Thin Solid Films, 178 (1989) 73-79
LATERAL CONDUCTION AT A MONOLAYER-WATER INTERFACE H. MORGAN,D. M. TAYLORAND O. N. OLIVEIRA,JR.* Institute of Molecular and Biomolecular Electronics, University of Wales, Bangor, Dean Street, Bangor, Gwynedd LL571UT (U.K.)
(ReceivedApril 25, 1989;acceptedApril 28, 1989)
Measurements are reported o f the lateral conductance and surface potential o f stearic acid and phospholipid monolayers at the air-water interface. Both quantities show a rapid increase and strong dependence on packing density, once the monolayer has been compressed below a critical area per molecule. It is argued that these effects are likely to arise from the formation o f a hydrogen-bonded network between the monolayer head group and nearest-neighbour water molecules. The concomitant decrease in local permittivity would contribute to the increase in surface potential while proton transport would be facilitated along the interface, leading to the enhanced conductance observed.
l. INTRODUCTION Phospholipid monolayers at the air-water interface are widely regarded as models of biomembranes 1 and have been used to elucidate the mode of proton transport between membrane-bound proton donors and acceptors. Recent fluorescent probe measurements by Prats et al.2 provide evidence in favour of the localized chemiosmotic theory 3 in which proton transport is confined to the membrane surface. Direct electrical evidence for this model was provided in a recent communication 4 in which it was shown that monolayers of dipalmitoyl phosphatidylethanolamine (DPPE) exhibited an enhanced lateral conductance when the molecular packing density exceeded a critical value. In this paper further results are presented which indicate that the effect is common to a wider range of compounds. 2. MATERIALSAND METHODS The lipids investigated were D P P E and dipalmitoyl phosphatidic acid (DPPA) obtained from Koch Light Ltd. The stearic acid samples were obtained from Sigma Ltd. For spreading, stearic acid was dissolved in chloroform and the
*On leaveof absencefrom Instituto de Fisicae Quimicade Sao Carlos, USP, Brasil. 0040-6090/89/$3.50
© ElsevierSequoia/Printedin The Netherlands
H. MORGAN, D. M. TAYLOR, O. N. OLIVEIRA, JR.
lipids in a 1 : 4 mixture of methanol and chloroform (both solvents were o f H P L C grade). The experiments were performed on a polytetrafluoroethylene trough (60 cm x 26cm x 2 cm) placed on a thermostatically controlled metal baseplate located on an antivibration table. The whole apparatus was housed in a temperature and humidity-controlled semiconductor clean room. Ultrapure water for the experiments was obtained from a Millipore Super-Q system comprising reverse osmosis, ion exchange, organex and 0.2g m ultrafiltration cartridges. When in equilibrium with carbon dioxide, the p H of the trough water was 5.6. The surface pressure n was measured with a Wilhelmy plate and electrobalance to better than 0.1 mN m - 1 accuracy. A vibrating probe with accuracy better than 10 mV was used to measure surface potential A V. The lateral conductance G was measured 4 by inserting two plane parallel bright platinum electrodes into the water surface, applying a constant bias of 1.5 V and recording the resultant current flow with a Keithley model 616 electrometer. Because of the small conductance changes involved, mechanical disturbances o f the trough must be minimized. It is essential also that only water of the highest purity is used and that the electrodes are free from contamination. Electrodes must be cleaned regularly by ultrasonication in hot chloroform and boiling in pure water for 5 min. Unless these precautions are taken, a stable background current through the subphase water cannot be achieved. During monolayer compression, n, A V and G were continuously recorded as a function of area per molecule A and the data stored in a computer. 3. RESULTS Figure 1 gives representative plots, from at least ten experiments per compound, of surface pressure, surface potential and lateral conductance as a function of area per molecule for stearic acid, DPPA and DPPE. The n vs. A isotherms are consistent with published work. As can be seen, in a well-expanded monolayer the lateral conductance is independent of the area per molecule and is determined entirely by the conductivity of the subphase water and the depth of immersion of the electrodes. However, when the area per molecule decreases below a critical value Ac, the conductance begins to increase and continues to increase until the surface pressure begins to rise. At this area, the reduction in surface tension of the subphase surface causes the meniscus to move down the electrode, effectively reducing the contact area, and so reducing the measured current. Examination of the meniscus using a microscope and video recording equipment confirmed that the height of the meniscus decreased by 0.55 mm as the surface pressure rose from zero to 5 0 m N m -1. Within the 0.01mm resolution of the microscope, no change occurred in the height of the meniscus during the first stages of compression. Some hysteresis was observed in the lateral conduction d u r i n g a compression-expansion cycle. However, the conductance returned to its initial value after fully expanding the monolayer for about 15min, indicating that the conductance changes are reversible. To eliminate the meniscus effect, microelectrodes of the type described by Widrig et al. s were constructed. These contact only the monolayer and the underlying water to a depth o f approximately 40 nm. Preliminary experiments with
LATERAL C O N D U C T I O N AT MONOLAYER--WATER INTERFACE
the microelectrodes have shown that the lateral conductance continues to increase with film compression until collapse occurs. The critical area Ac at which the conductance increases is a characteristic of the compound investigated and is more than double the molecular area in the condensed phase. In our previous publication4, the value reported for Ac in DPPE was larger than in Fig. 1. This difference has now been traced 6 to the quality of the water drawn from the purification system used previously. The surface potential plots in Fig. 1 show that for each compound the area per molecule at which the surface potential becomes non-zero corresponds exactly to the critical area for the onset of lateral conduction. 4. DISCUSSION It is well known that, at the pH of the experiment, monolayers of stearic acid and DPPA will be negatively charged so that at the monolayer-water interface a double-layer forms with a diffuse layer of protons extending into the water subphase. It could be argued therefore that the increased conductance for these two compounds arises simply from the higher proton concentration in the diffuse part of the interfacial double layer as discussed below.
4.1. Conduction through the double-layer The enhanced conductance AG of the interfacial water resulting from the formation of the double layer will be given by W
f ' ~o
AG = T q # ]" p(z)dz
where Wis the width of the electrodes, L is the interelectrode spacing, q and/~ are the charge and mobility of the proton and p(z) is the excess proton concentration appearing at a depth z below the water surface as a consequence of the formation of the double layer. Equation (1) applies so long as the length of the electrodes protruding into the water greatly exceeds the double layer thickness. Applying Gauss' theorem, eqn. (1) may be simplified to A G - W#IQsl L
where Qs is the surface charge density of the monolayer. For stearic acid the maximum change in conductance is 2.6 x 10-8S (obtained at 50mNm-1 after correcting for the meniscus effect). On the assumption that the mobility of the excess protons in the double layer is equal to the value in bulk water, Qs is estimated to be 5.4 x 10-2 cm-2. This suggests that about 1 in 15 of the stearic acid molecules is ionized, corresponding to a degree of dissociation a of 0.067 for the carboxyl head group. Simultaneously solving the Henderson-Hasselbalch equation v pKcoo. = PH,--loglo (]~_~)
H. MORGAN, D. M. TAYLOR, O. N. OLIVEIRA, JR.
3 £6- ~
%'=-o.__~ - ~-~x'-x'X~'x'x-x"x'x'x'x-~-x-x- x -x.x.x-x-x-x
z~O molecule, A (~z}
3 .~. 8O
o 0.2 x, x
cu X~ X-- X-- X ~ X - - X - - X - X
AV "l'1--1.14.1~1 _ i_1__ I
I 120 A ( A zl
LATERAL CONDUCTION AT MONOLAYER--WATER INTERFACE
,x x/ . . . .
.,_ x,×, ~
X'x, 0 ~'e~ - "X'x'x-x-x-x-x-x-X-X
area per molecule . A (A2)
(c) Fig. 1. Dependences o f surface pressure n ( ), surface potential A V (O) and lateral conductance G ( x ) on the area per molecule for monolayers of(a) stearic acid, (b) DPPA and (c) DPPE. All monolayers were formed on pure water (of pH 5.6, because of dissolved CO2) at a temperature o f 20°C. When the barriers are closed in the absence of a monolayer, the measured conductance remains constant at a value determined by the length o f electrode immersed in the subphase.
for the carboxyl head group and the Boltzmann equation ~so = 2 " 3 k T ( p H i - p H b ) q
where pH i and pH b are the interfacial and bulk pH respectively and ~ko is the doublelayer potential, enables ~b0 to be calculated from the conductance measurements assuming that pKcoon and pHb are known. The value presented in our earlier paper 4, - 1 2 5 m V , seemed in good agreement with the surface potential data obtained by Spink a in acid titration experiments on stearic acid monolayers formed on a 0.01 M NaCI subphase. However, in recent experiments 9 we have shown that for dilute NaCI subphases, ~cdepends on the ionic strength o f the subphase even when the pH is held constant, ct decreasing from 0.034 at 0.01 M to 0.002 in a pure water subphase (2.5 x 10 -6 M). The latter value o f ~ is at least an order of magnitude lower than predicted from conductivity data. Calculations for D P P A lead to a similar discrepancy in the values o f 0t determined from conductance and surface potential measurements. It appears therefore that the degree of monolayer dissociation on a
H. MORGAN, D. M. TAYLOR, O. N. OLIVEIRA, JR.
pure water subphase is much smaller than originally thought; so the enhanced conductance associated with the charged monolayers cannot arise from the higher proton concentration in the double layer. To some extent this is corroborated by the fact that a monolayer of the neutral lipid DPPE, which does not form a double layer, nevertheless displays an enhanced lateral conductance when compressed. 4.2. Proton conduction through an interfacial hydrogen-bonded network There is much evidence, at least for phospholipids, to support the view that the enhanced lateral conductance observed in the present work arises from the formation of a hydrogen-bonded network at the monolayer-water interface which allows facile proton transport. For example, it has been shown that the conductivity of lipids is independent of lipid chain length lo but depends strongly on the degree of hydration of the head groups 11. Acid injection experiments 2'12 have shown that proton diffusion rates at the interface of a condensed phospholipid layer and water are some 20 times greater than in bulk water. A hydrogen-deuterium isotope dependence has also been reported for the lateral conductance of poorly characterized lipid "monolayers" 13. Furthermore, lipid head groups possess both hydrogen-donating and hydrogen-accepting moieties and these are known 14 to participate in intramolecular and intermolecular hydrogen bonding in lipid membranes. X-ray diffraction studies on crystalline phospholipids is reveal the presence of structured sheets of water hydrogen bonded to the lipid head groups on either side. The carboxyl head groups of alkanoic acids may also form hydrogen bonds 16. The increase in A V which occurs at Ac suggests that substantial molecular reordering occurs at the onset of the enhanced conductance. Ellipsometric measurements ~? confirm that for a number of fatty acid and phospholipid monolayers the increase in AVis accompanied by a change in the refractive index of the layers, again indicative of structural changes in the monolayer. Finally, Prats et al. 2'12 were unable to detect proton transport when DPPE monolayers were expanded above 120/~ 2 per molecule and showed that the efficiency of proton transport (conductivity) depended on molecular packing but that the kinetics (mobility) did not. Taken together, the evidence is consistent with the formation of a hydrogenbonded network between monolayer head groups and adjacent water molecules. The sudden decrease in local permittivity as the water becomes structured will manifest itself as a rapid change in A 1I. As the packing density is increased, longrange connectivity between hydrogen-bonded molecules becomes more likely, giving rise to the observed increase in conductance. 5. CONCLUSIONS
It has been shown that, when stearic acid or phospholipid monolayers are compressed, a simultaneous increase occurs in the surface potential and the lateral conductance of the monolayer. On the basis of these measurements and from the evidence available in the literature, it is suggested that these events are the result of hydrogen bonding between the monolayer head group and the adjacent water
LATERAL CONDUCTION AT MONOLAYER--WATER INTERFACE
molecules such that a two-dimensional proton conducting network is established at the monolayer-water interface. ACKNOWLEDGMENTS
The authors wish to acknowledge the support of the Science and Engineering Research Council for supporting this work. One of us (O.N.O.) also wishes to thank Fundaq~o de Amparo $ Pesquisa do Estado de S~o Paulo (Brazil) and Overseas Research Students Award (U.K.) for a research studentship and financial support. REFERENCES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
D.M. Teide, Biochim. Biophys. Acta, 811 (1985) 357. M. Prats, J. Teissie and J. F. Tocanne, Nature (London), 322 (1986) 756. D.B. Kell, Biochim. Biophys. Acta, 549 (1979) 55. H. Morgan, D. M. Taylor and O. N. Oliveira, Jr., Chem. Phys. Lett., 150 (1988) 311. C.A. Widrig, C. G. Miller and M. Majda, J. Am. Chem. Soc., 110 (1988) 2009. D.M. Taylor, O. N. Oliveira, Jr., and H. Morgan, Thin Solid Films, 173 (1989) L141. E. Havinga and M. D. Hertog-Polag, Rec. Tray. Chim. Pays-Bas, 71 (1952) 64. J, A. Spink, J. ColloidSci., 18 (1963) 512. D.M. Taylor, O. N. Oliveira, Jr., and H. Morgan, Chem. Phys. Lett., in the press. G.J. Jendrasiak and J. H. Hasty, Biochim. Biophys. Acta, 348 (1974) 45. G.J. Jendrasiak and J. C. Mendible, Biochim. Biophys. Acta, 424 (1976) 149. M. Prats, J. F. Tocanne and J. Teissie, Fur. J. Biochem., 149 (1985) 663. I. Sakurai and Y. Kawamura, Biochim. Biophys. Acta, 904 (1987) 405. J.M. Boggs, Biochim. Biophys. Acta, 906 (1987) 353. H. Hauser, I. Pascher, R. H. Pearson and S. Sundell, Biochim. Biophys. Acta, 650 (1981) 21. T.H. Haines, Proc. Natl. A cad. Sci. U.S.A., 80 (1983) 160. D. Ducharme, C. Salesse and R. M. Leblanc, Thin Solid Films, 32 (1985) 83.