摘要:

Surface dilational behaviour of spread dipalmitoyl phosphatidyl glycerol monolayers d ln A (refs. 11–14). d R. Wüstneck,a P. Enders,b N. Wüstneck,c U. Pison,c R. Millerd and D. Vollhardtd aUniversität Potsdam, Institut für Physik, Physik kondensierter Materie, Kantstr. 55, 14513 Teltow , Germany, b Fischerinsel 2, 10179 Berlin, Germany c Humboldt-Universität Berlin, Virchow-Klinikum, Anaesthesiologie, Augustenburger Platz 1, 13344 Berlin, Germany d Max-Planck-Institut für Kolloid- und Grenzflächenforschung, Max-Planck-Campus, Haus 2, Am Mühlenberg 2, 14476 Golm, Germany Received 11th August 1999, Accepted 27th September 1999, Published 7th October, 1999 The stress relaxation process in spread dipalmitoyl phosphatidyl glycerol (DPPG) monolayers was investigated by using the pendant drop technique as a microfilm balance in combination with the axisymmetric drop shape analysis.The stress relaxation caused by a transient drop volume change was analysed by a new model, which takes into account material inhomogeneity by means of a distribution of relaxation times. The surface pressure decay in the DPPG monolayers at 35 and 40 ºC is governed by 2 main relaxation times. One is of the order of 20 s, the other one is in the range of 200–600 s. The p/A isotherms, the surface dilational modulus, and the higher relaxation times are influenced by temperature and by the presence of electrolyte in the subphase; 0.15 mol dm–3 NaCl yields a smaller area demand per molecule in the monolayer. Harmonic oscillation experiments were carried out and the dilational elasticities and viscosities were determined for different frequencies in the range between the two main relaxation times.Both the elasticities and the viscosities, depend on frequency. The surface rheological behaviour is structurally viscous, i.e. high viscosities at small frequencies, which decrease at faster deformations. The results are related to the surface dilational rheological behaviour of dipalmitoyl phosphatidyl choline (DPPC) monolayers. exposes most of all its elastic properties under the conditions of breathing. The relaxation times of the other components of the pulmonary surfactant are unknown. Although these components represent only a small part of the lipids the surface rheological properties of mixtures may differ from those of the single components.It was suggested for instance that mixtures of DPPC containing saturated acyl chains and of palmytoyl oleoyl phosphatidyl choline or glycerol containing one unsaturated chain differ in mobility or fluidity of the respective lipid mono- and bilayers.9 For a better understanding of the surface rheological properties of a complex mixture each lung surfactant component has to be characterised separately. Therefore the first aim of the present paper is to investigate the stress relaxation process of DPPG monolayers and to compare it with that of DPPC under comparable conditions.10 Such data will be the basis of further investigations of model mixtures of pulmonary surfactant.It should be noticed that usually the mechanical properties of pulmonary layers are characterised by monolayer compressibility, which is calculated by the first derivative of the p/A isotherm, k = Usually, p/A isotherms are measured at a definite surface compression rate and, consequently, the equilibrium properties are not always obtained. Therefore the compressibility calculated in such cases represent only a relative rheological parameter. The value of the compressibility for a certain time-dependent process is essentially unobtainable, because the relative value depends on the relaxation time, which cannot be determined from p/A isotherms. Hence an alternative Introduction Dipalmitoyl phosphatidyl glycerol (DPPG) is a lipid, which is found in many biological systems, in particular, in pulmonary surfactant,1,2 where palmitoyl glycerols represent about 5–10% of the total phospholipid content.A part of them is unsaturated. Contrary to the main lipid in lung surfactants, dipalmitoyl phosphatidyl choline (DPPC), DPPG is ionic and forms charged monolayers on aqueous electrolyte subphases. It is often used in mixtures with DPPC and apoproteins as a model resembling biological systems like the alveolar lining layer in mammalian lungs.3 One of the main characteristics of pulmonary surfactant is its ability to resist strong monolayer deformation. One of the main parameters of monolayer rheology is the relaxation time.4–6 Depending on the time scale of the surface stress and strain this time constant allows a prediction of which rheological behaviour, elastic or viscous, is influencing a given process.In the range of seconds and minutes, it characterises different rearrangement processes within the monolayer structure. This may include rearrangement of the condensed phase textures and the formation of networks. In some cases, exchange of matter between bulk and surface may occur, i.e., a slight desorption of monolayer material7 that would affect the relaxation time. The time scale of seconds is interesting because it coincides with the breathing frequency. For this, it has been investigated earlier, but only for DPPC and cholesterol.8 Two main relaxation times were found for DPPC; one in the order of 100 s, the other one in the range 103–5 � 103 s.These relaxation times are not in the frequency range of breathing. Therefore the main phospholipid of the lung surfactant PhysChemComm, 1999, 11surface rheological characterization of pulmonary layers is preferable.15 Under physiological conditions DPPG is a charged phospholipid. Its ionisation state depends on the electrolyte concentration in the subphase. Therefore, the second aim of this study is to compare the surface dilational behaviour of DPPG monolayers spread on water and on an electrolyte solution of 0.15 mol dm–3 NaCl, which resembles the physiological environment. The other experimental conditions were also chosen close to the real situation in human lungs. The experimental temperatures were 20, 35 and 40 ºC.This is of interest not only in terms of the in vivo circumstances, but also because systematic data for DPPG are not available at physiological temperatures. The reasons for this gap in knowledge are obviously experimental difficulties. The only p/A isotherm for DPPG cited in the Handbook of Monolayers and determined by El Mashak and Tocanne16 looks quite unusual for phospholipids. Recently Grigoriev et al.17 used Brewster angle microscopy to characterise DPPG monolayers on subphases containing different amounts of NaCl. They assumed that after spreading DPPG is completely hydrolysed, and that the dissociation of the resulting acid depends on the electrolyte concentration in the subphase.They claimed that monolayer inhomogeneities appear after spreading even at large mean areas per monolayer molecule and subphase electrolyte concentrations lower than 0.15 mol dm–3 NaCl. Undissociated acid molecules were assumed to form solid-like aggregates in the monolayer while the dissociated part surrounds them. However, some of the experimental area values of the compressed monolayers were found to be much smaller than the cross section of a DPPG molecule. A very recent study of DPPG monolayers has shown that over a large temperature region p/A isotherms exist with a two-phase coexistence region, which provides limiting values of the compressed monolayer in good agreement with the cross-section of a DPPG molecule.18 In this work, we will try to verify these results with drop techniques. The pendant drop technique is well suited for higher temperatures and it can be used for both, measurements of surface pressure/molecular area isotherms (p/A) and surface dilational characteristics.The former is possible by reducing the volume of a pendant drop at the surface of which a monolayer was spread. The method used to investigate monolayer dilational rheology is the transient drop deformation, where the spread monolayer at the drop surface is rapidly compressed or expanded by drop volume jumps.10,19 Usually, such relaxation processes areescribed through a single exponential decay. For inhomogeneous material, a better description has been achieved by means of a Gaussian distribution of relaxing elements.19 When the relaxation is determined by different processes with quite different relaxation times, such as in DPPC monolayers, each of these processes should be modelled by an appropriate decay function.Here, we generalise the ansatz given elsewhere19 by using the sum of two exponential decay functions and extend our model of distributed relaxers to the case of two separate distributions. Another method to characterise the surface rheological behaviour is based on sinusoidal oscillation of the drop volume thus causing harmonic oscillation of the monolayer area and therefore an oscillation of the film pressure. This experiment yields also surface elasticity and viscosity for a definite frequency.20,21 This method was used mainly to characterise protein layers, but it can also be used for monolayer characterisation when the monolayer deformation is sufficiently weak to exclude destruction.An important point for lipid systems is that their monolayers may be over-compressed,1 which means the layers are compressed to area below the minimum molecular area demand. In contrast to the usual collapse, which is indicated by a constant or decreasing film pressure, p, further increases under over-compression leading to extremely low surface tensions. Overcompression is thought to be relevant for respiration, although this relevance is controversially discussed.14,22 Nevertheless many authors try to characterise lipid monolayer properties at very high surface pressures.Monolayer over-compression is not possible by using the pendant drop technique because of monolayer material spreading over the outer wall of the capillary at very high surface pressure.10 Therefore the results presented here are limited to surface pressures < 55 mN m–1. An investigation of over-compressed monolayers will be possible by using a captive bubble surfactometer,23 which will be applied in the future experiments. In this report we present: (a) p/A isotherms for DPPG spread on water and on NaCl containing solution at different temperatures, (b) results of transient drop stress relaxation and harmonic oscillation experiments, (c) a mathematical model to describe the monolayer stress relaxation, and (d) surface dilational rheological properties of DPPG monolayers.The results will be discussed in comparison to those found for DPPC under comparable conditions. The isotherms will be compared with isotherms determined by using a Langmuir trough. It will be shown that the surface rheological properties of DPPG and DPPC monolayers remarkably differ and that a surface rheological characterisation of mixtures of different phospholipids and proteins is required for a better understanding of the surface rheological behaviour of pulmonary surfactant. Materials The dipalmitoyl phosphatidylglycerol was purchased from Sigma and was of 99% purity. It was stored at –20 ºC until use. The DPPG was dissolved in chloroform and methanol (3 : 1), and 2 � 10–4 and 10–3 mol dm–3 spreading solutions were prepared.The chloroform and the methanol were p.a. grade and purchased from Baker (J. T. Baker B.V. Deventer-Holland). The water used was bidistilled. Its surface tension was 72.7 0.1 mN m–1 at 20 ºC. The pH value in the subphase was controlled periodically and was around 6.2. Experimental details Surface tension measurements were carried out by using the pendant drop as a microfilm balance in combination with the ADSA-technique (axisymmetric drop shape analysis).24 A water drop (0.025 cm3), formed at the tip of a PTFE capillary of a Hamilton micro syringe dispenser system was brought into contact with small drops of thespreading solution (0.07–0.5 ml), thus creating the insoluble monolayer. Before spreading on the surface, the drop was compressed from 0.4 cm2 (0.025 cm3) down to 0.14 cm2 to check subphase purity.A surface tension decrease of 0.3 mN m–1 during this compression was tolerated. The drop was kept in a closed chamber of 3.5 cm3 volume, which contained 1 ml water at the bottom.25 The evaporated chloroform was sucked off continuously from the chamber during the first 2 min after spreading. Then the drop was kept in the chamber until water saturation of the atmosphere was guaranteed. The temperature was kept constant to ±0.1 degree. p/A isotherms were determined by changing the drop volume stepwise by 0.25 ml. After each step, the monolayer was kept for 20 s at this particular compression state, and a drop image was captured using a frame grabber and stored for further analysis.The resulting compression speed is 1.11 � 10–2 nm2 molecule–1 min–1. p/A isotherms determined by using the Langmuir trough and the pendant drop technique might differ.7,19,26–28 Therefore it seems to be useful to compare the results of the pendant drop technique with those measured by using a Langmuir film balance. For this aim a computerinterfaced film balance with a Wilhelmy-type surface pressure measuring system was used.29 During the compression the relative compression rate was kept constant at typically 210 � 10–3 s–1, which is equivalent to an initial compression rate of 0.05 to 0.10 nm2 molecule–1 min–1, slowed down to 0.01 nm2 molecule–1 min–1.For transient drop relaxation experiments, a drop volume change of 1 µl was used. Steps less than 1 µl were found to cause too small changes of the film pressure. For steps larger than 3 µl, the parameters became dependent on the step size indicating a non-linear rheological behaviour.30 There are drop vibrations immediately after the pressure jump, which prevent data acquisition for drop shape analysis. Therefore, the starting surface pressure p2 , see eqn. (1) below] was determined by extrapolation of (log t) at t = 0. Usually, stress relaxation experiments were carried out by stepwise monolayer compression up to the region where monolayer collapse was expected and subsequent stepwise dilation.7,23 After each step, the drop volume was kept constant for a certain time. Differences between the surface rheological parameters for compression and expansion were found to be small when the monolayer structure was not damaged during the experiment.In the present study, a long time period of 1 h was used to characterise a wide range of the relaxation spectra, as it was done for DPPC (ref. 10). We found, however, that in such cases, it is quite improbable to determine reasonable rheological parameters for monolayer expansion. This is due to the fact that after a stepwise monolayer compression the monolayer was already kept compressed for a long time. Therefore the experimental protocol was changed and the following investigation was restricted to monolayer compression up to a maximum pressure of 50 mN m–1 only. This also agrees with the conditions chosen by Joos et al.8 Surface dilational moduli result from periodic compression and expansion of a monolayer.These moduli contain both the storage elasticity and the dissipative component, which is affected by the retarded flow. This temporal retardation is determined by the phase angle and results in an intrinsic surface dilational viscosity, if there is no mass-transfer between the bulk and the interface. The complex modulus is given by i sin cos (1) i e q e e e= i = r + e with er and ei being the real and the imaginary part of the complex modulus respectively, and q the phase angle. The dilational elasticity is given by q+ d g ln A d e - = From the phase angle the dilational viscosity h can be calculated31 from (2) esin q = h w where w = 2pf is the circular frequency of the area oscillation.The physical meaning of the dilational viscosity is complex. It contains any relaxation process in the surface layer and the range of frequency used, as well as the intrinsic viscosity.32 There have been only few attempts to separate the intrinsic viscosity from the entire dilational viscosity.33,34 The present experiments dallow such separation because the frequency range accessible is restricted and an effective dilational viscosity results. The rheological parameters determined from harmonic oscillations may depend on the amplitude. In cases of high amplitudes the surface layer may be destroyed, which results in a breakdown of the rheological parameters.For adsorbed b-lactoglobulin layers at different interfaces it was found that amplitudes of drop volume changes of ml are small enough to guarantee independent elasticities and viscosities.21 In the present study all measurements were restricted to ±0.5 l, which corresponds to changes of the drop surface by ±2.5–5 %. Drop volume changes were realised by using the Hamilton syringe pump PD/2 and a 25 ml syringe. This syringe pump was step motor controlled and the minimal volume per step was 0.025 ml. The highest speed of image acquisition was 1 image per second. By using these values different cycles of sinusoidal drop oscillation were created in the range of 0.025 and 0.00625 Hz.Before starting oscillation experiments the drop volume was stepwise compressed to a definite surface coverage keeping the drop volumes constant (for 60 s) after each step. Usually 10 repeated oscillations (1 cycle) at 5 constant frequencies were combined to a program of experiments. The drops were kept 10 min at constant volume between the cycles. (3) 1) does not change Model of stress relaxation A transient relaxation experiment starts from an equilibrium state. Equilibrium is assumed to be adapted, when the surface pressure (pnormal distribution, where the negative wing (1/ t< 0) is neglected. 1 1 æ 1 B exp significantly over a certain time interval. After a rapid drop volume change by means of surface compression or expansion from A1 to A2 at time t = 0, the pressure jumps to a value p2.The corresponding dilational elasticity is defined as - 1 1 æçççç2 1 s A A (4) ; A , e 2 1 2 1 2 g(t) º The temporal behaviour is described by the stress relaxation function (t). For systems containing only one relaxation process, the single linear relaxer model consisting of a spring and a dashpot, reads35 t Y( ) D DA A == =e 1 -÷ (5) æ -t ö r ø t1 + B e B 2 æç ç( ) 1- B è 2 1 1 2 2 2 s 2 s 1 2 B ) æç çè d - 1 è è æç ç- è 1 s æ 1 ö÷ ÷ø 1d ö = ÷ø 1 ( = B ) - Y 1 1 è yçè ) (t Y with 1 D= expç = In practice, due to model approximations and experimental uncertainties, B1 + B2 will not exactly give 1.For real systems certain inhomogeneities need to be taken into account by means of a set of relaxers acting in superposition and having more or less different relaxation times. It turns out that for real systems it is favourable to express this distribution in terms of the inverse relaxation time, ¥, 1 æç çè , 11 t - ç ç2ç çè è Here, t1 and t2 are the most probable relaxation times, while 1/ t1s and 1/ t2s measure the widths of the distributions and, thus, characterise the (in)homogeneity of the structure under investigation. For vanishing widths of the distributions, 1/ t1(2)s®¥, it reduces to model eqn. (6) for homogeneous material. It should be noted that the width of the distribution (the material inhomogeneity) does not result from the scattering of the experimental data.Of course, real experimental data do not exactly obey an exponential decay law, so that the introduction of a distribution should improve the fit anyway. For DPPG, however, we did not find a correlation between the magnitudes of 1/ t 1/t2s and the standard deviation of the fit curve. This means that our numerical results for t1s and t2s are physically relevant. For the case of two separate distributed relaxation processes, model eqn. (7) becomes æ 1 e 2 0 0 ç t y ÷ öø 1 -t ¥ ¥ -t ò ò 1 2 1 i , 1 12 2 i i ¥ Dt2 - r ttd ; 0 g g(t) g(0) 0 = Y(0) Y º -Y¥ 0 è where g(t) is the stress (surface tension) at time t and g (0) denotes its value at t=0.¥ ¥ ò- r 0 0 º g For systems containing two distinct relaxation processes of relative amounts B1 and B2 with relaxation times t1 and 2, eqn. (5) can be generalised to (6) (7) tY(t) t) ( Y Inserting eqn. (8) into eqn. (9) yields the model function ö÷ ÷øæ ö ç ÷ è tø ( ) Y -Y e 1 - = t =¥ ti ö÷ ÷y ø 1 æ t yç çè t; t, model max y =t(t; Y; ( X t Ymodel = æ 1 ö ç + ÷ è ø t Here, we have added the final stress, y¥ º t(t = ¥) = g(t = ¥)/ g(t = 0). Y¥ is a material parameter depending on the prehistory of the monolayer (such as compression or dilation, deformation speed, surface coverage). It may also account for a residual non-equilibrium state of the monolayer.Similar to B1+B2 �1, the numerical fit of the model curve, y(t), to the experimental stress data, gexp(t), can result in y0 � 1 and y¥ � gexp(t = ¥)/ gexp(t = 0). At the same time, this flexibility of the model enables one to automate the calculation of fit parameters via a reasonable initial guess from the experimental data. y(1/ t) describes the distribution of relaxation times. It is written in terms of the inverse relaxation time, because this turns it into the inverse Laplace transform of y(t). In practice, however, due to the limited number of data and their scattering, the numerical inversion of the experimental data, y(t), is difficult. Thus, it is favourable to assume some functional form for y(1/ t) and to fit the containing model parameters.36 In fact, one can assume that the inverse relaxation times are Poisson distributed.Due to the large number of relaxers, the Poisson distribution can be approximated through the Gauss Y ( ) ( )å The function 2 2 2 t t ) , 1 erf í × 12 , ¥ t max = e ) t ® 1 t Y ( ) - éê+ êë describes the deviation from the exponential decay caused by the existence of different relaxation times. The faster relaxers ( t1) accelerate the initial relaxation (t< t1), while 2) slow down the long-time the slower relaxers ( t relaxation (t > t2). For very long times, the latter is no longer exponential, but proportional to 1/t 12 p In practice, however, the fluctuations in the experimental data prevent the observation of this asymptotic.1 e ìïæ t ç çï 2 è î2 2 2max t -t × tt 2 2 1 1 1 - 2 1 1 exp - ö ö ÷ ÷ ÷ø2 ÷÷÷ æçççç +1 ( ¥ çè ö÷ ÷ø ö÷ ÷øø (9) ÷ ÷ø (8) 1s, + Y ç çèe æ ö ç ÷ t ö ö ÷ ÷ ÷÷ ø ÷÷ ÷ ÷ø æ 1 ö + ÷ tø , 1 æçt y ÷ öø t t -t / ti t 1 Y æ t yç çè e B i t X × maxs ; t, t ( )+ (10a) t (10b) - I ; t Y¥ ü öïù ÷ ÷ú ý ø û ï�ú t >> t + (11)For practical computation, the rational approximations for the error function are most favourable, since they cancel the exponential factor exp{t2/2 t 2}.37 i s In order to quantify the (in)homogeneity of the stress relaxation process, the dimensionless homogeneity parameter 1 ti (12) i Q 1 i s t = has been introduced.19 It is an analogue to the quality parameter of microwave resonators and other resonance circuits.The larger its value is, the more homogeneous the relaxation behavior is; Q = ¥ corresponds to a pure exponential decay. When Q becomes too small, Q » 1, the distribution becomes wide and the mos its validity. The parameters of our model are the maximum probable relaxation times ti , the distribution widths tis, the homogeneity parameters Q, and the relative amounts of relaxer groups, Bi. From the first part of the experimental decay, a modulus can be determined, which measures the initial elastic deformation (t < t1 < t2).The unambiguous determination of dilational viscosities needs additional assumptions about the relationship between elasticity and relaxation time, however. The physical interpretation of the rheological behaviour may be associated to different interfacial models. The model by Joos et al.8 assuming a transfer between different monolayer states leads to similar formulations. Fig. 1 p/A isotherms of spread DPPG monolayers on water determined by using the Langmuir film balance (lines) and the pendant drop technique (symbols) at different temperatures. Experimental results (a)p/A isotherms Fig. 1 shows the p/A isotherms for DPPG spread on water which were recorded by using different measuring devices. The deviation between the p/A isotherms at 20 °C obtained from the Langmuir trough and the drop microfilm balance are negligible.35 ºC, 40 ºC. Measurements by using the 20 ºC, microfilm balance were carried out with 60 s between every step of volume reduction. At 35 °C there are however remarkable differences. In contrast to the isotherm recorded by using the Langmuir film balance, there is no pronounced plateau in the data obtained from the pendant drop experiment. Such discrepancies are common for p/A isotherms obtained from Langmuir and microfilm balance, respectively.26–28 Usually these deviations are explained by differences in the spreading conditions; most of all higher concentrations of the spreading solutions are used for the pendant drop.This however is not the case for the present experiments. It seems that more systematic investigations are necessary to explain these differences. This will be done in a future work only, and for the time being we will compare p/A isotherms determined under comparable conditions. All isotherms show a comparable shape and fit well into the temperature dependence of other phospholipids.38 Furthermore, all isotherms approach an equal minimum molecular area demand. In this sense the isotherm at 40 °C spread on water confirms the results reported by Grigoriev et al.17 In contrast, however, the isotherms obtained at 20 °C differ remarkably from those given in ref. 17. The shape of the p/A isotherm shows that the pure fluid (gaseous) phase of the DPPG monolayer does not exist at T = 20 ºC.Different to the DPPC monolayers, at this temperature the DPPG monolayer is in the state of the two-phase coexistence already at the high area values under the conditions of spreading (A ca. 1 nm2 molecule–1) at a pressure of about zero.29 That means 2D condensed phase aggregates are formed after evaporation of the spreading solvent. The situation is different in the DPPG monolayers at the higher temperatures. Similar to the Langmuir monolayers investigated in the film balance, the p/A isotherms show an inclined plateau region after the main phase transition point Ac. At high area values for A > Ac, the DPPG monolayers are in the liquid (gaseous) state. The condensed phase domains are formed and grow only in the two-phase coexistence region (A îAc).Fig. 2 compares the p/A isotherms at 35 and 40 ºC in absence and presence of 0.15 mol dm–3 NaCl determined by the microfilm balance. Fig. 2 p/A isotherms of DPPG monolayers spread on water ( ) and 0.15 mol dm–3 NaCl. (D). Measurements by using the 35 °C, and 40 °C. pendant drop technique. There are some differences caused by the presence of electrolyte in the subphase. Obviously the minimum area demand is increased by about 0.05 nm2 molecule–1. Furthermore, the surface pressure in general is increased over a wide range of surface coverage. It levels off however already at lower surface pressure. Such levelling off indicates the formation of collapsed and/or folded structures at the interface.23(b) Transient drop relaxation experiments Fig.3 shows some examples monitored at a relatively high surface coverage up to a surface coverage. Fig. 3 Surface stress recovery of DPPG monolayers after a compression step (1 ml drop volume reduction) at 35 ºC, spread on water at A = 0.45 nm2 molecule–1, spread on a 0.15 mol dm–3 NaCl solution at A = 0.70 nm2 molecule–1, and spread on a 0.15 mol dm–3 NaCl solution at A = 0.45 nm2 molecule–1. The first two examples ( and ) at 0.45 and 0.47 nm2 molecule–1 show similarities, but also remarkable differences between the two subphases. Similar are the time independent surface pressures measured after about 103 s. The constant values, which are reached after the surface stress recovery, are in good agreement with those of the p/A isotherms at comparable surface coverage, as shown in Fig.2. The absolute surface pressure decay for the two subphases is different. It is clearly smaller for the monolayer spread on water. This must result in a higher surface elasticity. Furthermore from the recovery process it can be concluded that the shape of the p/A isotherms should depend on the compression rate at least for a stepwise monolayer compression regime. In the monolayer coverage range < 0.45 nm2 molecule–1 ( ) the monolayers became brittle comparable to DPPC monolayers as reported in ref. 10. It was shown in ref. 23 that at high surface pressure even some visible collapsed crystalline parts of the phospholipid monolayer might be separated from the surface.(c) Model fitting to stress relaxation data We applied 3 different models to fit the stress relaxation of the DPPG monolayers, i.e. a single and a double exponential fit, and the model with 2 different main relaxation times, which are distributed thus taking into account the inhomogeneity of the two relaxation processes. Fig. 4 compares the experimental p(t) data with these three different models. The surface pressure p was normalised such that p(0) = 1 and p(¥) = 0. The single exponential decay curve 1 clearly fails to reasonably approximate the experimental data. The double exponential decay curve 2 represents a distinct improvement, in particular, for the initial stage, but is still not satisfying at medium time range (t » 400–1200 s).The model with distributed relaxing elements curve 3 yields an excellent fit for the entire time interval. To determine the relaxation times Fig. 5 depicts the distribution function y(1/ t) of curve 3 in Fig. 4. There are two separate narrow distributions. The first distribution with a maximum at t1 » 15 s has a width of only 0.5 s. The second one is found at t2 = 475 s, its width is about 20 s, much larger than that of the first distribution. Fig. 4 Normalised surface pressure over time for DPPG spread on water at 40 °C (A = 0.61 nm2 molecule–1) using three different approaches. Experimental points: 1 – fit to single exponential decay; 2 – fit to double exponential decay; 3 – fit to model eqn.(9). Fig. 5 Distribution function of eqn. (8) for curve 3 in Fig. 4. Further investigations revealed small distribution widths, too, with Q ranging between 5 and 500, where, typically, Q1 > Q2. The higher the Q-values, the more justified is the application of a double exponential decay, so that the introduction of distributions often yields no significant improvement. Because of large data scattering of Q(A) the parameter Q was used only to check the quality of the approximation. Unfortunately, Q does not exhibit a systematic dependence on A. The reason is, perhaps, the fact that it is evaluated as the quotient of two fit parameters with different fluctuations. Note, that the accuracy of the Q value is the lower, the higher its value, because the numerical computation of ts becomes less accurate.There were no cases of Q < 1, where the assumption of a Poisson distribution is invalid. Thus, the Q values obtained support the description of the stress recovery process by a model containing two different relaxation times/mechanisms. (d) Surface dilational rheological properties of spread DPPG monolayers The following results on the rheological parameters were obtained by stepwise transient monolayer compression, as demonstrated in Fig. 3, and keeping it constant for 1 h at a definite state of compression. During a time of 1 h the drop volume changes usually by about 1–2% due to evaporation, which was neglected.Figs. 6 and 7 summarise our findings concerning the surface dilational behaviour of DPPG monolayers.The model parameters extracted from these experiments are the dilational modulus, the two different relaxation times, and the amount B1 of relaxers with a relaxation time close to t1. The stress relaxation behaviour of DPPG monolayers is governed by two relaxation times that differ more than one order of magnitude. In general, these relaxation times are much smaller than that of DPPC.8,10 Fig. 6 Stress relaxation parameters of DPPG spread on water ( ) and on 0.15 mol dm–3 NaCl ( ) at 35 ºC. Fig. 7 Stress relaxation parameters of DPPG spread on water ( ) and on 0.15 mol dm–3 NaCl ( ) at 40 ºC. The relaxation time t1 is of the order of 20 s for the whole range of surface coverage, i.e.it is little affected by temperature and the presence of electrolyte. The relaxation time t2 decreases slightly at higher surface coverage. The scattering of the experimental data seems to be larger at 35 ºC than at 40 ºC. The dilation modulus e increases up to 0.45 nm2 molecule–1 at 35 ºC. At 40 ºC the modulus exhibits a maximum at 0.55 nm2 molecule–1. The highest elasticity values were determined for DPPG spread on water at 35 ºC. Both, the increasing temperature and the presence of electrolyte decrease the elasticity, where the effect of electrolyte becomes negligible at 40 ºC. The amount of fast relaxation processes, B1, is definitely higher at higher temperatures and in the presence of electrolyte. Correspondingly, the observed relaxation speeds up with increasing temperature and electrolyte presence.(e) Surface rheological behaviour characterised by harmonic oscillations of the drop surface Unfortunately our present experimental equipment is restricted to a small frequency interval. Therefore the experiments do not cover the whole range of relaxation times found by the stress relaxation experiments and do not really reach the frequency of the respiration process of mammalian lung. Nevertheless these experiments give a more detailed insight into the surface rheological behaviour of lipid monolayers and may show if the two experimental methods lead to comparable results. The investigations were carried out at 3 different surface coverages for both monolayers spread on water and on 0.15 mol dm–3 NaCl.To improve the comparability of the surface rheological parameters the surface coverage chosen for the NaCl containing solution was increased by 0.05 nm2 molecule–1 in comparison to that of the water subphase. p usually remained constant when the starting surface coverage was established by sufficiently slow compression rate. The film pressure therefore oscillates around a constant p. Only in the range of the minimum area demand the oscillation was superimposed by a decrease of the surface pressure, which is demonstrated in Fig. 8. Fig. 8 Oscillating film pressure of a DPPG monolayer spread on water. T = 35 ºC, A = 0.40 nm2 molecule–1, f = 0.00625 Hz. It was assumed that the film pressure decay is caused by a partial film collapse or folding.23 To avoid monolayer destruction the measurements were restricted to the range 0.8 to 0.45 nm2 molecule–1, where usually no superimposed changes of p were observed.Fig. 9 gives an example of two drops recorded at the minimum and at the maximum of an oscillation.It is clearly seen that there is a well-pronounced change of the drop profile, although the changes of the drop volume DV are small in comparison to V. Fig. 9 Drop recorded at the minimum and at the maximum of an oscillation (DV = ±0.5 ml). DPPG monolayer spread on 0.15 mol dm–3 NaCl , T = 40 ºC. (a) A = 0.4475 nm2 molecule–1; (b) A = 0.4896 nm2 molecule–1. Fig. 10 shows an example of one oscillation of the surface tension and the drop surface at 0.61 nm2 molecule–1 and 40 ºC.The surface coverage in Fig. 10 was much lower than that shown before (Fig. 8), nevertheless, there is also a surprisingly large amplitude of p. Note, for comparison, the interfacial tension changes by about 0.2 to 1 mN m–1 for adsorbed b-lactoglobulin layers at different interfaces and a drop volume amplitude of ±1 ml and ±1.5 ml,21 i.e. much higher amplitudes of drop volume oscillation, whereas the changes of the drop surface were comparable. Furthermore, Fig. 10 shows a small but distinct phase angle between the oscillation of the drop surface and the surface tension. Fig. 10 Oscillating surface tension g and surface area of a DPPG monolayer spread on water. T = 40 ºC, A = 0.61 nm2 molecule–1, surface area, surface tension, and f = 0.0167 Hz.approximated harmonic oscillation. The following 3 figures show the dilational elasticity and viscosity as a function of frequency for monolayers spread on water, and on 0.15 mol dm–3 NaCl. Fig. 11 shows the determined elasticity. All moduli display a slight increase with frequency. At surface coverages of 0.8 and 0.6 nm2 molecule–1 and at low frequency the elasticity values are in good agreement with those found by the stress relaxation experiments. The moduli increase with increasing surface coverage in general, and show very high values in the range of minimum area demand per molecule. A maximum as shown for 40 ºC in Fig. 7 was not observed. This maximum is probably caused by the different experimental regimes.The presence of NaCl obviously decreases the modulus. It is slightly higher at 35 ºC and at high surface coverage. Fig. 12 collects the results of dilational viscosity. At moderate surface coverage the viscosities show a well pronounced dependence on frequency with a minimum between the two relaxation times. The dependence on temperature is not significant, but the viscosities are clearly decreased by the presence of NaCl in the subphase. Fig. 11 Surface dilational elasticity of a spread DPPG monolayer depending on the frequency of harmonic oscillation. Red and yellow symbols at 40 ºC, green and blue at 35 ºC. Red and blue symbols spread on water, yellow and green spread on 0.15 mol dm–3NaCl.0.8 nm2 molecule, 0.42 nm2 molecule–1, 0.6 nm2 0.47 nm2 molecule–1. molecule, Fig. 12 Surface dilational viscosity of a spread DPPG monolayer depending on the frequency of harmonic oscillation. Red and yellow symbols at 40 ºC, green and blue at 35 ºC. Red and blue symbols spread on water, yellow and green spread on 0.15 mol dm–3 NaCl. 0.8 nm2 molecule–1, 0.6 nm2 molecule–1.Fig. 13 collects the results near the minimum area demand per molecule. Although it is complicated to get reliable results in this range of the isotherm, there is a strong increase of viscosity at low frequencies for all systems investigated with maximum values for monolayers spread on water at 40 ºC. Nevertheless, when the structure is stressed more rapidly (up to 0.018 Hz) the viscosity decreases dramatically.Fig. 13 Surface dilational viscosity of a spread DPPG monolayer depending on the frequency of harmonic oscillation. Red symbols at 40 ºC, yellow at 35 ºC. monolayer spread on water at 0.42 nm2 molecule–1, monolayer spread on 0.15 mol dm–3 NaCl at 0.47 nm2 molecule–1. The dependency of the surface rheological parameters shown in Fig. 11 confirms a slight increase of elasticity with rising frequency. It is clearly seen that the elasticity strongly increases with increasing surface coverage. The surface viscosity also shows a marked dependency on frequency. There is a minimum in the range between the relaxation times found by stress relaxation experiments. It is slightly pronounced in the surface coverage range 0.6– 0.8 nm2 molecule–1.The viscosity however strongly increases for a surface coverage of 0.42 nm2 molecule–1, where extremely high dilational viscosities are observed at low frequencies, i.e. when the oscillation of the drop surface is slow and the process becomes strongly influenced by the higher relaxation time. Discussion p/A isotherms of spread DPPG monolayers were presented for different temperatures, with and without electrolyte. The results determined by using a Langmuir trough and the drop microfilm balance agree in tendency, but show also some differences. The agreement was acceptable for the isotherms at 20 ºC. Several authors report differences between the isotherms found by these two methods, which is not satisfactory and needs further investigation.The p/A isotherms determined by using the pendant drop microfilm balance lead to reasonable minimum values of the area per molecule of about 0.4 nm2 molecule–1. This is in agreement with the cross-section area of phospholipids. The p/A isotherms fit well into the isotherms known for other phospholipids. In contrast to DPPC, DPPG is a negatively charged lipid. As a consequence, the surface behaviour of DPPG is influenced by the presence of an electrolyte. The p/A isotherms are shifted to higher film pressures in the presence of NaCl. It has been discussed17 that the pure DPPG monolayer is hydrolysed into the respective acid (DPPGH) immediately after spreading on pure aqueous subphase.Very recent thermodynamic studies of DPPG monolayers have shown that a possible dissociation effect on the monolayer properties of the ionic DPPG, which is spread on a pure water surface, can be largely ignored.18 On the other hand, the presence of electrolyte increases obviously the degree of dissociation. This may be understood by pushing out protons of the monolayer and by increasing the number of Na+ ions in the diffuse surface layer when NaCl is present in the aqueous subphase. The driving force for the increase in minimum area demand and a surface pressure increase at lower surface coverage is the electrostatic repulsion between the dissociated DPPG molecules in the monolayer. The results reported are compatible with the findings in refs.3 and 39. The peculiarities of DPPG are the reason for a more fluidlike behaviour, when compared with DPPC, which is obviously reflected by the formation of less brittle monolayers allowing to characterise the stress relaxation behaviour even in the range of higher surface pressure. The surface stress relaxation behaviour of spread DPPG monolayers has been characterised in the same time scale as it was done previously for DPPC.10 The relevant surface dilational rheological parameters are derived from a new model, which unifies the idea of using an individual (exponential) decay function for each relaxation process and the introduction of distributed relaxing elements (relaxation times) accounting for material inhomogeneity.The fit of experimental data reveals two distinct relaxation processes, which are connected probably with the formation and rearrangement of monolayer domains. The main relaxation times differ by more than one order of magnitude. The distribution of relaxation times is quite narrow for both groups of processes. The interpretation of experimental stress relaxation curves in terms of a continuous distribution of relaxation times yields both a quantitatively and qualitatively better description of the stress relaxation behaviour and, moreover, a characterisation of the homogeneity of the structure. Each relaxation mechanism needs its own distribution; otherwise, the model simulates large inhomogeneities. The model is applicable for both, the p(t) decay (monolayer compression) and p(t) increase (monolayer dilation) using (t) instead of the stress relaxation function (t).It was used here only for the case of monolayer compression, because of the influence of the monolayer prehistory that also changes the surface rheological behaviour. This however is no restriction of universality, because differences between the surface dilational parameters for monolayer compression and dilation are usually small30,40 when the monolayer was not destroyed. In general, the relaxation times of DPPG are smaller than those of DPPC determined under similar conditions. This means that DPPC monolayers exhibit mainly elastic behaviour in the range of the breathing frequency.14 In contrast DPPG monolayers probably may show viscoelastic behaviour already at usual breathing frequency, as the lower mean relaxation time is in the range of 15–20 s.The differences in the surface dilational behaviour between DPPG and DPPC are so strong that the present results do not allow conclusions for the surface rheological behaviour of these two phospholipids in mixtures.The influences on the dilational behaviour exerted by the stress relaxation of both temperature and presence of NaCl conflict. For a better understanding of surface dilational rheology and the possible function of DPPG in pulmonary surfactants additionally harmonic oscillation experiments were carried out and the dependence of the surface dilational behaviour on the frequency was characterised.Unfortunately the frequency range accessible is restricted by experimental conditions and covers only the time scale between the two relaxation times, t1 and t2. It is important however that both the surface dilational elasticities determined by stress relaxation and harmonic oscillation agree in the range of surface coverage 0.8–0.6 nm2 molecule–1 and at low frequencies. In the range up to 0.42 nm2 molecule–1, or 0.45 nm2 molecule–1, respectively, in the case of NaCl, the elasticity strongly increases, which is in good agreement with the strong increase of the film pressure in this part of the isotherm. A small oscillation of the surface area yields a strong oscillation of the surface pressure.Nonetheless, a destruction of the monolayer by the amplitude chosen can obviously be avoided, if the maximum of the film pressure does not exceed 50 mN m–1, and the surface tension oscillates around a definite constant value. However, a maximum of surface dilational elasticity as found at 40 °C by stress relaxation was not confirmed. Probably the compression steps of 1 ml cause some monolayer defects, thus leading to a decrease of surface dilational elasticity in the range 0.6–0.45 nm2 molecule–1. A new information resulting from the oscillation experiments is a pronounced dependency of the surface dilational elasticity on the frequency of oscillation. e slightly increases with frequency. The main influence of NaCl on the elasticity reflexes in all cases a decrease. There is also a well-pronounced dependency of the dilational viscosity on frequency and on the presence of electrolyte.h becomes minimal at a frequency between the main relaxation times, t1 and t2, and increases for frequencies close to the relaxation times. In the range of the liquid condensed phase the dilational viscosity strongly rises up for slow oscillations by the influence of the higher main relaxation time t2. This dependence, however, also clearly shows that the viscosity strongly decreases, as soon as the disturbance of the monolayer is rapidly executed. In such cases the overall surface dilational behaviour becomes more elastic. Therefore the results do not really confirm a more fluid-like behaviour in the presence of NaCl or a higher compressibility.Both elasticity and viscosity decrease by the influence of electrolyte and the rheological behaviour is much more complex. This is in good agreement with the increase of the minimum area demand, which prevents the formation of a mechanically stable network at the interface by electrostatic repulsion between the head groups. This is supported by the starting collapse or formation of folded structures at lower surface pressure. Usually 10 subsequent oscillations were carried out at each frequency, but no dependence of viscosity on the time of oscillation has been found. Therefore the rheological behaviour of the DPPG layers is structurally rather viscous than thixotropic. Nevertheless, there is also an increase of the viscosity at higher frequencies when getting closer to t1.Therefore it cannot definitely be excluded that the rheological character of the film in that frequency range is mainly elastic. It should be concluded that a better understanding of the behaviour of different pulmonary surfactant components via surface rheological measurements requires both the relaxation times and the dependence of the rheological parameters on frequency, especially in the frequency range of breathing. Acknowledgements This work was financially supported by projects of the European Community (INCO ERB-IC15-CT96-0809), the Deutsche Forschungsgemeinschaft (Mi 418/9, Mi 418/7, Pi 165/7, Kn 593/2, Wü 187/6, Wü 187/8) and the Ministerium für Wissenschaft, Forschung und Kultur, Land Brandenburg.Many thanks to Klaus Dannenberg from Virchow-Klinikum, Forschungswerkstatt, for his kind help in building the experimental set-ups. References 1 K. M. W. Keough, in Pulmonary surfactants, ed. B. Robertson, L. M. G. van Golde and J. J. Batenburg, Elsevier, Amsterdam, 1992, p.109. 2 S. Koppenol, F. H. C. Tsao, H. Yu, and G. Zografi, Biochim. Biophys. Acta – Biomembranes, 1998, 1369, 221. 3 J. N. Israelachvili, Intermolecular and surface forces, Academic Press, London, 1985. 4 F. V. Vader, T. F. Erkens, M. v.d. Tempel, Trans. Faraday Soc., 1964, 60, 1170. 5 M. Joly, Surface and Colloid Sci., ed. E. Matijevic, Wiley, New York, 1972, vol. 5. 6 R. Miller, R.Wüstneck, J. Krägel and G. Kretzschmar, Colloids Surf. A, 1996, 111, 75. 7 R. Wüstneck, S. Siegel, Th. Ebisch and R. Miller, J. Colloid Interface Sci., 1998, 203, 83. 8 P. Joos, M. van Uffelen and G. Serrien, J. Colloid Interface Sci., 1992, 152, 521. 9 S. Koppenol, H. Yu and G. Zografi , J. Colloid Interface Sci., 1997, 189, 158. 10 R. Wüstneck, N. Wüstneck, D. O. Grigoriev, U. Pison and R. Miller, Colloids Surf., in the press. 11 J. C. Smith and D. Stamenovic, J. Appl. Physiol., 1986, 60, 1341. 12 S. Schürch, D. Schürch, T. Curstedt and B. Robertson, J. Appl. Physiol., 1994, 77, 974. 13 R. Herold, H. Bünger and U. Pison, Colloids Surf. A, 1996, 114, 211. 14 S. Schürch, F. H. Y. Green and H. Bachofen, Biochim. Biophys. Acta, 1998, 1408, 180. 15 M. Ueno, K. Fukuda, N. Takeo, Y. Tanaka and F. Yuji, Tetsuro Nippon Kaimen Igakkai Zasshi, 1988, 19, 116. 16 E. M. El Mashak and J. F. Tocanne, Biochim. Biophys. Acta, 1980, 596, 165. 17 D. Grigoriev, R. Krustev, R. Miller and U. Pison, J. Phys. Chem., 1999, 103, 1013. 18 D. Vollhardt, V. Fainerman and S. Siegel, J. Phys. Chem., 1999, 103, in the press. 19 R. Wüstneck, P. Enders, Th. Ebisch and R. Miller, Thin Solid Films, 1997, 298, 39. 20 J. Benjamins, A. Cagna and E. H. Lucassen-Reynders, Colloids Surfaces, 1996, 114, 245. 21 R. Wüstneck, B. Moser and G. Muschiolik, Colloids Surf. B, in the press. 22 J. Goerke, Biochim. Biophys. Acta, 1998, 1408, 79.23 N. Wüstneck, R. Wüstneck, V. B. Fainerman, U. Pison and R. Miller Colloids Surf. B, in the press. 24 D. Y. Kwok, D. Vollhardt, R. Miller, D. Li and A. W. Neumann, Colloids Surf. 1994, 878, 51. 25 R. Wüstneck, N. Wüstneck, D. Vollhardt, R. Miller and U. Pison, Mater. Sci. Eng., C, in the press. 26 D. Y. Kwok, B. Tadros, H. Deol, D. Vollhardt, R. Miller, M. A. Cabrerizo-Vilchez and A. W. Neumann, Langmuir, 1996, 12, 1851. 27 D. Y. Kwok, D. Vollhardt, R. Miller, D. Li and A. W. Neumann, Colloids Surf. A, 1994, 88, 51. 28 A. Jyoti, R. M. Prokop, J. Li, D. Vollhardt, D. Y. Kwok, R. Miller, H. Möhwald and A. W. Neumann, Colloids Surf., 1996, 116, 173. 29 D. Vollhardt, Adv. Colloid Interface Sci., 1996, 64, 143. 30 R. Wüstneck, P. Enders, Th. Ebisch, R. Miller and S. Siegel, J. Colloid Interface Sci., 1998, 206, 33. 31 J. Lucassen and M. van den Tempel, Chem. Eng. Sci., 1972, 27, 1283. 32 S. S. Dukhin, G. Kretzschmar and R. Miller, in Dynamic Adsorption at the liquid Interface, ed. D. Möbius and R. Miller, Elsevier, Amsterdam, 1995, vol.1, ch. 3, 6. 33 C. I. Christov, L. Ting and D. T. Wasan, J. Colloid Interface Sci., 1982, 85, 363. 34 C. M. Maldarelli and K. J. Stebe, J. Colloid Interface Sci., 1994, 163, 177. 35 B. Biswas and D. A. Haydon, Proc. R. Soc. London, Ser. A, 1963, 278, 317. 36 W. Rabinovitch, R. F. Robertson and S. G. Mason, Can. J. Chem., 1960, 38, 1881. 37 N. Abramowitz and I. Stegun in Handbook of Mathematical Functions, NBS, Washington, 1964, no. 7.1.25.. 38 D. Chapman, in Biomembranes, Physical Aspects, ed. M. Shinitzky, Balaban Publishers,VCH, Weinheim, New York Basel Cambridge, 1993. 39 C. A. Helm, L. Laxhuber, M. Lösche and H. Möhwald, Colloid Polym. Sci., 1986, 264, 46. 40 R. Wüstneck, J. Reiche and S. Förster, Thin Solid Films, 1997, 307, 100. Paper 9/06516D PhysChemComm ©The Royal Society of Chemistry 1999

ISSN:1460-2733    

DOI:10.1039/a906516d

出版商:RSC    

年代:1999

数据来源: RSC