Dependence of interfacial film organization on lipid molecular structure.

Article pubs.acs.org/Langmuir

Dependence of Interfacial Film Organization on Lipid Molecular Structure Dorota Matyszewska,†,‡ Slawomir Sek,‡ Elzḃ ieta Jabłonowska,† Barbara Pałys,† Jan Pawlowski,‡ Renata Bilewicz,*,† Fabian Konrad,§ Yazmin M. Osornio,§ and Ehud M. Landau*,§ †

Faculty of Chemistry, University of Warsaw, Pasteura 1, 02-093 Warsaw, Poland Faculty of Chemistry, Biological and Chemical Research Centre, University of Warsaw, Ż wirki i Wigury 101, 02-089 Warsaw, Poland § Department of Chemistry, University of Zurich, Winterthurerstrasse 190, 8057 Zurich, Switzerland ‡

S Supporting Information *

ABSTRACT: Combination of surface analytical techniques was employed to investigate the interfacial behavior of the two designed lipidsN-stearoylglycine (1) and its bulky neutral headgroupcontaining derivative N-stearoylvaline ethyl ester (2)at the air− solution interface and as transferred layers on different substrates. Formation of monolayers at the air−water interface was monitored on pure water and on aqueous solutions of different pH. Crystallization effects were visualized at pure water by recording the hystereses in the Langmuir−Blodgett (LB) isotherms and by transferring the layers onto mica, gold (111), and ITO (indium−tin oxide on glass) electrodes. Subphase pH affects the morphology and patch formation in monolayers of 1, as evidenced by BAM measurements. At pH 8.2, formation of well-ordered crystallites is observed, which upon compression elongate according to predominantly 1-D growth mechanism to form a dense layer of crystallites. This effect is not observed in monolayers of 2, whose headgroup is not protonated. The orientation of layers of 1 transferred to the solid supports is also pH dependent, and their stability can be related to formation of a hydrogen-bonded networks. AFM images of 1 exhibited platelets of multilayer phase. The IR spectra of the ITO substrates covered by 1 indicated formation of hydrogen bonds between the amide groups. The nature of the adsorption layer and its organization as a function of potential were studied indepth by EC STM using Au(111) as the substrate. A model showing the arrangement of hydrogen bonds between adsorbed molecules is presented and related to the observed organization of the layer.

■

dependence of interfacial film organization on subtle structural elements of the lipid molecules is not unequivocally understood, and fine-tuning of the structure and ensuing dynamical and functional properties of such layers based on molecular design is a challenging field of research.14−16 The objectives of this study are to elucidate the effect of lipid’s charge, bulkiness, and propensity for hydrogen-bond formation on the ensuing layer structure, stability, and domain formation at the air− solution, air−solid, and solution−solid interfaces. To achieve these objectives, two analogous lipids were designed and synthesized, in which the saturated C18 hydrocarbon tail and the amide linkage to the headgroups are identical, but distinct variations in the headgroup region are used to address the objectives set: N-stearoylglycine (lipid 1) has a small ionizable polar headgroup whose charge is pH dependent and whose amide moiety can form H-bonded network between adjacent molecules in ordered films; N-stearoylvaline ethyl ester (lipid

INTRODUCTION Molecular organization in two dimensions has captured the imagination of scientists for over a century, starting with the pioneering investigations of Langmuir.1 The structure, function, and dynamics of Langmuir monolayers at the air−water interface can be controlled and measured with a plethora of analytical methods, and their transfer on various solid supports leads to formation of well-ordered Langmuir−Blodgett (LB) mono- and multilayers with predictable structures and interesting functionalities, which can be further studied using microscopic, electrochemical, or spectroscopic techniques.2−4 Science and technology of such thin films have advanced rapidly in recent years, and their applications range from 2D crystallization through membrane studies, molecular electronics, and microlithography to the use of chiral surfaces in chemistry.5 Another interesting field in which Langmuir layers proved to be very useful is formation of model cell membranes.6−8 They can be used to study the interactions with various substances such as drugs, toxins, and peptides.9−13 Whereas the requirements for formation of stable monolayers at the air−water interface are well established, the © 2014 American Chemical Society

Received: May 30, 2014 Revised: August 26, 2014 Published: August 29, 2014 11329

dx.doi.org/10.1021/la502092g | Langmuir 2014, 30, 11329−11339

Langmuir

Article

controlled with software version KSV 5000. A Wilhelmy balance made of a filter paper was used as a surface pressure sensor, and the paper plate was changed after each experiment. Following careful cleaning of the subphase, a few drops of the solution were spread on the subphase and was left for 15 min for complete solvent evaporation. Barrier speed during compression was 10 mm/min (7.5 cm2 /min) at room temperature (21 ± 1 °C). Brewster Angle Microscopy (BAM). BAM pictures were recorded using an UltraBAM objective (Accurion) with lateral resolution of 2 μm and a field-of-view of 800 μm × 430 μm. Pictures were captured during the simultaneous compression of the monolayers at the air−water/buffer interface. Transfer of Layers onto Solid Support. Prior to spectroscopic and microscopic experiments, the layers of lipids 1 and 2 were transferred onto different solid supports (ITO and mica, respectively) by means of the Langmuir−Blodgett technique. The substrate immersed into the subphase (water or buffer) was withdrawn at selected surface pressures with a speed of 25 mm/min to obtain the value of TR = 1.0 ± 0.3. Infrared Spectroscopy (IR) Experiments. Monolayers compressed at the air−buffer solution interface were transferred onto hydrophilic indium−tin oxide substrates (ITO) following careful washing of the ITO with hot acetone, 0.01 M LiOH solution, and water. Water or 0.1 M citric acid buffer, pH 8, was used as the subphase. Layers of 1 or 2 transferred at surface pressures of 35 and 10 mN/m were studied by polarization modulation infrared reflection adsorption spectroscopy (PM-IRRAS) in a Thermo Nicolet 8700 spectrometer equipped with an external table-top optical mount, MCT detector cooled with liquid nitrogen, photoelastic modulator, PEM (PM-100 Hinds Instrument, Hillsboro, OR), and synchronous sampling demodulator, SSD (GWC Instruments, Madison, WI). The IR spectra were acquired the PEM set for the half-wave retardation at 2800 cm−1. The angle of incidence was set at 70°. Typically 1000 scans were performed, and the resolution was 4 cm−1. Spectrum of the bare ITO substrate was recorded under the same conditions and subtracted from the layer spectra. IR spectra of the solid compounds 1 and 2 were measured in KBr pellets. Microscopic Measurements. Monolayers of 1, formed at the air−water interface, were transferred by the Langmuir−Blodgett (LB) method onto mica substrates (KAl2(AlSi3O10)(F,OH)2, muscovite mica, highest grade V1, 20 mm diameter, Ted Pella, Inc.) at the following surface pressures: 10, 20, 35, and 55 mN/m. Electrochemical Scanning Tunneling Microscopy (EC-STM). Experiments were carried out using MultiMode SPM (Veeco, Santa Barbara, CA) equipped with Nanoscope IIIa controller and bipotentiostat. Images were recorded in 0.05 M KClO4 aqueous solution. Tungsten tips were obtained by electrochemical etching in 2 M NaOH aqueous solution and were subsequently isolated with polyethylene in order to minimize faradaic currents. Gold beads, prepared by melting gold wire (diameter of 0.813 mm) using the Clavilier method,17 served as the working electrode. Au(111) facets were used for imaging. Platinum wires were utilized as reference and counter electrodes, but the potential values are referred to the saturated calomel electrode. In EC-STM experiment, gold surface was examined at different potentials in solutions containing vesicles of either N-stearoylglycine (1) or N-stearoyl-L-valine ethyl ester (2) in 0.05 M KClO4. The total concentration of the compounds was ∼5 × 10−4 M. Vesicles were prepared according to an established procedure used for preparation of small unilamellar vesicles.18,19 Atomic Force Microscopy (AFM). Measurements were performed with an Agilent 5500 AFM instrument. Images were recorded in water using MAC Mode AFM with silicon probes (Type VI MAC cantilevers, nominal spring constant 0.2 N/m, resonance frequency in water in the range of 22−24 kHz). Force spectroscopy was performed using PNP-DB silicon nitride probes (Nanosensors).

2) has a neutral headgroup, in which steric constraints inhibit formation of H-bonded networks due to the bulky isopropyl moiety of valine. Interfacial behavior of films formed by these lipids was investigated with Langmuir balance, Brewster angle microscopy, infrared spectroscopy, AFM, and electrochemical STM and was contrasted with the interfacial behavior of stearic acid (lipid 3), thereby revealing the molecular determinants for interfacial film organization. We rationalize our findings in terms of capacity of formation of interfacial, hydrogen-bonded networks, which provides a predictive tool for the design of more elaborate molecular layers.

Figure 1. Surface pressure−molecular area isotherms of 1 (black), 2 (red), and 3 (blue) at the air−water interface. Structures of the lipids are shown adjacent to the respective isotherms. Inset: reciprocal of compression modulus (C s−1) vs surface pressure plot for 1 (black), 2 (red), and 3 (blue).

■

EXPERIMENTAL SECTION

Lipid Syntheses (Figure 1). N-Stearoylglycine (1). A solution of stearoyl chloride (818 mg, 912 μL, 2.7 mmol) in THF (5 mL) was added dropwise over a period of 15 min to a stirred solution of glicyne (211 mg, 2.8 mmol) in 1 N aqueous NaOH (10 mL) at 0 °C under argon. The reaction mixture was stirred for 6 h at the same temperature, diluted with H2O (20 mL), acidified to pH 2 using aqueous HCl (3 N), and stirred for an additional 3 h. The precipitate was filtered off and recrystallized with MeOH to afford 469 mg (51%) of N-stearoylglycine as a white solid. N-Stearoyl-L-valine Ethyl Ester (2). A stirred solution of L-valine ethyl ester hydrochloride (500 mg, 2.75 mmol) and DIPEA (711 mg, 958 μL, 5.5 mmol) in a mixture of anhydrous CHCl3/MeOH (20 mL/ mmol 3:1) was cooled to 0 °C and treated dropwise with a solution of stearoyl chloride (833 mg, 929 μL, 2.75 mmol) in CH2Cl2 (3.0 mL). The reaction mixture was stirred at 0 °C and then allowed to warm to room temperature and stirred overnight. The reaction mixture was poured into brine and extracted three times with CH2Cl2. The combined organic phases were dried over MgSO4 and concentrated under reduced pressure. The residue was purified by column chromatography on silica gel using Et2O/MeOH (98:2) as eluent system to afford 832 mg (73%) of N-stearoyl-L-valine ethyl ester as a white solid. Langmuir−Blodgett (LB) Experiments. Solutions of N-stearoylglycine (1) and its bulkier, neutral headgroup-containing derivative N-stearoyl-L-valine ethyl ester (2) in concentrations ranging from 0.02 to 1.0 mg/mL were prepared in CHCl3/MeOH 8:2 (v/v) (SigmaAldrich and POCh, respectively). Surface pressure vs molecular area isotherms were recorded using a KSV LB trough 5000 (KSV Ltd., Finland) equipped with hydrophobic barriers. The experiment was

■

RESULTS AND DISCUSSION Surface Pressure−Molecular Area Isotherms at the Air−Water and Air−Solution Interface. Compression 11330


Langmuir

Article

Figure 2. Compression−decompression hysteresis of surface pressure−molecular area isotherms of 1 (A, B) and 2 (C, D) at the air −water interface compressed to different surface pressures. Black, red, and blue curves represent first, second, and third compression−decompression cycle, respectively.

pressure plot (inset in Figure 1). The reciprocal of compression modulus is defined as21

isotherms of 1 and 2 at the air−water interface along with the dependence of the reciprocal of compression modulus are shown in Figure 1. They indicate that both lipids form stable monolayers at the air−water interface, whose surface behavior varies because of the differences in the structures of their headgroups (Figure 1). Because of the resemblance in the structure, the isotherm of 1 is first compared to that of stearic acid (3). The only difference between the structures of these two lipids is the presence of the hydrophilic spacer (OC−NH−CH2) between the carboxylate headgroup and the hydrocarbon chain in 1 (Figure 1). The uplift of the isotherm of 1 starts at approximately 35 Å2, and surface pressure rises sharply up to approximately 40 mN/m, where the surface pressure decreases and the monolayer collapses. Compared to the isotherm of stearic acid, the isotherm of 1 collapses at significantly lower surface pressure (approximately 40 mN m−1 compared to 55 mN m−1 for stearic acid). However, contrary to stearic acid, further compression of the monolayer beyond the collapse at 40 mN/m leads to increase of the surface pressure up to values higher than 70 mN/m. Such increase is attributed to the formation of multilayers, which was confirmed by AFM studies of the solid substrates with a layer of 1 transferred at surface pressures higher than the collapse pressure (for further discussion see the SPM section). The observed value of the area per molecule in a well-organized monolayer for 1 is 19.5 ± 1.8 Å2, which is comparable to the value of 22 Å2 for stearic acid.20 Additional information is obtained from the dependence of the reciprocal of compression modulus versus surface

⎛ dπ ⎞ Cs−1 = −A⎜ ⎟ ⎝ dA ⎠

where A is area per molecule and π is surface pressure. The maximal value of the reciprocal of compression modulus (∼150 mN/m) of 1 would indicate that the layer is in the liquidcondensed phase.22 Compared to the values obtained for stearic acid, it is rather small, since stearic acid forms solid-like monolayers with much larger maximal values of the compression modulus, reaching values of 700−800 mN/m.20 In the case of 1 the phase transition from the liquid-expanded to the liquid-condensed phase is not well developed. The small minimum on the compression modulus versus surface pressure plot occurs at a surface pressure of approximately 15 mN/m and may be attributed to the phase transition,23 which is also visible on the isotherms as a temperature-dependent kink at the same surface pressure. At higher surface pressures crystallization is clearly observed, as shown in the BAM images. Thus, the layer is not a typical simple monolayer as in case of stearic acid, where no amide moiety is present. The surface behavior and the isotherms of the bulkier and neutral lipid 2 at the air−water interface differ greatly from those of 1 (Figure 1). First, the uplift of the isotherm occurs at much larger areas per molecule compared to 1. The surface pressure increases up to approximately 10 mN/m, which corresponds to the area per molecule of 55 Å2, where a phase transition takes place. This transition is also clearly observed in the reciprocal of compression modulus versus surface pressure 11331


Langmuir

Article

Figure 3. Surface pressure−molecular area isotherm for monolayers of 1 at the air−solution interface over subphases of pH values of 2.6 (black), 6.0 (red), 7.0 (green), 8.0 (blue), and 10.4 (cyan). Solution spread at the interface: 0.075 mg/mL (1, CH2Cl2/MeOH 4:1, temperature 22 °C). Inset: Cs−1 vs surface pressure plot.

plot (inset in Figure 1). The maximal value of Cs−1 reaches 100 mN/m, which corresponds to the upper limits of liquidexpanded phase of the monolayer22 and is smaller than the maximal value of Cs−1 for 1. It is evident that the introduction of the bulky isopropyl and ethyl groups into the headgroup region of 2 leads to the formation of the monolayer of much more liquid character when compared to the monolayers of 1. In order to compare the surface properties of the two compounds, especially with respect to the possible formation of domains at the air−water interface, cyclic compression− decompression of the monolayers was performed. The monolayers of lipids 1 and 2 were compressed to different selected surface pressures corresponding to the appearance of distinct states of the monolayers. In the case of 1 compressed to surface pressure of 10 mN/m (Figure 2A) or 35 mN/m (data not shown) similar behavior is observeda significant hysteresis occurs for the first compression−expansion cycle, while for the second and third cycle, the hysteresis is much less pronounced, and is almost invisible. This may imply formation of aggregates or domains due to hydrogen bonds between the amide groups of adjacent molecules of the monolayer, which are not relaxed during the following expansion. For the second set of compression−decompression cycles, in which the surface pressure reaches 50 mN/m, the behavior is similar as far as the difference between the consecutive compression−expansion cycles is concerned (Figure 2B). If the observed hysteresis in the first cycle was a result of loss of lipid molecules from the interface due to solubilization in the subphase, the same hysteresis would be expected in the following cycles. Moreover, in the second compression cycle the shape of the isotherm changes significantly compared to the first compression cycle which can be understood assuming irreversible formation of aggregates and partial crystallization of the layers driven by hydrogen bonds between the amide spacers of adjacent lipid molecules. This effect would also account for the observed areas per molecule of the compressed layers which are smaller than the cross-sectional area of a saturated, all-trans-stretched

hydrocarbon chain of ∼0.19 nm2. Such an effect is not observed in the compression−expansion cycles of fatty acids such as stearic acid, which cannot form hydrogen bonds when compressed to surface pressures below the monolayer collapse. In contrast to 1, the second lipid investigated, 2, shows completely different surface behavior in the compression− decompression cycles (Figure 2C). When the monolayer was compressed to relatively low surface pressures, the compression−decompression cycles retrace each other without noticeable hysteresis and only upon compression to 50 mN/ m; i.e., much above the phase transition, clear hysteresis between the compression and decompression cycle is observed (Figure 2D). Interestingly, the following compression cycles show the same hystereses as the first one which is different to the case of lipid 1. A possible explanation of this observation is the formation of metastable domains upon compression which disintegrate when decompression to very large area per molecule takes place (Figure 2D). The difference in behavior between monolayers of 1 and 2 at the air−water interface can be explained by the structure and resulting inability of 2 to form hydrogen bonds because of its ethyl ester headgroup moiety and the isopropyl side group (Figure 1). Hydrogen bonds may be only formed between amide moieties of 2 molecules, but due to bulky substituents this interaction is also not favorable. The hydrogen bond formation is further discussed in SPM and IR section. The interfacial behavior of monolayers of 1 is strongly pHdependent. Increase of the pH from 2.0 to 6.9 results in expansion of the monolayers, evidenced by a shift of the isotherms to larger areas per molecule, an increase of the collapse pressure, and an increase of the compressibility coefficient (Figure 3). Starting from pH 7.0 a phase transition is observed, and the surface pressure at which it occurs increases with pH. In order to visualize the effects of subphase pH on the monolayer morphology of 1, Brewster angle microscopy (BAM) was employed. 11332


Langmuir

Article

Brewster Angle Microscopy at the Air−Water and Air−Buffer, pH 8.2, Interfaces. BAM images were taken at selected points of the compression isotherm of a monolayer of 1 prepared on pure water (Figure 4) and compared with those

Figure 4. Upper panel: BAM pictures of a monolayer of 1 formed on a pure water surface at selected points. Lower panel: surface pressure− molecular area isotherm of a monolayer of 1 on pure water, indicating the points of BAM recordings.

Figure 5. Upper panel: BAM pictures of a monolayer of 1 formed on a citrate buffer, pH = 8.2, at selected points. Lower panel: surface pressure−molecular area isotherms of a monolayer of 1 on a citrate buffer, pH = 8.2, indicating the points of BAM recordings.

for a monolayer prepared at citrate buffer subphase, pH = 8.2 (Figure 5). Significantly, the morphology of the film and formation of the domains and crystallites depend strongly on the pH of the subphase. When the monolayer is formed on pure water, even at large expansion and undetectable surface pressure, patches of the relatively well-organized film can be observed (Figure 4A). Increase in the surface pressure to approximately 8 mN/m leads to crystallization within the patches of the organized layer of 1 (Figure 4B). However, the dark area in the picture, corresponding to the film-free surface of the subphase, indicates that the layer is not yet uniform and does not cover the entire surface of the subphase. Further compression results in both an increase of the surface pressure and formation of a relatively uniform and thick film at the air− water interface (Figure 4C). The last image taken at high surface pressure following the kink observed in the isotherm of 1 reveals the formation of much thicker lipidic structures, as shown by the light color in the image (Figure 4D), confirming the formation of aggregates/crystallites on the water subphase. The BAM recordings of a monolayer of 1 on a subphase of citrate buffer at pH = 8.2 are very different. Compressed to surface pressures lower than approximately 10 mN/m, i.e.,

below the phase transition seen in the isotherm of 1 on buffer solution (Figure 5), the film is homogeneous without any patches or aggregates (Figure 5A). However, in the plateau region of the isotherm which corresponds to the phase transition, formation of well-ordered domains is observed (Figure 5B−D). Upon formation, these objects exhibit regular shapes of small dots with a diameter of approximately 300 nm (Figure 5B). Subsequently, they grow into star-like objects (Figure 5C) and elongate markedly in one dimension (Figure 5D). Following the phase transition the elongated domains grow and coalesce into a dense layer (Figure 5E). It is worth noting that the width of the elongated crystallites observed at surface pressure of approximately 22 mN/m (Figure 5E) corresponds to the diameter of the initially observed dots (Figure 5B), indicating a predominantly 1-D growth mechanism. Finally, a uniform homogeneous layer of 1 at the citrate buffer is formed (Figure 5F). Infrared Spectroscopy of the Layers Transferred onto Solid Hydrophilic Substrate. Layers of compounds 1 and 2 11333


Langmuir

Article

Figure 6. Infrared (IR) spectra. Upper panel (1): IR spectra of compound 1. Curve a depicts the IR spectrum of solid compound 1 in KBr pellet; curves b and c show the PM-IRRAS spectra of layers of 1 transferred onto the ITO plate at 35 mN/m from the buffer at pH = 8.2 and from the pure water subphase, respectively. Lower panel (2): IR spectra of compound 2. Curve a depicts the IR spectrum of solid compound 2 in KBr pellet; curves b and c show the PM-IRRAS spectra of layers of 2 transferred onto the ITO plate at 35 mN/m from the buffer at pH = 8.2 and from the pure water subphase, respectively.

surface. Consequently, the alkyl chains are oriented nearly perpendicular to the surface. Furthermore, the 1705 cm−1 band, which corresponds to the CO stretching motion of the COOH group,26 is absent in the reflectance spectrum (curve b). The two new bands at 1621 and 1412 cm−1 in curve b correspond to the asymmetric and symmetric modes of the carboxylate moiety, indicating that compound 1 is deprotonated. Moreover, the high intensity of these carboxylate bands suggests that COO‑ stretching motions have a large component perpendicular to the substrate, which suggests perpendicular orientation of the alkyl chains. The amide I and amide II modes of 1 are observed at 1644 and 1560 cm−1 for the solid compound (curve a). In the spectrum of 1 transferred onto the ITO these bands are probably shifted and, therefore, overlap with the COOasymmetric stretching mode at 1621 cm−1. Such shift of

on ITO substrates were examined by PM-IRRAS. Because of the surface selection rule for infrared reflectance, only vibrations having components perpendicular to the conductive support contribute to the spectrum.24,25 Consequently, orientation of molecules in the layer can be evaluated by comparison of PM-IRRAS spectra and IR spectra of unbound molecules measured in the transmission mode. Figure 6 shows a comparison between the transmission IR spectrum of compound 1 (curve a) and the PM-IRRAS spectra of layers of 1 transferred onto ITO substrates at 35 mN/m by withdrawing the substrate through the monolayer covering the air−buffer, pH 8.2 (curve b), and the air−water (curve c) interface. The spectrum of the layer transferred from the buffer subphase (curve b) reveals very low intensity of the CH2 stretching modes at 2918 and 2849 cm−1, suggesting that the CH2 plane is oriented nearly parallel to the solid support 11334


Langmuir

Article

Figure 7. MAC AFM images recorded for films of 1 transferred onto mica using the Langmuir−Blodgett method at surface pressures of 55 (A), 35 (B), 20 (C), and 10 mN/m (D).

from the water subphase (curve c). The strong band at 1735 cm−1 is accompanied by a weak amide I band at 1645 cm−1 and a strong amid II band at 1545 cm−1. Such combination of intensities suggests that compound 2 at the surface may attain a specific twisted conformation in which the two carbonyl groups are not in the same plane; e.g., CO bond of the ester group is oriented perpendicular to the surface, while the CO bond of the amide group is parallel. In summary, only layers of compound 1 transferred from the buffer subphase are oriented perpendicularly to the solid support. The stability of such layers is enhanced by the formation of a hydrogen bonding network between amide groups of neighboring molecules. The carboxylate headgroups are hydrophilic and charged and thus easily interact with the cations of the buffer, which may contribute to the facile organization of the film on the buffer surface. In the case of compound 2, where dissociation is not possible, similar layers are formed at both water and buffer subphases. Scanning Probe Microscopic Examination of the Layers. Based on the isotherms and changes of the reciprocal of compression modulus, four values of the surface pressure were chosen for transfer of monolayers of 1 from the air−water interface onto the solid substrates, thereby forming Langmuir− Blodgett films. Figure 7 shows typical images recorded for a layer of 1 transferred onto mica at surface pressures of 55 (A), 35 (B), 20 (C), and 10 mN/m (D).

amide bands suggests the formation of hydrogen bonds between the amide groups of neighboring molecules of compound 1. The spectrum of a layer of 1 transferred at 10 mN/m (Supporting Information Figure 1S) at pH = 8.2 reveals much broader band at 1621 cm−1 with a shoulder at frequency matching the amide II position in the spectrum of the solid sample. This may be ascribed to disorder in the layer and the presence of a small fraction of molecules, not connected by hydrogen bonds. When the layer of 1 is transferred from the pure water subphase (curve c), the CH stretching modes at 2918 and 2849 cm−1 attain high intensity. The CO stretching mode of the COOH group at 1705 cm−1 is not observed in the reflectance spectrum, but also COO− at 1621 cm−1 is not present. The amide I band at 1644 cm−1 has very low intensity while amide II band at 1560 cm−1 is strong. All these observations are expected for a layer of 1 oriented parallel to the ITO surface. In contrast to 1, the IR spectra suggest that the conformation and orientation of 2 at the surface are similar for a layer transferred from the buffer and from the water subphase. Compound 2 contains an ethyl ester group in place of the COOH in the headgroup region, which does not dissociate in buffer or in the pure water subphase. (Figure 6). The CO stretching mode of the ester group is observed at 1735 cm−1 in all spectra: for the pure solid compound (curve a), the layer transferred from the buffer (curve b), and the layer transferred 11335


Langmuir

Article

Figure 8. EC-STM images of Au(111) electrode surface taken in 0.05 M KClO4 solution in the presence of 1. Images were recorded at the potentials of −0.6 (A), −0.3 (B), and +0.2 V (C) vs SCE. Total concentration of 1 was 4.9 ×10−4 M.

In each case, the substrate is covered with irregular patches of the film with significant height differences, indicating that the transfer of 1 does not result in uniform coverage of the hydrophilic mica surface. The thickness of the film transferred at 55 mN/m varies from 5 to 25 nm. Considering that the length of the lipid 1 molecule is about 2.6 nm, and assuming perpendicular orientation of the molecules with respect to the surface, we conclude that the film contains between 2 and 10 monolayers (or 1 and 5 bilayers). This result is reasonable, as a monolayer of 1 collapses at approximately 38 mN/m (Figure 1) and compression to higher surface pressures leads to multilayer formation. However, a quite unusual behavior was noticed for films of 1 transferred at lower surface pressures. Measurements of the reciprocal of compression modulus indicate that between 15 and 38 mN/m the monolayer is in a liquid-condensed state. Within this range, it can be expected that multilayer formation will be less frequent. Surprisingly, transfer at 35 and 20 mN/m results in bilayer formation. Representative topography for each sample is shown in Figure 7. The surface of mica was only partially covered with patches of the film and thickness of the visible islands was about 5 nm. Similar results were obtained for the film transferred at surface pressure of 10 mN/m, which is below the phase transition, where the monolayer is expected to be in the liquid-expanded state. Again the height of the patches covering the mica surface was about 5 nm, indicating strong tendency of lipid 1 to form bilayers. In order to verify the data obtained from the topography imaging, a series of force spectroscopy measurements were performed for the three systems transferred on mica at surface pressures below the collapse pressure. In the force spectroscopy method, the tip approaches the substrate modified with the film and the latter is elastically deformed until indentation occurs. By monitoring the changes in the measured force as a function of the distance separating the tip from the substrate, it is possible to determine the thickness of the film.27 An exemplary force−distance curve is presented in Figure 2S of the Supporting Information. The mean thickness measured for films of 1 transferred onto the mica at 35, 20, and 10 mN/m was found to be 5.2 ± 0.4, 5.4 ± 0.4, and 5.2 ± 0.6 nm, respectively, confirming formation of bilayers. However, the thickness measured for the bilayers in our experiments is either equal to or slightly higher than double of the length of lipid 1. Thus, our data may indicate that a layer of water is present between the bottom of the film and the surface of mica. Because mica in pure water is negatively charged, and the headgroup moieties of 1 are composed of partially negatively charged carboxylates, we may expect repulsive interaction between them. The presence of an aqueous layer separating the film and the substrate may help to screen such electrostatic repulsions.

Interestingly, bilayer formation and coverage are remarkably similar for different surface pressures, representing distinct physical states of the monolayer at the air−water interface. This finding strongly suggests that at surface pressures below the collapse pressure the conditions of the transfer and the actual state of the monolayer at the air−water interface do not affect the final structure of the films deposited on mica. This unusual behavior of 1 is thus related to the intrinsic properties of these molecules and to their strong tendency to self-assemble. The latter can be driven by hydrogen bonding and van der Waals interactions. To examine the self-assembly behavior of molecules of 1, their adsorption behavior on Au(111) electrode was investigated in situ using electrochemical scanning tunneling microscopy (EC-STM). An aliquot of the solution of 1 was added to the electrochemical cell, and high-resolution imaging was performed starting from negative potentials and gradually stepping the potential to more positive values. Figure 8 presents the sequence of the EC-STM images recorded at different potentials applied to the gold electrode. At the negative potential of −0.60 V, bare Au(111) surface is observed with characteristic reconstruction stripes giving rise to a 22 × √3 structure. This confirms the high quality of the electrode surface and its defined crystallographic structure. At this potential adsorption of 1 does not occur, which is similar to previous observations for other compounds possessing long alkyl chain and polar headgroup, including ionic surfactants.28 Upon stepping up the potential to less negative value (−0.30 V), adsorption of 1 sets in, leading to formation of well-ordered monolayers on the Au(111) surface. Interestingly, the molecules are organized in stripe-like structures, and the ordered film consists of several domains which are rotated by 120° with respect to each other. Faint reconstruction lines of the underlying gold substrate are also visible in this image. It is noteworthy that molecules of 1 form ordered adlayers exclusively on the reconstructed Au(111) surface, and upon lifting the reconstruction the ordered structure disappears. Identical behavior is typically observed for alkanes and their derivatives adsorbed with flat-lying orientation on the Au(111) surface.29 The highly organized structure of a monolayer of 1 on Au(111) exists until the potential is switched to the positive value of +0.20 V, at which point the morphology of the film changes and the highly organized monolayer is replaced with an irregular structure. Height analysis indicates that at this stage the molecules change their orientation with respect to the electrode surface, giving rise to a thicker film, in which the molecules most likely adopt a more vertical orientation. This seems to be also related to the fact that at +0.20 V the reconstruction of gold was lifted. 11336


Langmuir

Article

Figure 9. (A) EC-STM image of ordered adlayer of 1 formed on Au(111) surface. (B) Molecular model giving the interpretation of the STM contrast. (C) High resolution image revealing the details of the arrangement of the molecules and the region of hydrogen bonding. (D) Model showing the arrangement of hydrogen bonds between adsorbed molecules, where blue dotted lines correspond to hydrogen bonding between carboxylic acid groups while orange dotted lines correspond to hydrogen bonds between amide moieties. EC-STM images were taken at −0.15 V vs SCE in 0.05 M KClO4. Total concentration of 1 in a solution was 4.9 ×10−4 M.

well as some surfactants on the reconstructed Au(111) surface. It was suggested in several reports that the alignment of the molecules along [011̅], [1̅01], or [11̅0] directions ensures maximum van der Waals interactions between the hydrocarbon chains and the gold atoms at the surface.32 Moreover, this particular arrangement of the molecules also enables maximization of hydrogen bonding between neighboring molecules. Thus, hydrogen bonds can be formed between terminal carboxyl groups of antiparallel molecules, where each headgroup is bound to two others from the opposing row. In addition, the amide groups in neighboring parallel molecules are involved in hydrogen bonding along the direction of the stripe, i.e., perpendicular to the molecular axis. In order to verify the importance of hydrogen bonding in organization of the molecules at the interface, similar experiments were performed with lipid 2. At relatively large negative potential bare gold was observed. Once the potential was stepped to −0.3 V adsorption of 2 occurred, albeit with a highly irregular structure of the adlayer (see Figure 3S in the Supporting Information). Further increase of the potential to +0.2 V resulted in highly disordered structure with some aggregates. Such behavior was reversible; i.e., the same sequence was obtained when the potential was cycled between −0.6 and +0.2 V. Lack of ordered adlayer is strictly related to the molecular structure of lipid 2. In this case formation of the hydrogen bonds is prevented by the presence of bulky isopropyl and ethyl groups in the headgroup region. Moreover, isopropyl side chain hinders also dense packing of the alkyl chains. Taking this into account, the interactions that facilitate

Figure 9 presents an EC-STM image taken for an individual ordered domain at −0.30 V, that is, at the potential at which monolayers of 1 form well-organized films. Each stripe consists of two parallel rows formed by flat-lying molecules which are oriented in an alternating head-to-head and tail-to-tail packing arrangement. This arrangement of molecules on Au(111) surface is reasonable considering that the width of the stripe is 5.0 ± 0.2 nm, which corresponds to nearly double of the length of a single molecule of 1, i.e. 5.2 nm. The distance between adjacent molecules is 0.5 ± 0.1 nm. This value is close to the distance separating the alkyl chains in ordered monolayers of alkanes on Au(111) surface as well as in a 3D crystalline phase.30 Interestingly, high resolution EC-STM images reveal that the positions of molecules in adjacent rows are shifted by half of the molecular width, implying that the tails from one row are aligned along the bridge positions, between molecules of the neighboring row. Such arrangement of the molecules resembles that observed for odd alkanes adsorbed on Au(111), indicating the importance of van der Waals interactions between the substrate and the hydrocarbon chains.31 It was also found that the molecular rows are preferentially oriented along the [12̅1] direction of the reconstructed Au(111) surface. As demonstrated in Figure 9 the angle between the long axes of lipid 1 molecules and the row direction is 90°. Thus, the molecular axes are oriented at an angle of 30° with respect to the reconstruction stripes. As a consequence, the [1̅01] crystallographic direction is the preferred orientation for the adsorption of the molecules. Again such arrangement is characteristic for self-assembled monolayers of odd alkanes as 11337


Langmuir

Article

Insoluble Monolayers at Liquid-Gas Interfaces; Wiley-Interscience: New York, 1966. (2) Organized Monolayers and Assemblies: Structure, Processes and Function; Möbius, D., Miller, R., Eds.; Elsevier: Amsterdam, 2002. (3) Kycia, A. H.; Wang, J.; Merrill, A. R.; Lipkowski, J. Electrochemical and STM studies of 1-thio-β-d-glucose self-assembled on a Au(111) electrode surface. Langmuir 2011, 27, 10867−10877. (4) Agopian, A.; Castano, S. Structure and orientation study of Ebola fusion peptide inserted in lipid membrane models. Biochim. Biophys. Acta, B 2014, 1838, 117−126. (5) Structure and Dynamics of Membranous Interfaces; Nag, K., Ed.; Wiley: Hoboken, NJ, 2008. (6) Maget-Dana, R. The monolayer technique: a potent tool for studying the interfacial properties of antimicrobial and membrane-lytic peptides and their interactions with lipid membranes. Biochim. Biophys. Acta 1999, 1462, 109−140. (7) Brezesinski, G.; Möhwald, H. Langmuir monolayers to study interactions at model membrane surfaces. Adv. Colloid Interface Sci. 2003, 100 −102, 563−584. (8) Su, Z. F.; Jiang, Y. X.; Velázquez-Manzanares, M.; Leitch, J. J.; Kycia, A.; Lipkowski, J. Electrochemical and PM-IRRAS studies of floating lipid bilayers assembled at the Au(111) electrode pre-modified with a hydrophilic monolayer. J. Electroanal. Chem. 2013, 688, 76−85. (9) Dynarowicz-Łątka, P.; Hąc-Wydro, K. Edelfosine in membrane environment - the Langmuir monolayer studies. Anti-Cancer Agents Med. Chem. 2014, 14, 499−508. (10) Leitch, J. J.; Brosseau, C. L.; Roscoe, S. G.; Bessonov, K.; Dutcher, J. R.; Lipkowski, J. Electrochemical and PM-IRRAS characterization of cholera toxin binding at a model biological membrane. Langmuir 2013, 29, 965−976. (11) Dathe, M.; Wieprecht, T. Structural features of helical antimicrobial peptides: their potential to modulate activity on model membranes and biological cells. Biochim. Biophys. Acta 1999, 1462, 71−87. (12) Yuan, C.; Johnston, L. J. Atomic force microscopy studies of ganglioside GM1 domains in phosphatidylcholine and phosphatidylcholine/cholesterol bilayers. Biophys. J. 2001, 81, 1059−1069. (13) Reed, R. A.; Mattai, J.; Shipley, G. G. Interaction of cholera toxin with ganglioside GMl receptors in supported lipid monolayers. Biochemistry 1987, 26, 824−832. (14) Bringezu, F.; Brezesinski, G.; Nuhn, P.; Möhwald, H. Chiral discrimination in a monolayer of a triple-chain phosphatidylcholine. Biophys. J. 1996, 70, 1789−1795. (15) Minones, J., Jr.; Rodriguez Patino, J. M.; Conde, O.; Carrera, C.; Seoane, R. The effect of polar groups on structural characteristics of phospholipid monolayers spread at the air-water interface. Colloids Surf., A 2002, 203, 273−286. (16) Minones, J., Jr.; Dynarowicz-Łątka, P.; Minones, J.; Rodriguez Patino, J. M.; Iribarnegaray, E. Orientational changes in dipalmitoyl phosphatidyl glycerol Langmuir monolayers. J. Colloid Interface Sci. 2003, 265, 380−385. (17) Clavilier, J. Role of anion on the electrochemical behavior of a (111) platinum surface − unusual splitting of the voltammogram in the hydrogen region. J. Electroanal. Chem. 1980, 107, 211−216. (18) Barenholtz, Y.; Gibbes, D.; Litman, B. J.; Goll, J.; Thompson, T. E.; Carlson, F. D. A simple method for the preparation of homogeneous phospholipid vesicles. Biochemistry 1977, 16, 2806− 2810. (19) Sek, S.; Xu, S.; Chen, M.; Szymanski, G.; Lipkowski, J. STM studies of fusion of cholesterol suspensions and mixed 1,2-dimyristoylsn-glycero-3-phosphocholine (DMPC) cholesterol vesicles onto a Au(111) electrode surface. J. Am. Chem. Soc. 2008, 130, 5736−5743. (20) Roberts, G. Langmuir-Blodgett Films; Plenum Press: New York, 1990; p 21. (21) Davies, J. T.; Rideal, E. K. Interfacial Phenomena, 2nd ed.; Academic Press: New York, 1963; p 265. (22) Harkins, W. D. The Physical Chemistry of Surface Films; Reinhold: New York, 1952; p 107.

formation of highly organized monolayers of lipid 1 are substantially hindered in lipid 2. Thus, the resulting adlayer is disordered.

■

CONCLUSIONS The interfacial behavior of the two designed lipids Nstearoylglycine (1) and N-stearoylvaline ethyl ester (2) in monolayers at the air−solution interface and in layers transferred onto different substrates was investigated by a combination of surface techniques, thereby establishing the molecular determinants for film organization. The ability of 1 to form a hydrogen-bonded network between the headgroups of adjacent molecules, as opposed to 2, in which the bulky valine moiety in the headgroup region precludes such hydrogen bonding, greatly affects the interfacial packing arrangement and the ensuing film properties. The formation of hydrogen-bonded networks decreases the fluidity of the resulting layers. Thus, monolayers of 1 are less liquid than those of 2. Compression−expansion experiments reveal irreversible formation of aggregates/domains or partial crystallization of layers of 1, which is also due to hydrogen bond formation, in contrast to the behavior of fatty acids such as stearic acid or lipid 2. BAM images performed at pH 8.2 show formation of well-ordered crystallites of lipid 1 but not of the neutral lipid 2. At high pH, the molecules of 1 transferred onto gold or mica form uniform monolayers stabilized by the interaction of terminal charged groups with the cations of the subphase. On the basis of EC STM, we conclude that the ordered structure of adlayer of 1 results from two main contributions: van der Waals interactions between long alkyl chains and the underlying substrate and hydrogen bonding between terminal carboxylic groups and neighboring amide moieties.

■

ASSOCIATED CONTENT

S Supporting Information *

Figures 1S−3S; NMR spectra of the designed lipids. This material is available free of charge via the Internet at http:// pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (R.B.). *E-mail: [email protected] (E.M.L.). Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Livia Salvati Manni for help with the NMR of the synthetic lipids. This work was supported by grant PSPB 079/ 2010 from Switzerland through the Swiss Contribution to the enlarged European Union. The SPM part of this study was carried out at the Biological and Chemical Research Centre, University of Warsaw, established within the project cofinanced by European Union from the European Regional Development Fund under the Operational Programme Innovative Economy, 2007−2013.

■

REFERENCES

(1) The Collected Works of Irving Langmuir; Suits, C. G., Way, H. E., Eds.; Pergamon Press: London, 1961; Vols. 1−12. Gaines, G. L., Jr. 11338


Langmuir

Article

(23) Kaganer, V. M.; Möhwald, H.; Dutta, P. Structure and phase transitions in Langmuir monolayers. Rev. Mod. Phys. 1999, 71, 779− 819. (24) Greenler, R. G. Infrared study of adsorbed molecules on metal surfaces by reflection techniques. J. Chem. Phys. 1966, 44, 310−315. (25) Kycia, A. H.; Su, Z. F.; Brosseau, C. L.; Lipkowski, J. In Vibrational Spectroscopy at Electrified Interfaces; Wieckowski, A. , Korzeniewski, C., Braunschweig, B., Eds.; Wiley: Chichester, 2013; Chapter 11. (26) Socrates, G. Infrared and Raman Characteristic Group Frequencies; John Wiley & Sons: Chichester, 2001; Chapter 10. (27) Butt, H. J.; Capella, B.; Kappl, M. Force measurements with the atomic force microscope: Technique, interpretation and applications. Surf. Sci. Rep. 2005, 59, 1−152. (28) Sek, S.; Chen, M.; Brosseau, C. L.; Lipkowski, J. In situ STM study of potential-driven transitions in the film of a cationic surfactant adsorbed on a Au(111) electrode surface. Langmuir 2007, 23, 12529− 12534. (29) Yamada, R.; Uosaki, K. Two-dimensional crystals of alkanes formed on Au(111) surface in neat liquid: Structural investigation by scanning tunneling microscopy. J. Phys. Chem. B 2000, 104, 6021− 6027. (30) Denicolo, I.; Douset, J.; Craievich, A. X-ray study of the rotator phase of paraffins (III): Even-numbered paraffins C18H38, C20H42, C22H46, C24H50, and C26H54. J. Chem. Phys. 1983, 78, 1485−1469. (31) Uosaki, K.; Yamada, R. Formation of two-dimensional crystals of alkanes on the Au(111) surface in neat liquid. J. Am. Chem. Soc. 1999, 121, 4090−4091. (32) Marchenko, A.; Xie, Z.-X.; Cousty, J.; PhanVam, L. Structures of self-assembled monolayer of alkanes adsorbed on Au(111) surfaces. Surf. Interface Anal. 2000, 30, 167−169.

11339


The molecular organization of asymmetric lipid bilayers and lipid-peptide complexes.

Molecular tailoring of interfacial failure.

film system speed: its dependence on x-ray energy.

Interfacial water. The structure of interfacial water on gold electrodes studied by x-ray absorption spectroscopy.

Interfacial structure and chemistry of GaN on Ge(111).

Organization of lipids in the tear film: a molecular-level view.

Curvature dependence of the interfacial heat and mass transfer coefficients.

Creation of high-density and low-defect single-layer film of magnetic nanoparticles by the method of interfacial molecular films.

The dependence of lipid monolayer lipolysis on surface pressure.

Helical organization of blood coagulation factor VIII on lipid nanotubes.

Model for anodic film growth on aluminum with coupled bulk transport and interfacial reactions.

Molecular organization, localization and orientation of antifungal antibiotic amphotericin B in a single lipid bilayer.

Micron-scale assessment of molecular lipid organization in human stratum corneum using microprobe X-ray diffraction.

Spectral dependence of glandular tissue dose in screen-film mammography.

Dependence of hydrophobicity of apolar molecules on their molecular volume.

Lipid organization of the plasma membrane.

Functional organization of the HIV lipid envelope.

Peroxidation-induced perturbations of erythrocyte lipid organization.

An unconventional bilayer ice structure on a NaCl(001) film.

Oxygen dependence of lipid peroxidation in mice.

Specific drug binding by purified lipid-reconstituted P-glycoprotein: dependence on the lipid composition.

silica composites via sol-gel.

Temperature dependence of resistive switching behaviors in resistive random access memory based on graphene oxide film.

Structure of human plasma low-density lipoproteins: molecular organization of the central core.