; Smith, R. L.; McIntosh, T. J. Biochim. Biophys. Acta 1997, 1329, 159. (21) Cevc, G. Biochim. Biophys. Acta 1990, 1031, 311. (22) McConnell, H. M. Annu. ReV. Phys. Chem. 1991, 42, 171. (23) McConnell, H. M.; De Koker, R. Langmuir 1996, 12, 4897. (24) Andersen, O. S.; Finkelstein, A.; Katz, I.; Cass, A. J. Gen. Physiol. 1976, 67, 749. (25) Franklin, J. C.; Cafiso, D. S. Biophys. J. 1993, 65, 289. (26) Le Fevre, P. G.; Marshall, J. K. J. Biol. Chem. 1959, 234, 3022. (27) Levine, S.; Franki, N.; Hays, R. M. J. Clin. InVest. 1973, 52, 1436. (28) Voglino, L.; McIntosh, T. J.; Simon, S. A. Biochemistry 1998, 37, 12241. (29) Voglino, L.; Simon, S. A.; McIntosh, T. J. Biochemistry 1999, 38, 7509. (30) Davis, J. T.; Rideal, E. Can. J. Chem. 1955, 33, 947. (31) Fox, H. W. J. Phys. Chem. 1957, 61, 1058. (32) Bernett, M. K.; Zisman, W. A. J. Phys. Chem. 1963, 67, 1534. (33) McIntosh, T. J.; Simon, S. A.; Vierling, P.; Santaella, C.; Ravily, V. Biophys. J. 1996, 71, 1853. (34) Santaella, C.; Fre´zard, F.; Vierling, P.; Riess, J. G. FEBS Lett. 1993, 336, 481. (35) Taylor, D. M.; Bayes, G. F. Phys. ReV. E. 1994, 49, 1439. (36) Petrov, J. G.; Polymeropoulos, E. E.; Mo¨hwald, H. J. Phys. Chem. 1996, 100, 9860. (37) Petrov, J. G.; Mo¨hwald, H. J. Phys. Chem. 1996, 100, 18458. (38) Petrov, J. G.; Polymeropoulos, E. E.; Mo¨hwald, H. Langmuir 2000, 16, 7411. (39) Petrov, J. G.; Brezesinski, G.; Krasteva, N.; Mo¨hwald, H. Langmuir 2001, 17, 4581. (40) Petrov, J. G.; Brezesinski, G.; Andreeva, T. D.; Mo¨hwald, H. J. Phys. Chem. B 2004, 108, 16154-16162. (41) Haydon, D. A.; Hladky, S. B. Q. ReV. Biophys. 1972, 5, 187-282. (42) Taylor, D. M. AdV. Colloid Interface Sci. 2000, 87, 183. (43) Alexander, A. E.; Schulman, J. H. Proc. R. Soc. London, A 1937, 161, 115. (44) Davis, J. T.; Rideal, E. Interfacial Phenomena; Academic Press: New York, 1961. (45) Paltauf, F.; Hauser, H.; Phillips, M. C. Biochim. Biophys. Acta 1971, 249, 539. (46) Vogel, V.; Moebius, D. J. Colloid Interface Sci. 1988, 126, 408. (47) Taylor, D. M.; Oliveira, O. N.; Morgan, H. J. Colloid Interface Sci. 1990, 139, 508. (48) Fort, T.; Alexander, A. E. J. Colloid Science 1959, 14, 190. (49) Demchak, R. J.; Fort, T., Jr. J. Colloid Interface Sci. 1974, 46, 191. (50) Taylor, D. M.; Bayes, G. F. Mater. Sci. Eng. 1999, C8-C9, 65. (51) Gaines, G. L., Jr. Insoluble monolayers at liquid-gas interfaces, Interscience Publishers: New York, London, Sydney, 1966; p 281. (52) Shapovalov, V. L. Thin Solid Films 1998, 327-329, 599. (53) Oliveira, O. N., Jr.; Taylor, D. M.; Lewis, T. J.; Salvagno, S.; Stirling, C. J. M. J. Chem. Soc. 1989, 85, 1009. (54) Gawrisch, K.; Ruston, D.; Zimmerberg, J.; Parsegian, V. A.; Rand, R. P.; Fuller, N. Biophys. J. 1992, 61, 1213. (55) Adam, N. K. Proc. R. Soc. London, A 1929, 126, 366. (56) Mathieson, McL. A.; Welsh, H. K. Acta Crystallogr. 1965, 18, 953. (57) Bondi, A. J. Phys. Chem. 1964, 68, 441. (58) Gege, Ch.; Schneider, M. F.; Schumacher, G.; Limozin, L.; Rothe, U.; Bendas, G.; Tamnaka, M.; Schmidt, R. R. ChemPhysChem 2004, 5, 216. (59) Elena, J. F.; Dominey, R. N.; Archer, S. J.; Xu, Zh. Ch.; Cafiso, D. S. Bochemistry 1987, 26, 4584. (60) Adam, N. K.; Askew, F. A.; Danielli, J. F. Biochem. J. 1935, 29, 1786. (61) Papahadjiopoulos, D. Biochim. Biophys. Acta 1968, 163, 240. (62) Stenhagen, E. Determination of organic structures by physical methods; Braude, E. A., Nachod, F. C., Eds.; New York, 1955. (63) Adam, N. K. The physics and chemistry of surfaces, 3rd ed.; Oxford University Press: London, 1941. (64) Smaby, J. M.; Brockman, H. L. Biophys. J. 1990, 58, 195. (65) MacDonald, R. C.; Ashley, G. W.; Shada, M. M.; Rakhmanova, V. A.; Tarahovsky, Y. S.; Pantazatos, D. P.; Kennedy, M. T.; Pozharski, E. V.; Baker, K. A.; Jones, R. D.; Rosenberg, H. S.; Choi, K. L.; Qui, R.; McIntosh, T. J. Biophys. J. 1999, 77, 2612. (66) Rajan, R.; Balaram, P. Int. J. Peptide Protein Res. 1996, 48, 328. (67) Cammers-Goodwin, A.; Allen, T.; Oslik, S.; McClure, K.; Lee, J.; Kemp, D. S. J. Am. Chem. Soc. 1996, 118, 3082.