4004
Langmuir 2004, 20, 4004-4010
Dynamic Adsorption and Surface Tension of Aqueous Dilauroylphosphatidylcholine Dispersions under Physiological Conditions Tze-Lee Phang, Ying-Chih Liao, and Elias I. Franses* School of Chemical Engineering, Purdue University, 480 Stadium Mall Drive, West Lafayette, Indiana 47907-2100 Received August 5, 2003. In Final Form: February 18, 2004 The dynamic surface tension and equilibrium adsorption behavior of DLPC dispersions in phosphate buffer saline at 37 and 25 °C was studied with tensiometry, infrared reflection-absorption spectroscopy (IRRAS), and ellipsometry. The results are compared with those in water (Pinazo et al. Langmuir 2002, 18, 8888). Even though the pH and salinity have no apparent effect on the equilibrium surface tension and the surface pressure-area isotherm, they affect the dynamic surface tension by decreasing the adsorption rate and increasing the dynamic tension minima at a pulsating area of 20 or 80 cycles per minute. Moreover, IRRAS and ellipsometry results show that the adsorbed layers and the spread monolayers have larger area per molecule, or looser packing, in buffer than in water. A new hypothesis is proposed to elucidate the effect of pH/salinity on this zwitterionic surfactant: there is some specific interaction or binding between the ions from the buffer saline with the polar headgroups of DLPC. This interaction induces stronger intermolecular repulsions of the surfactant layer in buffer than that in water, despite the expected electrostatic screening effect, and causes higher dynamic surface tensions. The results have implications in designing lung surfactant replacement formulations.
1. Introduction The adsorption of aqueous solutions or dispersions of surfactants and lipids at the air/water interface is important in food processing,1,2 biological membranes,3 and lung surfactants.4-6 Dipalmitoylphosphatidylcholine (DPPC), a lecithin, is a key ingredient of lung surfactant, which is a lipid/protein mixture which stabilizes the lungs against collapse. Dilauroylphosphatidylcholine (DLPC) is a homologue of DPPC with C12 hydrocarbon chains. It has a chain melting transition temperature (Tc) of ca. 5 °C. It forms fluid liposomes and vesicles at 37 or 25 °C. Moreover, it has low toxicity, as evidenced by successful tests of asthma drugs on humans.7-9 We have reported the surface tension behavior of DLPC in water, mostly at 25 °C, and compared it to that of DPPC.10,11 DLPC in water shows very low dynamic surface tensions under pulsating area conditions. Its properties result from its molecular structure, low but * To whom correspondence should be addressed. Tel: (765) 4944078. Fax: (765) 494-0805. (1) Darling, D. F.; Birkett, R. J. In Food Emulsions and Foams; Dickinson, E., Ed.; Royal Society of Chemistry, Burlington House: London, 1986; Special Edition No. 58, p 1. (2) Dickinson, E.; Woskett, C. M. In Food Colloids; Bee, R. D., Richmond, P., Mingins J., Eds.; Royal Society of Chemistry, Thomas Graham House: Cambridge, 1989; Special Edition No. 75, p 74. (3) Tanford, C. The Hydrophobic Effect: Formation of Micelles and Biological Membranes; Wiley-Interscience: New York, 1980. (4) Clements, J. A.; Hustead, R. F.; Johnson, R. P.; Gribetz, I. J. Appl. Physiol. 1961, 16, 444. (5) Notter, R. H.; Finkelstein, J. N. J. Appl. Physiol. 1984, 57, 1613. (6) Notter, R. H. Lung Surfactants; Marcel Dekker: New York, 2000. (7) Vidgren, M.; Waldrep, J. C.; Arppe, J.; Black, M.; Rodarte, J. A.; Cole, W.; Knight, V. Int. J. Pharm. 1995, 115, 209. (8) Waldrep, J. C.; Gilbert, B. E.; Knight, C. M.; Black, M. B.; Scherer, P. W.; Knight, V.; Eschenbacher, W. Chest. 1997, 111, 316. (9) Saari, M.; Vidgren, M. T.; Koskinen, M. O.; Turjanmaa, M. H.; Nieminen, M. M. Int. J. Pharm. 1999, 181, 1. (10) Pinazo, A.; Infante, M. R.; Park, S. Y.; Franses, E. I. Colloids Surf. B 1996, 8, 1. (11) Pinazo, A.; Wen, X.; Liao, Y.-C.; Prosser, A. J.; Franses, E. I. Langmuir 2002, 18, 8888.
finite solubility (4 ppm), its ability to adsorb via a molecular mechanism, particle sizes and fluidity, and favorable monolayer properties, namely its ability to withstand high surface compression without collapse.10,11 We have used such properties in designing or identifying potential lung surfactant replacement formulations.12-15 In order for a formulation to be viable, it has to exhibit favorable properties at the physiological conditions prevailing in the lung alveolar lining layer. Because of the potential of DLPC as a lung surfactant replacement formulation, it is important to examine its behavior at 37 °C in phosphatebuffered saline, which is near physiological conditions. Little effect of temperature, from 25 to 37 °C is expected, as both temperatures far exceed the Tc. Moreover, since DLPC is zwitterionic (it has a positively and a negatively charged group in its polar headgroup), one would expect a slight, if any, effect of pH/salinity. This has not proven to be the case, as shown below with dynamic surface tension and direct probing experiments, even though its equilibrium, particulate, and monolayer behavior seem equally viable in buffer as in water. 2. Experimental Section 2.1. Materials. Synthetic L-R-dilauroylphosphatidylcholine (DLPC, 99% pure) was purchased from Sigma Chemical Co. (St. Louis, MO). n-Hexane (99+%) was purchased from Sigma Chemical Co. (St. Louis, MO), and ethyl alcohol (200 proof) was purchased from Pharmco Products, Inc. (Brookfield, CT). Sodium chloride (NaCl) and sodium dihydrogen phosphate (NaH2PO4‚ H2O) were analytical reagent grade from Mallinckrodt Specialty (12) Park, S. Y.; Peck, S. C.; Chang, C.-H.; Franses, E. I. In Dynamic Properties of Interfaces and Association Structures; Shah, D. O., Ed.; American Oil Chemists Society Press: Champaign, IL, 1996; p 1. (13) Franses, E. I.; Chang, C.-H.; Chung, J. B.; Coltharp, K. A.; Park, S. Y.; Ahn, D. J. In Micelles, Microemulsions, and Monolayers: Science and Technology; Shah, D. O., Ed.; M. Dekker: New York, 1998; Ch. 18, p 417. (14) Park, S. Y.; Hannemann, R. E.; Franses, E. I. Colloids Surf. B 1999, 15, 325. (15) Wen, X.; Franses, E. I. Langmuir 2001, 17, 3194.
10.1021/la035424w CCC: $27.50 © 2004 American Chemical Society Published on Web 04/13/2004
Behavior of Aqueous DLPC Dispersions Chemicals Co. (Paris, KY). Disodium hydrogen phosphate dodecahydrate (Na2HPO4‚12H2O) was purchased from Fluka Chemical Corp. (Milwaukee, WI). All materials were used as received. The lipid dispersions were prepared on a weight basis. The phosphate-buffered saline solution with pH of 7.2 contained 150 mM of NaCl, 32 mM of NaH2PO4, and 93 mM of Na2HPO4. The pure water used for all samples was first distilled and then passed through a Millipore four-stage cartridge system, resulting in a water resistivity of 18 MΩ cm at the exit port. 2.2. Protocols of Dispersion Preparation and Methods of Dispersion Characterization. Aqueous DLPC dispersions were prepared with different protocols to vary the sizes of the dispersed liquid crystalline particles, or liposome droplets, which were readily observed with polarizing microscopy (Leitz Ortholux microscope), or characterized with dynamic light scattering (DLS). In protocol 1, DLPC dispersions were shaken vigorously at room temperature for about 2 min, decreasing the size of the liposome droplets. In protocol 2, DLPC dispersions were first shaken and then magnetically stirred for an hour to provide a narrower range of liposome sizes. In protocol 3, DLPC dispersions were first shaken, then stirred, and then sonicated (usually for 3 h or more) using a sonicator bath (Branson 1200 Ultrasonic cleaner, Branson Cleaning Equipment Co., Shelton, CT), to further decrease the size of the droplets. By contrast, in protocol 4, dispersions were shaken and then sonicated using a sonicator probe (Sonicator ultrasonic liquid processor, model W-370, Heat Systems-Ultrasonic, Inc., Plainview, NY) until they appeared translucent. Microscopic observations of protocol 1 and 2 systems showed that normally turbid/milky dispersions contain fluid multilamellar liquid crystalline liposomes (sometimes called “onion” structures), mostly with diameters from 1 to 10 µm. Sonication reduced the average sizes considerably. For protocol 3 and 4 systems, few, if any, particles larger than 1 µm were seen in translucent dispersions. A detailed discussion on the proper interpretation of visual observation of aqueous polymer latex or surfactant dispersions can be found in refs 16 and 17. Hydrodynamic radii (RH) of sonicated DLPC dispersions were obtained with DLS using a photon-correlation spectrometer (PhotoCor Complex, PhotoCor Instruments, MD). A heliumneon laser, λ ) 633 nm, was used at 25 °C at a scattering angle of 90°. Samples were placed into 5 mL capped cylindrical vials. The scattered light intensity signals were processed with a PhotoCor-FC correlator, and the data were analyzed using PhotoCor Soft software. The Stokes-Einstein equation was used to determine RH. Dispersions for 1 wt % (10 000 ppm) of DLPC in phosphatebuffered saline (“buffer”), when prepared with protocols 1 and 2, looked white and opaque, showing mostly large droplets as described above. Dispersions prepared with protocols 3 and 4 were translucent and were characterized with DLS. Protocol 3 samples, with DLPC concentration of 1 wt %, have RH ranging from 20 to 24 nm (Figure 1). When diluted to 0.1 wt % (1000 ppm), RH increased to ca. 40 nm. Samples prepared with protocol 4 have RH values ranging from 42 to 50 nm for 1 wt %. The particle sizes were slightly larger than those prepared with protocol 3 and also increased upon dilution. The particle sizes tended to increase with time over days and weeks, because these liposomes or vesicles are thermodynamically unstable. Because the dispersions looked unchanged upon heating from 25 to 37 °C, we inferred that no significant particle size changes occurred at 37 °C during the experiments. Dispersions prepared with different protocols have different dynamic surface tension behavior. 2.3. Surface Tensiometry. A pulsating bubble surfactometer, purchased from Electronetics Co. (now General Transco, Largo, FL), was used for measuring the dynamic surface tension under either constant or pulsating area conditions of DLPC dispersions. Details of its operation have been reported.10,18,19 This instrument uses a pressure transducer for measuring the pressure drop (∆P) (16) Franses, E. I.; Scriven, L. E.; Miller, W. G.; Davis, H. T. Am. Oil Chem. Soc. J. 1983, 60, 1029. (17) Franses, E. I.; Scriven, L. E.; Miller, W. G.; Davis, H. T. Am. Oil Chem. Soc. J. 1983, 60, 1043. (18) Enhorning, G. J. Appl. Physiol. 1977, 43, 198. (19) Electronetics Corp., Amherst, NY (now General Transco, Largo, FL), Pulsating Bubble Surfactometer Surface Tension Measuring Apparatus Operating Manual; 1990.
Langmuir, Vol. 20, No. 10, 2004 4005
Figure 1. Hydrodynamic radii, by DLS, of dispersed particles of DLPC in buffer at 25 °C. The 10 000 ppm dispersions were prepared by sonication in a sonicator bath, protocol 3 ([), or with a sonicator probe, protocol 4 (2); the lower two concentrations were prepared by dilution from 10 000 ppm dispersions. across the air/water interface of a bubble. The surface tension (γ) is then calculated from the Laplace-Young equation for a spherical surface [∆P(t) ) 2γ(t)/R(t)]. At constant area, the radius of the bubble R is 0.40 mm, and the tension measurements are recorded every 50 ms after an initial 1 s delay upon forming a new bubble. At pulsating area conditions, the bubble oscillates between R ) 0.40 and 0.55 mm at frequencies from 1 to 100 cycles/min. The bubble radii were controlled in the instrument by precalibration, and the minimum radius was checked via a microscope. The bubble/liquid chamber was thermostated at 37 or 25 °C. For the experiments performed, the following standard procedure was used. First, the dynamic surface tension [γ(t)] was measured under constant area conditions until the equilibrium or steady-state surface tension was reached. Then, the surface tension was measured at area pulsation at 20 rpm for 45 s, and the pulsation was stopped to recover the original equilibrium surface tension. Then pulsation was resumed for 10-20 min, stopped, and resumed at 80 rpm, similarly as at 20 rpm. A KSV 5000 computer-controlled Langmuir trough (purchased from KSV Instruments, Finland) with a platinum Wilhelmy plate connected to an electrobalance was used for measuring the surface pressure-area (Π-A) isotherm for spread DLPC monolayers. One hundred microliters of 1 mg/mL DLPC solution in 9/1 volume ratio of hexane/ethanol were spread on buffer or water at 37 or 25 °C. Isotherms were recorded by compressing the film (reducing the area with barriers) at a constant rate of 10 mm/min. 2.4. Direct Probing Methods. Ellipsometry and infrared reflection-absorption spectroscopy (IRRAS) are direct probing methods that were used to study the properties of spread DLPC monolayers or adsorbed DLPC layers. Since DLPC does not absorb light at the three wavelengths used in ellipsometry, the absorptivities for the adsorbed layers k1 and the subphase solution or dispersion k2 are nearly zero, as sketched in Figure 2. Moreover, the real refractive index of the subphase (n2) changes little for DLPC concentrations less than 10 000 ppm.11 Therefore, ellipsometry probes primarily the monolayers. The ellipsometric parameters ∆ and Ψ, which are defined in standard monographs,20,21 are not affected by the concentration of the solute in (20) Azzam, R. M. A.; Bashara, N. M. Ellipsometry and Polarized Light; North-Holland: Amsterdam, 1979. (21) Tompkins, H. G. A User’s Guide to Ellipsometry; Academic Press: New York, 1993.
4006
Langmuir, Vol. 20, No. 10, 2004
Phang et al.
Figure 2. Schematic (not to scale) of ellipsometry and IRRAS experiments at the air/water interface, with the subphase containing dissolved molecules or dispersed particles. the subphase solution or dispersion.11,15 By contrast, in IRRAS, not only k1 but also k2 can increase strongly with the concentration of the dissolved molecules or the dispersed particles. Hence, the IRRAS reflectance-absorbance measurements are affected by the bulk subphase absorption for DLPC concentrations more than 1000 ppm, as documented previously11,15,22 and further shown below. A Rudolph Research (now Rudolph Technologies, Flanders, NJ) Auto ELII automatic null ellipsometer was used for measuring the ∆ and Ψ of adsorbed DLPC layers at the air/water interface. Measurements were taken at wavelengths of 633, 546, and 405 nm, with an incident angle of 60° or 70° measured from the surface normal. A Petri dish, filled with DLPC dispersions, was placed on the standard sample stage. Up to 10 measurements of ∆ and Ψ pairs were made at a given wavelength and then averaged. Then ∆o and Ψo for the solvent (water or buffer) was subtracted from the measured ∆ and Ψ values of the DLPC samples to yield δ∆ ≡ ∆ - ∆o and δΨ ≡ Ψ - Ψo. A typical precision in measuring δ∆ and δΨ is (0.005° to (0.2°, depending on the angle and wavelength. A Nicolet Prote´ge´ 460 Fourier transform infrared spectrometer equipped with a liquid nitrogen cooled mercury-cadmiumtelluride (MCT) detector and an external reflection attachment (Graseby Specac Inc.), which has a removable Teflon Langmuir trough, was used to obtain infrared reflection absorption spectra. An incidence angle of 40° measured relative to the surface normal was used. The instrument was continuously purged with dry air from a Balston purge gas generator for reducing the water vapor and carbon dioxide in the sample chamber. Spectra of DLPC dispersions in water were collected using 1024 scans, or 64 scans for the first spectrum, at 8 cm-1 resolution. The Happ-Genzel apodization and one level of zero-filling were employed, yielding the same data spacing as when the spectra were taken at 4 cm-1 resolution. Spectra of DLPC dispersions in phosphate-buffered saline were collected using 512 scans at 4 cm-1 resolution. Again, the Happ-Genzel apodization and one level of zero-filling were employed, yielding the same data spacing as when the spectra were taken at 2 cm-1 resolution. Most peaks were determined quite reproducibly with a standard deviation of ( 0.5 cm-1 or better; hence, differences of 1 and 2 cm-1 in wavenumber are significant. All the spectra were taken using unpolarized light. IRRAS data are reported as plots of reflectance-absorbance (RA) vs wavenumber. RA is defined as -log (R/Ro), where Ro and R are the reflectivities of the pure and the film-covered water (or buffer) surfaces, respectively.
3. Results and Discussion 3.1. Equilibrium and Dynamic Surface Tensions. At 37 °C, the equilibrium surface tension of 1 wt % DLPC dispersion in buffer prepared by protocol 3 is ca. 23 mN/m (Figure 3). Equilibrium surface tensions for other proto(22) Prosser, A. J.; Franses, E. I. Langmuir 2002, 18, 9234.
Figure 3. Dynamic surface tensions of 1 wt % DLPC at 37 °C at constant area: [, in buffer; ], in water. Dispersions were prepared by sonication in a sonicator bath (protocol 3). Results for other protocols and 25 °C for 1 wt % of DLPC in buffer are shown in Table 1.
cols, at 37 or 25 °C, can be found in Table 1. The equilibrium surface tensions of DLPC dispersions in buffer are the same for both temperatures and for the four protocols of preparation. The equilibrium surface tensions for 1010 000 ppm showed no significant effect in either buffer or water (Figure S1 in Supporting Information). The equilibrium surface tension at 37 °C of 1 wt % of DLPC in water and for protocol 3 was similar to that in buffer (Figure 3). This similarity may not necessarily imply that the surface densities of the adsorbed layers are the same (see section 3.2 later). Little effect of concentration at 25 °C and dispersion preparation protocol in water was found.10,11,23 Even though the equilibrium surface tensions are independent of pH/salinity, temperature, concentration, and preparation protocol (except for protocol 1, for which they are 10% higher), the dynamic surface tensions (DSTs) at constant area are quite sensitive to these parameters. The time scales t50 and t95, which are the times for the DST to drop by 50 or 95% of its total drop from γo ) 72 mN/m to its equilibrium value, are used to represent the dynamic adsorption curves conveniently. These equilibrium time scales vary for the same concentration and protocol of preparation by up to 5 times. This variability is due, we believe, to the variability in the particle size distribution, which cannot be fully controlled. The variability in the particle sizes and the equilibration times can be minimized by using identical procedures, as was done for Figure 3. For the same concentration, many direct comparisons indicate that the equilibration times are shorter at 37 °C than at 25 °C and that the shorter equilibration times are obtained with the smaller particles, produced with protocols 3 or 4 (Table 1). For example, the time scales t95 for 1 wt % DLPC dispersion at 37 °C with protocol 3 are 80 s in buffer vs 20 s in water (Figure 3). (23) Phang, T.-L. M.S. Thesis; Purdue University, West Lafayette, IN, 2004.
Behavior of Aqueous DLPC Dispersions
Langmuir, Vol. 20, No. 10, 2004 4007
Table 1. Equilibrium Surface Tensions, Equilibration Times, and Ranges of Tension Maxima γmax and Tension Minima γmin for 1 wt % of DLPC in Phosphate-Buffered Saline ν ) 20 rpm
ν ) 80 rpm
protocol
T, °C
γeq, mN/m
t50,a s
t95,a s
γmin,b mN/m
γmax,b mN/m
γmin,b mN/m
γmax,b mN/m
1 1 2 2 3 3 4 4
37 25 37 25 37 25 37 25
26 ( 2 27 ( 2 24 ( 2 24 ( 2 24 ( 2 24 ( 2 24 ( 2 24 ( 2
150-400 1000-1200 50-400 70-400 20-50 100-150 20-70 20-70
1500-2000 1800-2500 700-900 600-1200 50-200 200-400 40-150 200-300
18-21 19-23 16-23 16-23 8-20 14-20 8-11 10-14
45-53 56-60 42-60 42-60 40-50 45-50 39-42 39-42
16-20 19-21 14-23 15-23 6-18 10-18 4-11 6-10
45-57 57-60 44-60 45-60 45-55 42-50 44-50 44-50
a t b 50 and t95, times for dynamic tension to drop by 50 and 95%; respectively. γmin and γmax, minimum and maximum values of final (b) limit cycle (see Figure 4) for area oscillations at 20 and 80 cycles per minute; boldfaced numbers highlight γmin values less than 10 mN/m.
The faster adsorption at 37 vs 25 °C may arise from a larger diffusivity or larger adsorption rate constants. We speculate that the slower adsorption in buffer than in water may arise either from an expected lower solubility (due to ionic strength), which may lower the dissolution rate, or from an additional adsorption barrier. As the DLPC concentration increases, the adsorption equilibration times decrease. For 10, 100, 1000, and 10 000 ppm in buffer at 37 °C (protocol 3), the average equilibration times, t95, are ca. 5600, 1500, 800, and 80 s, respectively.23 One infers that the adsorption rate increases with increasing particle concentration or with decreasing particle size. These two changes would cause a higher particle surface area, which would cause the net dissolution rate of the particles to increase. After surface tension equilibration of 1 wt % in buffer (protocol 3 at 37 °C), the area was pulsated at 20 rpm for about 45 s, with the radius oscillating between Rmin ) 0.40 mm and Rmax ) 0.55 mm, and the DST was recorded (Figure 4A-a). At the first stage (Figure 4A-a), the local surface tension maximum increased gradually with successive pulsations to ca. 44 mN/m. The small gradual increase is attributed to a small error in Rmax, due to some imperfection or error in the control of the chamber liquid volume. This error decreases with the number of pulsations, and has no impact on our conclusions. Significant adsorption of DLPC molecules to the interface during the area expansion was inferred. This is because the surface tension maximum was less than 60 mN/m, which would be expected on the basis of the pressure-area isotherm if there were no adsorption. The local surface tension minimum decreased, as expected, with successive pulsation cycles to 18 mN/m, which is lower than the equilibrium surface tension of 23 mN/m, because desorption of the extra adsorbed DLPC molecules during the compression of the bubble area back to its original size is not fast enough. Thus, the molecules can accumulate at the interface, causing the maximum nonequilibrium surface density (at or near the area minimum) to be higher than the equilibrium surface density. This is termed “dynamic adsorption hysteresis” and has been discussed and explained before.24,25 After the first stage of pulsation stops, the surface tension at constant area did return to the 23 mN/m equilibrium value, as the surface density relaxed to its equilibrium value. After that, pulsation was started again, continued for 10 min, stopped for about 1 min, and then restarted. The result for 10 cycles is displayed in Figure 4A-b. At this stage, the limit cycle (steady-state oscillation) minimum and maximum surface tensions, γmin and γmax, were 9 and 40 mN/m. Both values are lower (24) Chang, C.-H.; Franses, E. I. J. Colloid Interface Sci. 1994, 164, 107. (25) Chang, C.-H.; Franses, E. I. Chem. Eng. Sci. 1994, 49, 313.
Figure 4. Dynamic surface tensions of 1 wt % DLPC (same systems as in Figure 3), at 37 °C, at area pulsating at 20 rpm: A, in buffer; B, in water; a, pulsation right after equilibration; b, pulsation recorded after 10 min of continuous pulsation. Bubble deformation8 was observed in the B-b case but not in the A-b case.
than those in the first stage of pulsation. Although with the above procedure of pulsation γmin and γmax decreased, pulsation beyond ca. 10 min had no effect (data shown in ref 23). For samples from protocols 1 and 2, the values of γmax were close to 60 mN/m and the values of γmin were close to the γeq of ca. 25 mN/m, indicating less adsorption of additional molecules during the area expansion, and thus less extensive adsorption hysteresis. The γmin values for protocol 3 samples were variable, probably because of variations in the state and size distribution of the dispersed particles. Even though the bath temperature and time of sonication were the same in protocol 3, the net effect of sonication depended on the dispersion volume and possibly other parameters that cannot be controlled well. When DLPC dispersions were prepared with protocol 4, low γmin
4008
Langmuir, Vol. 20, No. 10, 2004
Figure 5. Maximum (4, 2) and minimum (], [) surface tensions during bubble area oscillations at 20 rpm in the final (b) limit cycles (see Figure 4), in buffer (2, [), or in water (4, ]), with protocol 3. The error bars indicate the variability for this preparation protocol; see the text. The broken lines are used to aid the eye.
( (3.0 ( 0.2) × 10-6 mol/m2 in buffer, (26) Dluhy, R. A. J. Phys. Chem. 1986, 90, 1373. (27) Gericke, A.; Hu¨hnerfuss, H. J. Phys. Chem. 1993, 97, 12899. (28) Snyder, R. G.; Strauss, H. L.; Elliger, C. A. J. Phys. Chem. 1982, 86, 5145.
Behavior of Aqueous DLPC Dispersions
Figure 7. RA intensities and wavenumbers of νa-CH2 bands for adsorbed monolayers from DLPC dispersions: in buffer (9, 2) or in water (0, 4) at 25 °C.
the RA intensities at both wavenumbers remain unchanged. We infer that the additionally applied lipid material does not appear in the monolayer and is probably lost because of monolayer collapse. Since the collapsed DLPC monolayer in buffer cannot be detected by IRRAS, it is probably not close to the surface. By contrast, in water, the RA intensity increases slightly and the wavenumber decreases slightly. One infers that some of the additional material remains close enough to the surface and is detected by IRRAS.11 The difference of about 30% in the RA plateau value for buffer vs water indicates that the maximum surface density is smaller in buffer than in water by about 30%. IRRAS was used to study the adsorption of DLPC from bulk DLPC dispersions (Figure 7). At 100 and 1000 ppm of DLPC dispersions in buffer, the RA intensities were about 0.0017 ( 0.0001, similar as those of the close-packed DLPC monolayer in buffer. The RA intensity increased slightly to 0.0020 ( 0.0001 for 5000 ppm and 0.0026 ( 0.0001 for 10 000 ppm. The wavenumber decreased to 2920 ( 0.5 cm-1 for 5000 ppm and 2918 ( 0.5 cm-1 for 10 000 ppm. The RA intensity for 10 000 ppm in water was higher than for the close-packed DLPC monolayer, 0.0032 ( 0.0002 vs 0.0024 ( 0.0002. The RA intensities increase is attributed to the IRRAS detecting some bulkphase particles, which contribute to the reflectance absorbance by affecting the absorption coefficient of the subphase (as explained in section 2.4). The higher RA intensity in water than in buffer for 1000 ppm (for which the RA is not affected by bulk phase absorption) implies a higher surface density in the adsorbed monolayer from water, consistent with the results on the spread monolayers. On the basis of the wavenumbers and the RA intensity results, we concluded that the monolayer could have a higher maximum surface density in water than in buffer. Our hypothesis, sketched in Figure 8, is as follows. In water the electrostatic repulsion between the zwitterionic headgroups is weak, allowing the hydrocarbon chains to
Langmuir, Vol. 20, No. 10, 2004 4009
Figure 8. Schematic (not to scale) of postulated monolayer conformations in water or in buffer. In the buffer the area per molecule is larger, and the chain conformations are more gauche than trans, which implies shorter chains lengths. The phosphate ions are not shown. Table 2. Ellipsometry δ∆ Data for Adsorbed Monolayers from 1000 ppm DLPC in Buffer or Water at 25 °C buffer
water
φo, deg
λ, nm
|δ∆|, deg
φo, deg
λ, nm
|δ∆|, deg
60 60 60 70 70 70
633 546 405 633 546 405
0.53 ( 0.07 0.60 ( 0.13 0.76 ( 0.27 0.30 ( 0.06 0.30 ( 0.09 0.53 ( 0.19
60 60 60 70 70 70
633 546 405 633 546 405
0.65 ( 0.17 0.72 ( 0.10 1.12 ( 0.18 0.32 ( 0.06 0.32 ( 0.09 0.48 ( 0.10
close-pack, with the area per molecule being ca. 0.40 nm2 (corresponding to Γ ) 4.0 × 10-6 mol/m2). The average distance between the polar groups centers, calculated on the basis of the area, the atom sizes, and the bond lengths, is ca. 0.6 nm; the distance between the groups’ edges is ca. 0.2 nm. With buffer, the electrostatic repulsions increase, despite the expected electrostatic screening, resulting in distances of 0.7 and 0.3 nm, respectively. If there is some preferential binding of Na+ or Cl- to the DLPC headgroup, the group will become charged. Such effects have been observed in micellar solutions of a zwitterionic surfactant29 and are analogous to preferential binding of H+ and OH- to proteins, from which the net protein charge is determined to be a function of pH. If the DLPC headgroups are charged, this charge will produce a net electrostatic repulsion, which may increase the distance between the headgroups. If then the chains are driven further apart, the conservation of chain volume implies that their average chain length is shorter, as sketched in Figure 8. With shorter chains, more gauche conformations are allowed. This hypothesis is consistent with the above data on RA intensities and wavenumbers. To further probe the consequences of this hypothesis for the adsorbed layer, ellipsometry measurements were done for 1000 ppm (Table 2). The values of |δ∆| at different incident angles φo ) 60° and 70° increase as the wavelength (29) Chorro, M.; Kamenka, N.; Faucompre, B.; Partyka, S.; Lindheimer, M.; Zana, R. Colloids Surf. A 1996, 110, 249.
4010
Langmuir, Vol. 20, No. 10, 2004
λ decreases from 633 to 405 nm, as expected. The δΨ values are negligible and are not shown here.23 The |δ∆| values of DLPC dispersions are the same as the maximum values for the spread monolayers.11,23 Thus, the adsorbed layers for both systems are monolayers with the same surface densities as the spread monolayers in buffer or water, respectively. At φo ) 60°, the |δ∆| values in water are higher than in buffer by up to 30%, suggesting that the DLPC surface density is slightly higher. Because the |δ∆| values are smaller at φo ) 70° than at 60°, the method is less sensitive at 70° for detecting monolayer thicknesses with a precision of better than 30%. The |δ∆| values for 10 000 ppm in buffer are similar to those at 1000 ppm (Table S1 in Supporting Information). Thus, the ellipsometry results at 70° are not inconsistent with the above hypothesis, and the results at 60° are consistent. 4. Conclusions The dynamic surface tension and equilibrium adsorption behavior of DLPC in buffer vs in water was probed with equilibrium and dynamic tensiometry, IRRAS, and ellipsometry. Adding buffer/saline into the DLPC dispersions in water causes small or no changes on the equilibrium surface tension and the surface pressurearea isotherm. However, the dynamic surface tension and the adsorbed layer, as probed by IRRAS and further supported by ellipsometry, are affected by buffer/saline, even though DLPC is a zwitterionic lipid. The results are consistent with a novel hypothesis that some ions from the electrolyte adsorb or bind selectively to the zwitterionic
Phang et al.
headgroups, thereby producing a net charge in the headgroup. The electrostatic repulsion resulting from this charge is stronger in buffer than in water, despite the expected electrostatic screening effect, which is expected at the higher ionic strength in buffer. Such an effect should be considered for DLPC and other zwitterionic surfactants when a salt or a buffer is used, as under physiological conditions. This effect has implications in the understanding and designing of lung surfactant formulations, in which very low dynamic surface tensions are required, when zwitterionic lipids are used. Acknowledgment. This research is supported in part by the National Institutes of Health (Grant HL-5464102), the National Science Foundation (Grant CTS 0135317), and the Indiana 21st Century Research and Technology Fund. We thank Prof. M. T. Harris and Mr. Sang-Yup Lee, both at Purdue University, for substantial assistance with the DLS results. Supporting Information Available: The equilibrium surface tensions of DLPC, in buffer or in water, as a function of concentration are shown in Figure S1. Figure S2 shows the relation of surface tension of spread DLPC monolayer to surface density. Table S1 shows ellipsometry δ∆ data for adsorbed DLPC layer, from 10 000 ppm DLPC dispersion in buffer or from 100 ppm DLPC dispersion in water. This material is available free of charge via the Internet at http://pubs.acs.org. LA035424W
