Fatty Amine

We describe the formation of lateral 2D−3D patterns in mixed multilayer LB films of stearic acid (SA) and octadecylamine (ODA) deposited from aqueou...
3 downloads 0 Views 1MB Size
Download PDF
J. Phys. Chem. B 2009, 113, 8581–8587

8581

Lateral 2D-3D Phase Segregation in Fatty Acid/Fatty Amine Monolayers Induced by Langmuir-Blodgett Deposition Elizaveta M. Lomova,† Dmitry S. Turygin,† Alexander A. Ezhov,‡ Vladimir V. Arslanov,† and Maria A. Kalinina*,† Department of Physical Chemistry of Supramolecular Systems, Frumkin Institute of Physical Chemistry and Electrochemistry RAS, 31 Leninsky prospect, Moscow 119991, and Faculty of Physics, M.V. LomonosoV Moscow State UniVersity, 2 Leninskije gory, Moscow 119992, Russia ReceiVed: July 17, 2008; ReVised Manuscript ReceiVed: March 20, 2009

We describe the formation of lateral 2D-3D patterns in mixed multilayer LB films of stearic acid (SA) and octadecylamine (ODA) deposited from aqueous subphases at a basic pH. The 3D particles of SA constituting the micrometer-scale linear assemblies in the LB film are assumed to segregate at the three-phase contact line in the course of film deposition. This 2D-3D phase separation of the two-component system presumably originates from the substrate-induced lowering of the collapse point of SA that leads to spontaneous 3D condensation of an acid on a solid support. The morphology of SA/ODA LB patterns is sensitively influenced by the deposition speed and surface pressure, while the chemistry of the solid support does not affect the resulting structures. The possible mechanism that controls the specific orthogonal arrangement of the 3D phase of SA in the LB film through wettability oscillations is suggested. Introduction In this paper we describe a substrate-induced 2D-3D phase separation in mixed Langmuir-Blodgett films of stearic acid (SA) and octadecylamine (ODA) that results in formation of linear assemblies of 3D micrometer-sized particles of SA in the ultrathin films on solid surfaces. The substrate-mediated formation of laterally ordered linear features in LB monolayers is an effect underlying the so-called LB patterning. This is the method used to create regular patterns with nanometer-sized features in monolayers transferred from the air/water interface onto solid surfaces.1 The substratemediated phase transition2-5 of phospholipids has been studied the most extensively. When deposited at a surface pressure below that of the phase transition point, the homogeneous liquidexpanded phase of the phospholipid spontaneously condensates on a solid surface.6-8 This condensation promotes wetting instabilities and meniscus oscillations that result in a switch from the deposition of the condensed phase to that of the expanded one. The process is repeated in the course of film deposition (i.e., substrate withdrawing from the subphase), yielding alternating zones of deposited material (stripes) and uncoated gaps (channels) on a substrate surface.9-11 The morphology of the resulting patterns can vary substantially (e.g., from parallel to perpendicular orientation with respect to the transfer direction) depending on the experimental conditions, particularly the surface pressure and deposition speed.12,13 As has been proposed in the recent work of Raudino and Pignataro,14 this switching of the structures’ orientation results from dynamic interplay among the surfactant concentration along the meniscus profile, lateral pressure, and deposition speed. * To whom correspondence should be addressed. Phone: (+7) 495-95544-08. Fax: (+7) 495-952-53-08. E-mail: [email protected]. † Frumkin Institute of Physical Chemistry and Electrochemistry RAS. ‡ M.V. Lomonosov Moscow State University.

For mixed monolayers, the LB patterning allows for fabricating linear arrays of laterally segregated components.13 The substrate-induced condensation of DPPC results in microphase separation between the condensed DPPC phase and liquid phase of another component (phospholipids,13 cholesterol,13 or some dyes15) at the three-phase contact line. The LB patterns have also been fabricated from binary DPPC/DLPC monolayers at a surface pressure that exceeds that of the phase transition point for DPPC.12,16,17 The authors assumed that the substrate-induced solidification of condensed DPPC domains controls the nucleation of lipid and formation of phospholipid stripes on a solid surface. Yet another elegant example of substrate-mediated LB patterning has been reported by Huang and co-workers for the stripe alignment of stabilized nanoparticles on silica via dewetting-induced self-organization.18 The stripelike patterns have also been observed in skeletonized arachidic acid/cadmium arachidate bilayers.19 The origin of this lateral phase segregation, which depends on the pH and the salt/acid ratio, differs from that for phospholipids. The formation of alternating stripes of fatty acid and salt was attributed to the complex dynamic interactions between a monolayer deposited from the air/water interface and the substrate covered with an already transferred monolayer.20 The deposition of the charged monolayer is accompanied by a concentrationpolarization21,22 andelectrohydrodynamicinstability22,23 near the contact line that induces a fast transition of fatty acid salt to the acid form. This transition results, in turn, in significant changes of the adhesion work and dynamic contact angle and leads to the meniscus oscillations at a three-phase contact line.20-22 In summary, the LB patterning is a method which is comparatively well-developed both theoretically and experimentally. This approach allows for generating planar linear structures over large areas in a controlled manner.24 The DPPCpatterned surfaces were used for numerous applications such as anisotropic wetting,10 templated electrodeposition of nanow-

10.1021/jp806317c CCC: $40.75  2009 American Chemical Society Published on Web 06/01/2009

8582

J. Phys. Chem. B, Vol. 113, No. 25, 2009

ires,25 guided self-assembly of nanoparticles1 and cell alignment,24,26 pattern transferring,9 etc. Nevertheless, several important opportunities still remain for fundamental investigations in this field. First, the variety of chemicals, which are used for LB patterning with mixed monolayers, is limited to the mixtures of well-studied DPPC though the substrate-induced separation may, in principle, occur in many other binary mixtures of compatible components with different physical and chemical properties. Second, the reported LB patterns have been produced in monomolecular LB films or bilayers, while multilayer structures fabricated via substratemediated phase transition have not been characterized. Finally, the major portion of the research has continued to focus on substrate-mediated LE-LC transition in diluted monolayers and on 2D solidification coupled with the phase separation in condensed films. The substrate-mediated collapse accompanied by the phase segregation in mixed systems remains unexplored. Herein, we report on 2D-3D phase separation of condensed mixed monolayers of SA and ODA at the solid surface during the deposition of LB films from aqueous subphases with basic pH. (These classic amphiphiles have been studied extensively in the one-component Langmuir monolayers,27-30 and their mixed monolayers have been described in detail.31,32) The substrate-induced decrease of the collapse point presumably promotes spontaneous nucleation of the 3D phase of SA, which segregates into linearly arranged particles at the three-phase contact line and thereby gives a linear surface pattern with micrometer-scale periodicity within the 2D LB matrix. The deposition speed and surface pressure affect the morphology (i.e., the size of the pattern features and separation distances) of the resulting patterns. The possible mechanism governing the process of 2D-3D segregation is discussed. Experimental Section Reagents. All chemicals used were of analytical reagent grade. SA was obtained from Fluka. ODA, sodium hydroxide, calcium chloride, and silver nitrate were purchased from Acros Organics (Belgium). Chloroform (Merk) was used as a solvent for the preparation of SA and ODA solutions. Standard solutions of metal salts were prepared with water deionized to 0.20 µS cm-1 conductivity (pH 6.25) through a “Vodoley” system (Russia). The subphase pH for Langmuir monolayer formation was adjusted by the addition of either sodium hydroxide or hydrochloric acid. Langmuir Monolayer and LB Film Formation. A homemade, fully automated Langmuir surface balance with a Wilhelmy glass plate was used for monolayer studies. Teflon barriers and a trough of surface area 268.2 cm2 were sequentially rinsed with a hot mixture of chromic and sulfuric acids and then with acetone and with pure water. To prepare the spreading solution of the SA/ODA (1:1) mixture, 38 µL of a 0.312 mg mL-1 solution of stearic acid in chloroform and 42 µL of a 0.262 mg mL-1 solution of octadecylamine in chloroform were mixed. Monolayers were formed by spreading 80 µL (4.96 × 1016 molecules) of freshly prepared SA/ODA mixed solution on the surface of a dilute aqueous solution of hydrochloric acid of pH 5.00 ( 0.05 and on the surface of a dilute aqueous solution of sodium hydroxide of pH 8.00 ( 0.05 and 10.00 ( 0.05 at 20 ( 1 °C. Equimolar SA/ODA monolayers were also formed on the aqueous subphases of CaCl2 (1 × 10-2 M) and AgNO3 (5 × 10-3 M). Spreading was done using an automatic micropipet (Gilson) delivering 5 µL drops onto a subphase surface in a chessboard-like pattern to distribute the monolayer uniformly. The subphase pH was controlled in situ during the

Lomova et al. experiments to keep its deviations within (0.05. The solvent was allowed to evaporate for 15 min prior to the monolayer compression with a speed of 10 mm min-1. The SA/ODA LB films were deposited vertically onto various supports (silicon wafers, thiolated gold, and quartz supports). To prepare gridlike patterns in LB films, the solid support was rotated by 90° in the substrate plane after the deposition of each bilayer. Optical Microscopy. Optical microscopy images of LB films supported by n-type silicon wafers or tiolated gold were obtained with an Olympus BX51 microscope interfaced to an Olympus C5050 camera. Scanning Electron Microscopy (SEM). Scanning electron microscopy images of LB films were obtained on a JSM-U3 scanning electron microscope equipped with a digital scanner detachable devise and supported by WIN EOS. The LB surfaces were scanned at 15 kV and an electron beam accelerating potential of 1 × 10-10 A. Atomic Force Microscopy (AFM). AFM images of patterned LB films were obtained with an EasyScan AFM microscope equipped with a power supply, LPS-1, an AFM scan head, and NCLR and CONTR sensors, and with a scanning probe microscope, NT-MDT Solver PRO, in AFM mode. All measurements were carried out in room atmosphere and at room temperature (20 ( 2 °C). Raman Spectroscopy. Local Raman spectra were obtained with a HORIBA Jobin Yvon LabRAM HR Raman microscope. The argon ion laser operating at 488.0 nm was used as a light source. The density of laser power on a sample surface was varied from 15 to 500 W cm-2. To study the composition of the phase-segregated LB film, the 15-layer SA/ODA LB film was deposited onto glass supports (TF-1 glass, 20 × 20 mm) covered with a Cr adhesion sublayer (5 nm) and a polycrystalline Au layer (50 nm) by the Analytical-µSystem (Germany). The LB film was transferred onto the support at a surface pressure of 37 mN m-1 with a deposition speed of 10 mm min-1. The reference SA and ODA samples were formed by spreading and further evaporating 20 µL of their individual chloroform solutions on gold-covered supports. Raman spectra were obtained from 3D particles with an average diameter d of about 2 µm (the area A for collecting the spectral signal was about 3.5 µm2), from the particle-free regions of the LB film (A ) ca. 1000 µm2, d ≈ 35 µm), and for reference SA and ODA (A ) ca. 4000 µm2, d ≈ 70 µm). Field Emission Auger Electron Spectroscopy (AES). A PHI 680 Auger system was used for surface elemental analysis of 2D and 3D phases in the seven-layer SA/ODA LB film. This instrument has an elemental detection limit of 0.1 atom %. Typical parameters for obtaining the Auger spectra were 5 keV (1 nA) and 10 keV (10 nA) electron beam energy to achieve lateral resolution of 80 and 40 nm, respectively. The depth resolution was ca. 5 nm. For 3D phase analysis, the exposition time varied from 48 s to 5 min. Five different points were measured for 3D particles and for the 2D surrounding phase, respectively. The survey scans were collected at 1 eV/step resolution, and each spectrum was averaged over 30 scans to obtain satisfactory signal-to-noise ratios. The exposition time ranged from 43.78 s to 4. 38 min. Results and Discussion We observed lateral structural patterns in multilayer LB films of stearic acid and octadecylamine deposited from aqueous subphases onto solid supports at a basic pH. The surface pressure isotherms for the monolayers of individual components and that

Phase Segregation in Fatty Acid/Fatty Amine Monolayers

Figure 1. Surface pressure vs area isotherms for stearic acid (1) and octadecylamine (2), and their equimolar mixture (3) on the aqueous subphases equilibrated at pH 8.0.

of the equimolar SA/ODA mixture at pH 8.0 are shown in Figure 1. The negative deviation from ideal behavior is attributed to the attractive electrostatic interactions between oppositely charged acid and amine entities.31,32 At the same deposition speed (10 mm min-1), the formation of linear surface patterns was observed for the LB films transferred from the subphase at a surface pressure g35 mN m-1 (Figure 2a). (At lower surface pressure, the deposition of mixed monolayers gives morphologically uniform LB films). The gridlike pattern found in the 21-layer LB film is formed by 3D round-shaped or elongated particles (Figure 2a). As shown in Figure 2b, these particles are aligned both perpendicular and parallel to the transfer direction (hereafter horizontally and vertically, respectively). The arrangement of particles in horizontal lines, which resemble the strings of beads, differs from that in vertical ones, which represent the assemblies of separate elements with continuously diminishing size. Figure 2c shows an AFM image of the single morphological element

J. Phys. Chem. B, Vol. 113, No. 25, 2009 8583 of such a pattern formed in the 17-layer LB film; the maximal diameter and height of the particles are ca. 6 and 0.6 µm, respectively. As illustrated by the SEM image of the 27-layer mixed LB film, the number and size of 3D particles in horizontal lines increases with the number of deposited layers (Figure 2d). The composition of 3D particles in the 15-layer SA/ODA LB film deposited onto a gold-evaporated glass was studied by local Raman spectroscopy (Figure 3a). The reference spectra obtained from solid ODA (Figure 3b) and SA (Figure 3c), equal volumes of which were spread from their chloroform solutions onto a gold-covered glass support and then evaporated, were used to make the band assignment. In the region 1500-1700 cm-1, which typically presents the -CdO vibration bands in the spectra of fatty acids,33 the distinct band assigned to the carbonyl group vibration mode was observed in the spectrum of reference SA; the smaller peak shifted toward lower frequencies was also found in the spectrum of 3D particles. In the region 3100-3400 cm-1, the spectrum of reference ODA showed the band near 3335 cm-1 assigned to the -N-H stretching vibration mode,33 while the spectra of 3D particles and reference SA remained smooth. We assumed therefore that 3D particles represented the segregated SA phase or, at least, they were enriched with a fatty acid. In the region 2800-3000 cm-1, the CH2 stretching bands were observed in both the local spectra of 3D particles and surrounding LB film (data are available in the Supporting Information). Unfortunately, the signal obtained from thin surrounding regions was too weak to distinguish the bands of carbonyl or amine groups. To confirm the lateral distribution of the components within the film, the elemental analysis of the three-layer LB film was carried out via Auger electron spectroscopy (spectral data not shown; the local AES spectra and the elemental maps are available in the Supporting Information). The formation of such an ultrathin film for elemental analysis allows for reducing the possible contamina-

Figure 2. Linear patterns in SA/ODA (1:1) mixed LB films obtained with a deposition speed of 10 mm min-1 at a surface pressure of 37 mN/m and subphase pH of 8.0: (a, b) representative SEM images of a 21-layer SA/ODA LB film, (c) AFM image of a single linear feature of the pattern formed in a 17-layer SA/ODA LB film, (d) SEM image of a 21-layer SA/ODA film on a silica surface. The scale bar for these images is 30 µm; the transfer direction is indicated by arrows.

8584

J. Phys. Chem. B, Vol. 113, No. 25, 2009

Lomova et al.

Figure 3. (a) Raman spectra obtained on 3D particles formed in a 15-layer LB film deposited onto gold-evaporated glass at a surface pressure of 37 mN/m and subphase pH of 8.0. (b, c) Raman spectra of reference solid ODA (b) and SA (c) evaporated from their individual chloroform solutions on gold-covered supports.

tion of the sample in the course of film deposition/drying. The AES analysis was performed on 3D particles with a diameter up to 500 nm and on surrounding 2D regions of similar area. For the 2D phase exposed to the electronic beam for ∼5 min, four elements were determined: Si, 45.3%; O, 27.0%; C, 25.2%; N, 1.5% (the atomic percentages are given). The determined amount of nitrogen comprises ∼5.6% relative to that of carbon. This ratio almost exactly corresponds to an 18:1 ratio of carbon to nitrogen, when estimated on the assumption that the 2D phase only consists of ODA. The large amount of oxygen was attributed to the SiO2 on the support surface, which also presents nonoxidized silica. For 3D particles, only three elements were determined (representative results for three different 3D particles are given): Si, 30.3-36.0%; C, 48.8-52.0%; O, 16.2-17.7%. The amount of nitrogen in the 3D phase was below the detection limit irrespective of the exposition time, although the total amount of carbon in 3D particles increased significantly compared to that in the 2D phase. These results also corroborate the segregation of the LB monolayer into the 2D phase enriched with ODA and the 3D phase consisting of SA. The lateral structure of the segregated LB film was sensitively influenced by the experimental conditions. At the same deposition speed, increasing the transfer surface pressure decreased the size of 3D elements and increased the length of vertical lines (Figure 4). The morphology of the LB patterns does not depend on a compression speed value; however, it is sensitively influenced by the deposition speed. Pattern formation was observed for the deposition speed values ranging from 5 to 50 mm min-1. As the deposition speed increased, the spacing of horizontal lines and size of 3D particles decreased, whereas the ordering of vertical lines appeared to be hindered. Instead, the surface between horizontal lines was densely covered by randomly distributed particles (Figure 5a). The already formed patterns do not guide the formation of linear assemblies in the subsequently deposited monolay-

Figure 4. (a, b) Optical microscopy images of a 15-layer SA/ODA LB film deposited onto a silica surface with a speed of 10 mm min-1 at a surface pressure of 45 mN/m and suphase pH of 8.0. The scale bar for these images is 60 µm; the transfer direction is indicated by arrows.

Figure 5. (a) Linear pattern in 15-layer SA/ODA LB films deposited onto the support with a fixed orientation with a speed of 50 mm min-1. (b) Optical microscopy images of a 15-layer LB film with alternating bilayers of different in-plane orientations, which was obtained using the 90° rotation of the support after the transfer of each subsequent bilayer. The scale bar for these images is 60 µm; the transfer direction is indicated by arrows.

ers. Figure 5b shows an optical image of the LB film with alternating bilayers of different in-plane orientations, which was obtained using the 90° rotation of the support during the fabrication of the film. The addition of some reactive ions into the subphase inhibits substrate-induced separation in the SA/ODA system. We did not observe ordered 3D

Phase Segregation in Fatty Acid/Fatty Amine Monolayers structures in the films transferred from aqueous solutions of calcium or silver salts. Unlike the chemistry of the subphase, the chemistry of solid supports had a negligible effect on the resulting patterns. At the same surface pressure and deposition speed, the pattern obtained on hydrophobic thiolated gold was very similar to that formed on a hydrophilic silica surface. The surface of one SA/ODA LB monolayer on silica and those of bilayer and six-layer LB films on thiolated gold were examined via the AFM tool. The 3D particles (∼1 µm in diameter) were found in the bilayer, while the surface of the LB monolayer was topologically uniform. An increase in the size of the particles in the six-layer film compared to that in the bilayer (from ∼20-30 to ∼70-90 nm height) is proportional to the number of deposited layers. (AFM data and the optical microphotograph of the SA/ODA LB film on thiolated gold are available in the Supporting Information.) We assumed therefore that the 2D-3D phase segregation is most likely induced by the interactions evolving between hydrophilic groups of a monolayer, which is already immobilized on a support surface, and those of a monolayer deposited from a subphase. We note especially that the equimolar ratio of components and basic pH are also required for the formation of patterns in SA/ODA LB films. Changing the component ratio or decreasing the subphase pH resulted in a visible decrease in the number and size of segregated 3D particles to the point of disappearance. These limitations empirically evidence the important role of electrostatic acid-base interactions, the maximum of which is attained for the equimolar SA/ODA mixtures,32 in the substrate-induced phase segregation. Although the obtained experimental data are not enough to establish a detailed mechanism of such patterning, we made some general assumptions regarding the origin of the process. When spread from premixed solution, SA and ODA form homogeneous mixed Langmuir monolayers below the collapse point that have been corroborated by Brewster angle microscopy studies.31 Hence, the SA particles constituting the pattern nucleate directly at the three-phase contact line during the LB deposition rather than in a floating film in the course of monolayer compression. As proposed by Spratte and Riegler,5 the contact of a Langmuir monolayer with the solid surface is equivalent to the decrease of the system’s temperature. The surface pressures of phase transitions are generally decreased with decreasing temperature. Monolayer collapse can also be considered as a 2D-3D phase transition of the Langmuir monolayer. Consequently, the LB deposition of a condensed mixed monolayer onto the solid support can promote phase microseparation and 3D condensation of the component with a higher melting point (the melting temperature of SA is 70-72 °C, and that of ODA is 50-52 °C). The formation of linear patterns in SA/ODA LB films might be therefore attributed to substrate-mediated microphase separation of the condensed SA/ODA monolayer at surface pressures >35 mN m-1. This process can be described by a simplified phase diagram (Figure 6a), which illustrates the collapse of floating and deposited SA/ODA monolayers on a water surface and on a solid substrate, respectively. The ordinates of Figure 6a show the collapse surface pressure πcollapse measured with a Wilhelmy balance for the floating monolayer. The dashed lines describe the collapse transition of a floating SA/ODA monolayer. (This diagram was plotted using the experimental data obtained in our laboratory and those reported in refs 31 and 32). The solid lines describe the deposited monolayer collapsing at a surface pressure π′collapse. The

J. Phys. Chem. B, Vol. 113, No. 25, 2009 8585

Figure 6. Schematics illustrating the possible mechanism for the substrate-induced phase separation in SA/ODA mixed LB films. (a) Simplified phase diagram for floating and deposited SA/ODA mixed monolayers collapsing on water and a solid surface, respectively. The diagram of the deposited monolayer (solid line) is shifted to lower surface pressures compared to that of the floating monolayer (dashed line). The deposited monolayer is segregated into 2D and 3D phases at a surface pressure πtransf ) πc, which corresponds to the composition χeq. Images a and b describe the stepwise formation of a linear assembly of 3D particles of SA in a mixed monolayer contacting the solid surface: (a) crystallization of SA into round-shaped aggregates resulting in the meniscus wavelike curvature and local instabilities due to the increased surface tension, (b) segregation of the unstable portions of a monolayer into drops (dashed lines), (c) equilibration of the meniscus contact line. The transfer direction is indicated by an arrow along the set of figures; the small arrows in (a) and (b) show the directions of the equilibrating forces at the meniscus. (d) The formation of the orthogonal assembly of 3D particles is followed by the deposition of the structurally uniform 2D phase depleted of SA.

horizontal line corresponds to the surface pressure πtransf, at which the monolayer is transferred onto the solid support. This pressure is lower than that for the collapsing floating monolayer with an equimolar SA/ODA ratio (χeq). For π′collapse = πtransf, the components are segregated and the portion of a monolayer with 3D particles of SA is deposited onto a solid surface. After a certain period of time t1, the composition of a floating monolayer at a distance l from the three-phase line is changed because of continuous diffusion/ depletion of SA. The molar ratio of components corresponds to that at point χ. The system reaches a state for which πtransf < π′collapse, and the mixed 2D phase depleted of SA is deposited. When the molar ratio is re-established and becomes equal to χeq, the successive line of SA particles is formed. However, this diagram alone does not adequately explain the morphology of the resulting pattern. In particular, the size of SA islands, which are comparatively large even in the bilayer LB film, and the microscale periodicity of the pattern’s features suggest that the diffusion of SA from a homogeneously mixed floating monolayer to the collapsing region at the contact line is much faster (by at least 3 orders of magnitude) than that determined for condensed floating monolayers of fatty acids with the diffusion coefficients on the order of 10-9 m2/s.20 This effect cannot be easily understood, in our opinion, by simple modeling of the diffusion of SA molecules within the floating 2D phase toward the 3D islands, which grow on a continuously moving solid support. The diffusion accelerationsas compared to that in a floating monolayersmight be related to the complex action of several common driving factors such as a large concentration gradient, Marangoni effect, and healing of defects (e.g., cracklike channels or gaps arising as a consequence of wetting instabilities on the nucleating particles) as well as with nonequilibrium hydrodynamic effects and the instability of the surface of the growing 3D phase. The development of an adequate theoretical model of the process therefore requires further investigations on the described phenomena. Nevertheless, one important experimental observation provides useful information about the mechanism of pattern

8586

J. Phys. Chem. B, Vol. 113, No. 25, 2009

formation. The surfaces of SA/ODA LB patterns were microscopically examined immediately after the deposition of the film, and small water drops rapidly evaporating from the support were observed. These drops were aligned at the horizontal lines in the LB pattern. We hypothesized that the line of hydrophobic particles retards the meniscus upward slipping (a similar effect is commonly observed in the course of wetting/dewetting processes on structurally heterogeneous or porous surfaces34) and thereby promotes a wavelike profile of the contact line (Figure 6b). The particle growth is most likely accelerated by the Ostwald ripening, which supplies the transfer of the material from smaller elements to the larger ones.35,36 The local surface tension on the tops of the “waves” increases due to the substantial depletion of these zones of SA molecules, and thus, the meniscus becomes thermodynamically unstable. The unstable portion of the monolayer on the meniscus, the composition and surface tension of which differ significantly from those in the floating monolayer,37 is spontaneously separated into drops, which thus isolate these instabilities from a monolayer (Figure 6c). This mechanism presumably guides the formation of each extended horizontal line with adjacent vertical ones. At a certain critical point of the SA depletion, the 3D condensation is switched to the transfer of the ODA-enriched expanded phase (Figure 6d). The process self-replicates during the deposition, yielding alternating 2D and 3D stripes (Figure 6e). Finally, the SA molecules condensate on the already formed 3D elements in the course of subsequent deposition steps that leads to a gradual increase in the particle’s size with the number of deposited layers. The deposition speed and surface pressure are therefore primary factors that control the morphology of forming patterns. The SA particles are nucleated and deposited more quickly than acid molecules diffuse from a floating monolayer toward the contact line as the deposition speed increases, and the size and lateral distribution of SA particles within the pattern change from large and orthogonal to small and random. The rate of SA diffusion decreases with increasing surface pressure (i.e., χ approaches χeq in the phase diagram, and l becomes shorter compared to that for lower πtransf), and thus, higher pressures yield patterns with smaller particle spacing and size. Although the electrostatic interactions seem to significantly impact the process, the suggested mechanism of 2D-3D patterning does not yet account for their influence. The origin of such an electrochemical effect is most likely complex and might be related to a number of interfacial phenomena, including the dependence of mechanisms of proton transfer within the monolayer on the subphase pH,38 lateral pH gradient along the meniscus toward the bulk,39 concentration polarization, electrodiffusion,20 etc. The detailed understanding of the effect of the electrochemical factor on the substrate-induced 2D-3D phase separation in SA/ODA LB films requires further investigations on the electrochemistry of these systems; our laboratory actively pursues this goal. Conclusion We described substrate-mediated phase separation in mixed Langmuir monolayers of stearic acid and octadecylamine on a basic subphase that results in formation of micrometer-scale linear patterns in multilayer LB films. The effect primarily originates from the decrease of the collapse point of SA in the mixed monolayer contacting a solid surface at high surface pressures that promotes spontaneous condensation of SA into 3D particles on a solid support. The specific morphology of formed LB patterns constituting orthogonally ordered SA

Lomova et al. particles presumably originates from the oscillations of wettability and segregation of the unstable portion of the monolayer into droplike islands at the three-phase contact line. The deposition speed and surface pressure control the periodicity and size of the pattern’s features. Although the mechanism of 2D-3D LB patterning remains poorly understood in several important aspects (namely, the micrometer-scale periodicity of the pattern features and the role of electrostatic interactions in this process), we believe that similar phenomena might occur in other mixed monolayers at certain conditions, and this type of substrate-mediated phase separation in Langmuir monolayers can be optimized to tailor various surfaces with different functionalities in a controllable fashion. Acknowledgment. We thank the Russian Foundation for Basic Research for financial support (Grant No. 07-03-13519). Supporting Information Available: Raman spectra (-CH2 stretching region) of patterned LB films, AES spectra for 2D and 3D phases in patterned LB films, AFM data on the morphology of one-layer, two-layer, and six-layer SA/ODA LB patterns, and optical microphotograph of an SA/ODA LB film deposited onto thiolated gold. This information is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Gleiche, M.; Chi, L. F.; Fuchs, H. Nature 2000, 403, 173. (2) Reigler, H.; Spratte, K. Thin Films 1995, 20, 349. (3) Sikes, H. D.; Woodward, J. T.; Schwartz, D. K. J. Phys. Chem. 1996, 100, 9093. (4) Sikes, H. D.; Schwartz, D. K. Langmuir 1997, 13, 4704. (5) Spratte, K.; Reigler, H. Langmuir 1994, 10, 3161. (6) Spratte, K.; Chi, L. F.; Reigler, H. Europhys. Lett. 1994, 25, 211. (7) Sanchez, J.; Badia, A. Thin Solid Films 2003, 440, 223. (8) Lenhert, S.; Gleiche, M.; Fuchs, H.; Chi, L. ChemPhysChem 2005, 6, 2495. (9) Lu, N.; Gleiche, M.; Zheng, J.; Lenhert, S.; Xu, B.; Chi, L.; Fuchs, H. AdV. Mater. 2002, 14, 1812. (10) Gleiche, M.; Chi, L.; Gedig, E.; Fuchs, H. ChemPhysChem 2001, 3, 187. (11) Hirtz, M.; Fuchs, H.; Chi, L. J. Phys. Chem. B 2008, 112, 824. (12) Moraille, P.; Badia, A. Langmuir 2002, 18, 4414. (13) Chen, X.; Lu, N.; Zhang, H.; Hirtz, M.; Wu, L.; Fuchs, H.; Chi, L. J. Phys. Chem. B 2006, 110, 8039. (14) Raudino, A.; Pignataro, B. J. Phys. Chem. B 2007, 111, 9189. (15) Chen, X.; Hirtz, M.; Fuchs, H.; Chi, L. AdV. Mater. 2005, 17, 2881. (16) Moraille, P.; Badia, A. Langmuir 2003, 19, 8041. (17) Moraille, P.; Badia, A. Angew. Chem., Int. Ed. 2002, 41, 4303. (18) Huang, J.; Kim, F.; Tao, A. R.; Connor, S.; Yang, P. Nat. Mater. 2005, 4, 896. (19) Mahnke, J.; Vollhardt, D.; Stockelhuber, K. W.; Meine, K.; Schulze, H. J. Langmuir 1999, 15, 8220. (20) Kovalchuk, V. I.; Bondarenko, M. P.; Zholkovskiy, E. K.; Vollhardt, D. J. Phys. Chem. B 2003, 107, 3486. (21) Kovalchuk, V. I.; Zholkovskiy, E. K.; Bondarenko, M. P.; Vollhardt, D. J. Phys. Chem. B 2004, 108, 13449. (22) Bondarenko, M. P.; Kovalchuk, V. I.; Zholkovskiy, E. K.; Vollhardt, D. J. Phys. Chem. B 2007, 111, 1684. (23) Kovalchuk, V. I.; Zholkovskiy, E. K.; Bondarenko, M. P.; Vollhardt, D. J. Phys. Chem. B 2001, 105, 9254. (24) Lenhert, S.; Zhang, L.; Mueller, J.; Weismann, H. P.; Erker, G.; Fuchs, H.; Chi, L. AdV. Mater. 2004, 16, 619. (25) Zhang, M.; Lenhert, S.; Wang, M.; Chi, L.; Lu, N.; Fuchs, H.; Ming, N. AdV. Mater. 2004, 16, 409. (26) Lenhert, S.; Meier, M.-B.; Meyer, U.; Chi, L.; Wiesmann, H. P. Biomaterials 2005, 26, 563. (27) Dynarowicz-Latka, P.; Dhanabalan, A.; Oliveira, O. N., Jr. AdV. Colloid Interface Sci. 2001, 91, 221. (28) Schwartz, D. K. Surf. Sci. Rep. 1997, 27, 241. (29) Gaines, G. L. Insoluble Monolayers at Liquid-Gas Interfaces; Interscience: New York, 1966. (30) Li, C.; Zhao, B.; Lu, Y.; Liang, Y. J. Colloid Interface Sci. 2001, 235, 59.

Phase Segregation in Fatty Acid/Fatty Amine Monolayers (31) Lee, Y. L.; Yang, Y. C.; Shen, Y. J. J. Phys. Chem. B 2005, 109, 4662. (32) Stosch, R.; Cammenga, H. K. J. Colloid Interface Sci. 2000, 230, 291. (33) Smith, E.; Dent, G. Modern Raman Spectroscopy: A Practical Approach; Wiley: New York, 2004; p 146. (34) DeGennes, P. G. ReV. Mod. Phys. 1985, 57, 827.

J. Phys. Chem. B, Vol. 113, No. 25, 2009 8587 (35) (36) (37) (38) (39)

Finsy, R. Langmuir 2004, 20, 2975. Madras, G.; McCoy, B. J. J. Colloid Interface Sci. 2003, 261, 423. Legrange, J. D. Phys. ReV. Lett. 1991, 66, 37. Bero, J. M.; Eriksson, L. G. T. Langmuir 1994, 10, 1213. Reigler, J. E.; LeGrange, J. D. Phys. ReV. Lett. 1988, 61, 2492.

JP806317C