Layer by Layer Self-Assembled Polyelectrolyte Multilayers with

Brigitte Städler , Rona Chandrawati , Kenneth Goldie and Frank Caruso ... Qingmin Ji, Masahiko Miyahara, Jonathan P. Hill, Somobrata Acharya, Ajayan ...
0 downloads 0 Views 181KB Size
7854

Langmuir 2005, 21, 7854-7859

Layer by Layer Self-Assembled Polyelectrolyte Multilayers with Embedded Phospholipid Vesicles Obtained by Spraying: Integrity of the Vesicles M. Michel,†,‡ A. Izquierdo,‡ G. Decher,‡ J.-C. Voegel,† P. Schaaf,*,‡ and V. Ball† Institut National de la Sante´ et de la Recherche Me´ dicale, Unite´ 595, Faculte´ de Me´ decine, 11 rue Humann, 67085 Strasbourg Cedex, France, and Centre National de la Recherche Scientifique, Unite´ Propre 22, Institut Charles Sadron, 6 rue Boussingault, 67083 Strasbourg Cedex, France Received February 24, 2005. In Final Form: May 11, 2005 In a previous paper (Michel, M.; Vautier, D.; Voegel, J.-C.; Schaaf, P.; Ball, V. Langmuir 2004, 20, 4835), we showed that phospholipid vesicles can be incorporated into poly(glutamic-acid)/poly(allylamine) (PGA/ PAH) multilayered polyelectrolyte films built by the alternated dipping of a surface in polyanion and polycation solutions. AFM imaging, quartz crystal microbalance, and ellipsometry suggested that the vesicles remain intact when adhering on the surface. In the present paper, we show that such films can also be realized by spraying both the polyelectrolyte solutions and the vesicles onto the surface. Using such vesicles filled with ferrocyanide ions, we prove by cyclic voltammetry that the sprayed vesicles remain intact when embedded in the multilayers. We show that multilayers containing two distinct layers of intact vesicles separated by several polyanion/polycation bilayers can also be constructed. Polyelectrolyte multilayers containing layers of phospholipid vesicles could act as reservoirs for drug or other biologically active molecules in controlled release bioactive coatings.

I. Introduction The alternate deposition of polyanions and polycations onto a charged solid substrate leads to the formation of polyelectrolyte multilayers. Such films, with thicknesses ranging from the nanometer up to several micrometers, depending upon the nature of the polyelectrolytes, their deposition conditions, and the number of deposited layers, are of particular interest because of their possible functionalization. Applications range among others from optical materials1,2 up to biomaterials,3-8 including filtration,9 self-cleaning surfaces,10 and corrosion prevention.11 Functionalization of these films is usually obtained by the incorporation of active molecules or particles into the film architectures. This is of particular interest to obtain biologically active films. Indeed the first bioactive multilayers were obtained by such an approach. However as research on multilayers progresses, the film architectures * To whom correspondence should be addressed. Phone: 0033 3 88 41 40 01. Fax: 0033 3 88 41 40 99. † Institut National de la Sante ´ et de la Recherche Me´dicale. ‡ Institut Charles Sadron. (1) Fou, A. C.; Onitsuka, O.; Ferreira, M.; Rubner, M. F.; Hsieh, B. R. J. Appl. Phys. 1996, 79, 7501. (2) Eckle, M.; Decher, G. Nano Lett. 2001, 1, 45. (3) Vautier, D.; Karsten, V.; Egles, C.; Chluba, J.; Schaaf, P.; Voegel, J.-C.; Ogier, J. J. Biomater. Sci. Polym. Ed. 2002, 13, 713. (4) Jessel, N.; Atalar, F.; Lavalle, P.; Mutterer, J.; Decher, G.; Schaaf, P.; Voegel, J.-C.; Ogier, J. Adv. Mater. 2003, 15, 692. (5) Richert, L.; Boulmedais, F.; Lavalle, P.; Mutterer, J.; Ferreux, E.; Decher, G.; Schaaf, P.; Voegel, J. C.; Picart, C. Biomacromolecules 2004, 5, 284-294. (6) Serizawa, T.; Yamaguchi, M.; Matsuyama, T.; Akashi, M. Biomacromolecules 2000, 1, 306. (7) Mendelsohn, J. D.; Yang, S. Y.; Hiller, J.; Hochbaum, A. I.; Rubner, M. F. Biomacromolecules 2003, 4, 96. (8) Thierry, B.; Winnik, F. M.; Merhi, Y.; Silver, J.; Tabrizian, M. Biomacromolecules 2003, 4, 1564-1571. (9) Tieke, B.; van Ackern, F.; Krasemann, L.; Toutianoush, A. Eur. Phys. J. E. 2001, 5, 29. (10) Zhai, L.; Cebeci, F. C.; Cohen, R. E.; Rubner, M. F. Nano Lett. 2004, 4, 1349. (11) Fahrat, T. R.; Schlenoff, J. B. Electrochem. Solid State Lett. 2002, 5, B13-15.

become more sophisticated, and multi-compartment films were recently reported.12 Such films are based on two different multilayered film architectures: linearly and exponentially growing ones. The linearly growing ones are usually tight and do not allow for the diffusion of large molecules; they play the role of walls or barriers. On the other hand, the exponentially growing films, which can play the role of reservoirs, are less dense and allow for polyelectrolyte diffusion. Reservoirs can also be created by the embedding of intact vesicles into multilayers. Vesicles are however fragile objects and their deposition onto a surface usually leads to spontaneous disruption. A first approach toward the deposition of intact vesicles on surfaces of poly(vinyl sulfate)/poly(diallyldimethylammonium chloride) multilayers as well as their embedding in such multilayer architectures was reported by Katagiri et al.13,14 They used N,N-dihexadecyl-N-(3-triethoxysilyl)propylsuccinamide vesicles which were stabilized by a polymerized silica layer at the vesicle surface resulting from a spontaneous condensation of triethoxysilyl groups in water. In a recent paper,15 referred here as paper I, we presented an alternative route to adsorb and embed vesicles, made only of phospholipids, into polyelectrolyte multilayers. These vesicles are constituted by POPC (1-palmitoyl-2-oleoylsn-glycero-3-phosphatidylcholine), rigidified with cholesterol and coated with poly-(D-lysine) (PDL). The vesicles were incorporated in poly(L-glutamic acid)/poly(allylamine) (PGA/PAH) multilayers prepared by bringing the surface alternately in contact with these polyelectrolyte solutions. AFM imaging and quartz crystal microbalance experiments suggested that the vesicles can be deposited (12) Mendez Garza, J.; Schaaf, P.; Muller, S.; Ball, V.; Stoltz, J. F.; Voegel, J. C.; Lavalle, Ph. Langmuir 2004, 20, 7298. (13) Katagiri, K.; Hamasaki, R.; Ariga, K.; Kiguchi, J. Langmuir 2002, 18, 6709. (14) Katagiri, K.; Hamasaki, R.; Ariga, K.; Kiguchi, J. J. Am. Chem. Soc. 2002, 124, 7892. (15) Michel, M.; Vautier, D.; Voegel, J.-C.; Schaaf, P.; Ball, V. Langmuir 2004, 20, 4835.

10.1021/la050497w CCC: $30.25 © 2005 American Chemical Society Published on Web 07/12/2005

Integrity of the Vesicles

Langmuir, Vol. 21, No. 17, 2005 7855

Table 1. Size and Zeta Potentials (ζ) Measurements of Nonstabilized Vesicles (NSV) and Stabilized Vesicles (SV) with or without Encapsulated Fe(CN)64dilution 30-fold NSV NSV Fe(CN)64SV SV Fe(CN)64a

50-fold

100-fold

size (nm)

ζ (mV)

size (nm)

ζ (mV)

size (nm)

Ζ (mV)

215 220 240 240

-55.2+/-2.2 -58.4+/-3.8 53.4+/-3.6 51.6+/-6.2

200 220 225 230

-50.4+/2.8 -58.2+/-4.2 48.2+/-3.2 45.0+/-2.6

205 200 215 220

-50.8+/-2.3 -54.0+/-2.7 51.2+/-1.8 49.0+/-4.6

Measurements were realized at 30-, 50-, or 100-fold dilutions

in such a way and remain intact: they neither disrupt nor coalesce even in the dry state. This important aspect could, however, not be firmly proven. Most of the polyelectrolyte multilayers reported up to now were obtained by dipping the substrate alternately into the polyelectrolyte solutions, by spin-coating the solutions or simply by flowing the polyelectrolyte solutions alternately over the surface.16,17 The films reported in paper I were obtained in a similar way. Yet, Schlenoff18 showed that polyelectrolyte multilayers can also be obtained simply by spraying alternately the polyelectrolyte solutions onto the surface. We recently verified ourselves that spraying leads to very regular film constructions. Moreover, this method considerably speeds-up their buildup without affecting their “quality” as measured by X-ray reflection. In the present article, we will make use of this deposition technique to fabricate PGA/PAH films with embedded vesicles. All of the build-up steps including the vesicle deposition have being realized by spraying. By charging the vesicles with ferrocyanide ions, an electroactive compound and making use of cyclic voltammetry, we will prove that the vesicles are not disrupted during deposition despite the fact that they are also brought onto the substrate by spraying. Finally, we will also show that more than one vesicle layer, separated by several PGA/ PAH bilayers can be incorporated into multilayer architectures. This should open the route toward the construction of sophisticated multi-functionalized films which could find applications in particular in biologically active coatings. II. Material and Method Solutions and Chemicals. All of the experiments were performed in 10 mM. Tris (tris(hydroxymethyl)aminomethane, Gibco BRL) buffer at pH ) 7.4 with 0.015 M of NaCl (Prolabo, France). The electrochemical experiments were performed in the presence of ferrocyanide (Fe(CN)64- CAS 14459-95-1, Sigma, P9383, batch 033K0128) at a concentration of 1 mM in the TrisNaCl buffer. Lipids used to prepare the vesicles were all purchased from Sigma and used without any further purification. We employed 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphatidylcholine (POPC) [(L-R lecithin, type XVI-E from egg yolk, Sigma, ref P-3556, lot 22K5223], cholesterol [(5-cholesten-3-ol), Sigma, ref C-8667, lot 050K304] and L-R-phosphatidyl-D,L-glycerol, (POPG) [β-oleoylγ-palmitoyl C18:1[cis]-9/C16:0, Sigma, ref P-6956, lot 112K5210]. The polyelectrolytes used were poly(ethylene imine) (PEI, Lupasol WF, Mw ) 25 000 g mol-1, CAS 14459-95-1, Sigma, P9383, batch 033K0128), poly(allylamine hydrochloride) (PAH, Aldrich, cat number: 28,322-3, Mw ) 65 000 g mol-1), poly(L-glutamic acid) (PGA, Sigma, ref. P-4886, Mw ) 17 000 g mol-1), and poly(Dlysine) (PDL) (Sigma, P-0296, lot 082K5112, Mw ) 2000 g mol-1). (16) Chiarelli, P. A.; Johal, M. S.; Casson, J. L.; Roberts, J. B.; Robinson, J. M.; Wang, H.-L. Adv. Mater. 2001, 13, 1167. (17) Jiang, C.; Markutsia, S.; Tsukruk, V. V. Adv. Mater. 2004, 16, 157-161. (18) Schlenoff, J. B.; Dubas, S. T.; Fahrat, T. Langmuir 2000, 16, 9968.

Vesicles Preparation. Vesicles preparation has been described in detail in paper I and will thus only be presented shortly here. They were obtained by dissolving a mixture of 50 mg of POPC, 2.5 mg of POPG and 2.5 mg of cholesterol in 5 mL of chloroform. POPG was used to confer a negative charge to the vesicles and cholesterol to rigidify the membranes.The lipid solution was dried under a nitrogen flux to obtain a lipid film that was stored overnight under vacuum. This film was then hydrated by adding 10 mL of a buffer solution to obtain a turbid suspension that was subjected to 15 freeze and thaw cycles before 10 successive extrusions though a 0.22 µm Millex GV (Millipore) membrane. Vesicles, either modified or not with a PDL layer, containing 1 mM of encapsulated Fe(CN)64- ions were prepared in the same manner as the vesicles without Fe(CN)64- ions but with the difference that the electroactive species were added to the 10 mL of 10 mM Tris, 15 mM NaCl buffer used for lipid hydration. The non encapsulated Fe(CN)64- anions were then eliminated after the vesicle preparation by dialysis using a membrane with a molecular weight cut off of 1000 g mol-1 (Spectra/Por, Cellusose Ester, Bioblock). The vesicles, with or without encapsulated Fe(CN)64- ions were diluted 30, 50, or 100 times in the buffer before deposition onto a PGA ending multilayer for PDL stabilized vesicles and onto a PAH ending multilayer for the non stabilized vesicles. As will be described later on, all polyelectrolytes and vesicles were deposited onto solid substrates by a spraying method described elsewhere.19 To stabilize the vesicles with PDL, the solution of nonstabilized vesicles was slowly dropped into the same volume of a 0.5 mg mL-1 of PDL solution prepared in the same Tris buffer. Both, the PDL chain length (Mw ) 2000 g mol-1) and its concentration (0.5 mg mL-1) were chosen small enough to avoid significant vesicle bridging. Moreover, the ionic strength of the buffer (15 mM NaCl) was also chosen small enough to avoid a too strong screening of the repulsive electrostatic forces between identically charged vesicles. Because of the negative charge of the unmodified vesicles, PDL is assumed to adsorb onto the vesicles. Diameters and zeta potentials of the vesicles were measured at different dilutions, with and without adsorbed PDL, by means of dynamic light scattering (Zetasizer 3000 HS, Malvern Instruments) and laser Doppler electrophoresis, respectively. The obtained results are gathered in Table 1. Table 1 shows that 1 mM Fe(CN)64- ion solution encapsulation does not significantly modify the size and the zeta potential of the vesicles. Moreover, the size of the stabilized vesicles (SV) was comparable to their nonstabilized vesicle (NSV) counterparts. This clearly shows that the adsorption of PDL does not increase the mean vesicle size. Moreover, the sign of the zeta potential was inverted for the stabilized vesicles with respect to the non stabilized ones, meaning that the PDL adsorption effectively takes place. Adsorption Substrates. Silicon wafers (Polylabo, Strasbourg, France) were used for AFM and ellipsometry measurements. These wafers were cleaned with a 1:1 CH2Cl2:MeOH mixture and then with a 10 mM H2SO4 solution before spraying. For cyclic voltametry, films were directly sprayed on a gold electrode previously polished and submitted to 2000 restructuration cycles in a 0.5 M H2SO4 solution. The cleaning of the gold working electrode used for the cyclic voltametry experiments will be described in the next paragraph. (19) Izquierdo, A.; Voegel, J.-C.; Decher, G.; Schaaf, P. Langmuir submitted for publication.

7856

Langmuir, Vol. 21, No. 17, 2005

Build-Up of the Polyelectrolyte Multilayers by Spray. The deposition by spray19 was carried out by using air pump spray cans made of polypropylene and polyethylene purchased from Roth (ref: 0110.1. Size: height 217 X diameter 55 mm. Nozzle: 0.40 mm). A different spray can was used for each polyelectrolyte solution, as well as for each vesicle solution and the rinsing buffer solution (10 mM. Tris buffer in 15 mM. NaCl). Each can was filled with the solution to be sprayed and pressurized by pumping cycles in such a way that the spray rate remained approximatively constant over the entire film deposition process. The spray rate was of the order of 0.6 mL s-1. Spraying was carried out perpendicularly ((20°) to the substrate which was fixed in a vertical orientation ((15°). This allows drainage of the fluid along the surface.19 The build-up of a polyelectrolyte multilayer by spray was made as follows: each polyelectrolyte solution was sprayed for 3 s onto the surface, the substrate was left at rest during 3 s, rinsed during 3 s with the buffer solution, left at rest for an additional 3 s. To avoid the rupture of the vesicles during the build-up process, the vesicle solution was sprayed during 3 s, the substrate was left at rest during 3 s and the following polyelectrolyte solution was then directly sprayed onto this surface. Preliminary experiments showed that it was important to avoid this rinsing step of the vesicle layer to keep the vesicles intact. Previous experiments, reported in paper I, where the film was constructed by the conventional dipping method, showed that the adsorption of vesicles onto a multilayer is fast (only a few seconds are necessary). Hence, the spraying of the vesicles onto the substrate during 3 s was sufficient to obtain good surface coverage. We verified that longer spraying times of vesicle solutions does not change the final surface morphology determined by AFM. Experimental Methods. The atomic force microscopy and ellipsometry experiments were conducted in exactly the same way as in paper I. Films were always dried under a nitrogen stream before carrying out the experiment. For cyclic voltametry experiments we used a conventional three electrode set up (CHI604B from CH instruments, Texas) in which the working electrode was a gold electrode, the reference electrode and the counter electrode being a saturated calomel electrode (SCE) and a platinum foil, respectively. Before each experiment, the gold electrode was first cleaned with a SDS (10 mM) and HCl (1 M) solution each during 15 min and then placed under a UV/ozone light for 1 h. Just before start of an experiment, electrodes were rinsed with ethanol and distilled water. The working electrode was then subjected to 2000 restructuration cycles in a 0.5 M H2SO4 bath at a scan rate of 10 V s-1 between 0 and 1.6 V/SCE. These fast oxidation and reduction cycles produce oxydes on the gold surface.20,21 The electrochemistry cell was then rinsed with distilled water. The polyelectrolyte multilayer with or without embedded vesicles (containing Fe(CN)64-) was then sprayed onto the gold electrode as previously described. Finally, the coated gold electrode was introduced in the electrochemistry cell filled with Tris-NaCl buffer. The cyclic voltammograms were recorded between 0 and 0.5 V/SCE at a scan rate of 0.2 V s-1 at different time intervals to follow the release of encapsulated Fe(CN)64- in the polyelectrolyte multilayer. In a previous study we showed that ferrocyanide ions present in a PGA/PAH multilayer do not desorb in a pure buffer solution and their presence in the film can thus be directly detected.22

III. Results and Discussion (a) PGA/PAH Film Growth. Mass and thickness of PGA/PAH multilayers constructed by the conventional method where the 1 mg mL-1 polyelectrolyte solutions are alternately brought in contact with the adsorption substrate increase exponentially with the number of deposition steps.23 This implies that at least one of the two polyelectrolytes diffuses “in” and “out” of the whole (20) Oesch, U.; Janata, J. Electrochim. Acta 1983, 28, 1237. (21) Vela, M. E.; Salvarezza, R. C.; Arvia, A. J. Electrochim. Acta 1990, 35, 117. (22) Hu¨bsch, E.; Fleith, G.; Fatisson, J.; Labbe, P.; Voegel, J.-C.; Schaaf, P.; Ball, V. Langmuir 2005, 21, 3664. (23) Boulmedais, F.; Ball, V.; Schwinte, P.; Frisch, B.; Schaaf, P.; Voegel, J. C. Langmuir 2003, 19, 440.

Michel et al.

Figure 1. Increase of the film thickness for a PEI-(PGA/ PAH)20 multilayer obtained by the alternated spraying method. The full line in the inset corresponds to an exponential fit of the experimental data up to layer 24 (the best fit is d ) 6.0967 exp[(0.1465)n] where d (in nm) corresponds to the thickness and n corresponds to the number of layers).

film during each bilayer deposition step.24 The occurrence of such diffusion processes has been demonstrated by adsorbing fluorescently labeled polyelectrolytes at different stages of the build up and by using confocal laser scanning microscopy.24 It was shown that PGA molecules are the diffusing species in the present system.23 Figure 1 represents the evolution of the film thickness of a PGA/ PAH multilayer obtained by the alternated spraying method. The film thickness increases exponentially up to the deposition of 24 layers as it can be seen in the inset of Figure 1. The 130 nm observed for a film composed of 10 PGA/PAH bilayers (i.e., 20 layers) are much thinner than the about 1 µm thickness estimated for a (PGA/PAH)10 film made by the conventional dipping method.23 This thickness reduction might well be related to the drying of the film before determination of the thickness by means of ellipsometry. As the film build-up continues by the spraying method, the evolution of the film thickness becomes linear with the number of deposition steps. Such a behavior is expected. Indeed, each deposition process takes place over a limited time (3 s in our case). During each PGA spraying step, PGA diffuses into the film which is filled up to a given concentration of free PGA chains. However, when the film becomes thick enough, 3 s are no longer sufficient to fill all of the film up to this concentration. The amount of PGA chains that diffuse into the film during the PGA deposition step becomes then constant. It depends on the spraying time. The amount of free PGA chains that diffuses out of the film during the subsequent PAH spray becomes then also constant so that the film should grow linearly. A detailed understanding of this transition from exponential to linear growth is currently under investigation but is out of the scope of this article, having no influence on the vesicle deposition. We always deposited the vesicles on PGA/PAH films constituted by less than 3 bilayers [(PGA/PAH)2-PAH or (PGA/PAH)2]. (b) Vesicle Deposition. Vesicles identical to those employed in paper I were sprayed onto a PEI-(PGA/ PAH)2-PGA film. The diameter of the vesicles dissolved in solution was of the order of 215 nm, and they were stabilized with PDL. Directly after their deposition, we covered the vesicle layer by two other (PGA/PAH) bilayers. We found (see paragraph on cyclic voltametry experiments) that the embedding of the vesicle layer greatly enhances their stability. The films were then rinsed with (24) Lavalle, Ph.; Picart, C.; Mutterer, J.; Gergely, C.; Reiss, H.; Voegel, J. C.; Senger, B.; Schaaf, P. J. Phys. Chem. B 2004, 108, 635.

Integrity of the Vesicles

Figure 2. AFM images (in the tapping mode) obtained by spraying stabilized vesicles (SV) solutions at three different lipid concentrations. On a PEI-(PGA/PAH)2-PGA film, SV were deposited and covered by a (PGA/PAH)2 bilayer. Panel a, b, c, d: spraying of (SV) at 0.2 (a), 0.1 (b), or 0.06 mg mL-1(c) after PDL stabilization, respectively. (d) Spraying of nonstabilized vesicles (NSV) at a concentration of 0.1 mg mL-1 onto PEI-(PGA/PAH)2 and covered by a further (PGA/PAH)2.film.

distilled water, dried under a stream of nitrogen, and imaged by AFM with silicium nitride cantilevers (k ) 0.03 N/m, Park Scientific). Figure 2 shows typical AFM images obtained by spraying vesicle solutions at 3 different lipid (and thus vesicle) concentrations: 0.2, 0.1, and 0.06 mg mL-1. It clearly comes out that there exist an optimum in the lipid concentration which leads to a surface covered regularly by particles whose lateral sizes are typically of the order of 200-300 nm (see Figure 2b), a size of the same order as the diameter of native vesicles. This image is very similar to what was observed in paper I for films obtained by the classical method where the solution was flowed over the surface. At the higher lipid concentration (Figure 2a), the surface is clearly covered by particles but it appears rather irregular. This points toward vesicle disruption as it will be confirmed later by electrochemistry measurements. On the other hand, for low vesicle concentration, no vesicles are further visible on the surface. It thus seems that at low concentrations, the vesicles disrupt on the surface so that the PGA/PAH film should be covered by a lipid monolayer or bilayer. One can only speculate about the stabilization mechanism. It seems that isolated vesicles that arrive onto the surface are fragile objects that relax on the surface which leads naturally to their rupture. This process will probably take place at small lipid concentrations. The rupture could be the consequence of the propensity of the vesicles to increase their contact area with the polyelectrolyte multilayer. If, in the meanwhile, other vesicles adhere in their neighborhood, they can be stabilized simply by hindrance of their lateral extension. Such a stabilization must be optimal for a given vesicle density on the surface. If the flux of vesicles arriving on the surface is too large, the vesicles that enter in contact with the multilayer have no time to adapt before a new vesicle already enters in contact with it. This could then lead to vesicle rupture and/or fusion which seems to be the case at high vesicle concentration (Figure 2a). The importance of the vesicle stabilization by PDL is seen in Figure 2d where nonstabilized vesicles were sprayed onto the multilayer from a solution with optimal vesicle concentration. The structures

Langmuir, Vol. 21, No. 17, 2005 7857

observed by AFM are much larger than expected for single vesicles, confirming the observations made in paper I about the importance of the PDL stabilization. It must be stressed that the critical phospholipid concentration at which vesicles seem to remain intact during spraying is independent of the spraying time for spraying times larger than 3 s. It is the spraying rate (number of vesicles arriving onto the surface per unit time) that should be the crucial parameter. We confirmed by ellipsometry (see Figure 1 of Supporting Information) that the spraying of a PDL stabilized vesicle layer from a vesicle solution of optimal concentration leads to an increase of the film thickness by almost 280 nm. Such a value is of the order of 250 nm as expected from the vesicle size. It must be pointed out that we used a monolayer model with a value of 1.456 for the refractive index of the film in order to analyze the ellipsometric data. This provides an order of magnitude of the film thickness. A multilayer model would certainly be more appropriate to describe the present film with a refractive index for the vesicle layer of the order of 1.333. However, such a model contains several thicknesses that have to be determined and our experimental setup works at a constant angle and single wavelength. This allows us to essentially get the optical mass nd where n is the film refractive index and d its thickness. The estimated error of the film thickness by using a monolayer model is, however, only of the order of 20%, the vesicle layer being slightly thicker than measured. This vesicle layer thus appears slightly thicker than expected from the vesicle size. This could be due to the fact that the vesicles are constrained on the surface and are slightly squeezed without affecting their integrity. Moreover, one must also take into account that 5 additional PGA/PAH layers have been deposited on top of the vesicle layer before measurement. (c) Cyclic Voltammetry Experiments. To prove that during the vesicle deposition the vesicles are not disrupted, even only very briefly, we sprayed vesicles containing ferrocyanide, an electrochemically active molecule, and used cyclic voltammetry experiments as the detection method. It was shown in a previous paper that when a PGA/PAH multilayer is brought in contact with a ferrocyanide solution the Fe(CN)64- ions diffuse into the multilayer giving rise to a cyclic voltammetry signal.22 This electrochemical signal is only sensitive to the ferrocyanide ion concentration in the immediate neighborhood of the electrode onto which the multilayer is built. We first verified that a ferrocyanide solution sprayed onto a PEI-(PGA/PAH)5 multilayer leads, almost instantaneously, to an electrochemical signal (see Figure 2 of the Supporting Information). We then deposited a PEI-(PGA/ PAH)2-PGA-SV-(PGA/PAH)2 film onto the surface. For these experiments, the vesicles were dissolved at the optimal conditions defined in Figure 2, namely at 0.1 mg mL-1 We used vesicles containing ferrocyanide ions. Figure 3a shows the evolution of the electrochemical signal with time, and in Figure 3b, we plot the evolution of the maximum of the oxidation current that takes place at around 300 mV/SCE. One observes that the electrochemical signal does not vary after the vesicle deposition during the first hour and the current measured before spraying the vesicles corresponds to the capacitive current. This proves that the vesicles remain intact and do not open, even briefly, during their deposition and during the subsequent multilayer deposition. However, a slow increase of the electrochemical signal is observed after 1 h of deposition and takes place over a time scale of about 12 h (see insert of Figure 3b). This indicates that a slow release of ferrocyanide ions

7858

Langmuir, Vol. 21, No. 17, 2005

Michel et al.

Figure 4. Cyclic voltammetry experiments on a PEI-(PGA/ PAH)5 film that was sprayed on the gold electrode and subsequently immerged in a vesicles solution containing 1 mM Fe(CN)6.4- Doted line: after 5min in a vesicles solution; Dashed line: after 18 h.; Full line: 5 min after the vesicle suspension was submitted to sonication before spraying onto the polyelectrolyte multilayer.

Figure 3. Typical cyclic voltametry experiments on PEI-(PGA/ PAH)2-PGA-SV-(PGA/PAH)2 films sprayed onto the surface of the gold electrode. The vesicles contained ferrocyanide ions. (a) Evolution of the electrochemical signal with time (each signal corresponds to: from bottom to top 0h (before spraying the vesicles), 1, 6, and 12 h). (b) Plot of the evolution of the maximum of oxidation current observed at around 300mV/SCE for short contact time between the modified electrode and the buffer. Inset: evolution of the current intensity over a time period of about 23 h.

from the embedded vesicles takes place over such a longer time scale. This increase in current over time cannot be due to an increase of the effective area since the oxidation and reduction peaks of the gold electrode, produced during the electrode restructuration cycles (in the presence of 0.5 M H2SO420,21) do not evolve during the time when the potential scans are performed between 0 and 0.7 V/SCE and at a scan rate of 200 mV s-1 as in the present experiments. The area under the reduction peak is directly proportional to the effective area which remains constant. Is this release due to the embedding of the vesicles into the multilayer or does it also take place for vesicles dissolved in solution? To answer this question, we stored a solution of vesicles containing ferrocyanide for 18 h before bringing it in contact with a gold electrode coated with a PEI-(PGA/PAH)5 multilayer. No electrochemical signal could be observed (see Figure 4). We then used the same vesicle solution, and disrupted the vesicles by ultrasound waves and brought this solution in contact with a gold electrode coated with a similar multilayer. We observed a strong electrochemical signal after only 5 min of contact with the electrode (see Figure 4). This indicates that stabilized vesicles in solution do not leak. It is thus the embedding of vesicles into multilayers that apparently slightly affects their tightness with respect to ferrocyanide ions. To demonstrate the importance of the stabilization of the vesicles by PDL adsorption, we embedded nonstabilized vesicles that contained ferrocyanide ions into a similar multilayer as the one used previously. In this case we, observed immediately after film build-up a significant

Figure 5. Evolution of the thickness increase measured by ellipsometry of a PEI-(PGA/PAH)2-(PGA/SV)-(PGA/PAH)6(PGA/SV)-(PGA/PAH)4 multilayer obtained by the alternated spraying method. The different steps correspond to the deposition of the following parts: step I: PEI; step II: step I + (PGA/ PAH)2; step III: step II + (PGA/SV)-(PGA/PAH)2; step IV: step III + (PGA/PAH)2; step V: step IV + (PGA/PAH)2-(PGA/ SV)-(PGA/PAH)2; step VI: step V + (PGA/PAH)2.

electrochemical signal, confirming the fact that nonstabilized vesicles can disrupt during deposition on multilayers (see Figure 3 of the Supporting Information). Similarly we sprayed stabilized vesicles onto a PEI-(PGA/ PAH)2-PGA multilayer but left it uncovered. In this case too, we directly observed a strong electrochemical signal, indicating the stabilizing role of subsequent deposited (PGA/PAH) bilayers (see Figure 4 of the Supporting Information). (d) Architectures Containing Two Vesicle Layers. One of the goals of the present study was also to construct multilayer architectures containing several vesicle layers. It is thus important to verify if a film made of two vesicle layers can be constructed and in particular if the vesicles deposited in the second layer remain intact during deposition and if this deposition does not induce a rupture of vesicles of the first layer. We first used ellipsometry to follow the construction of a PEI-(PGA/PAH)2-PGA-SV(PGA/PAH)6-PGA-SV-(PGA/PAH)4 film entirely constructed with the spraying method. We always used the vesicle solution of optimal lipid concentration (0.1 mg mL-1). Figure 5 shows the evolution of the film thickness at different stages of the construction. The two thickness jumps of the order of 300 nm each, corresponding to the deposition of the two vesicle layers,

Integrity of the Vesicles

Langmuir, Vol. 21, No. 17, 2005 7859

concerned or if both vesicle layers are equally concerned. This however is out of the scope of this article and will be investigated in a future work.

Figure 6. Short time evolution of the oxidation current maximum (around 300 mV/SCE) with time for a PEI-(PGA/ PAH)2-(PGA/SV)-(PGA/PAH)6-(PGA/SV)-(PGA/PAH)4 film sprayed onto the surface. The vesicles contained ferrocyanide ions. Inset: evolution of the signal over a long time period (13 h).

suggest that the vesicles remain intact during their deposition. To confirm this hypothesis, we constructed a similar architecture but used vesicles containing ferrocyanide ions. The evolution of the cyclic voltametry signal was qualitatively similar to that observed for a single vesicle layer. Figure 6 represents the evolution of the intensity at the maximum of the peak at aroung 300 mV/ SCE. The absence of an electrochemical signal over the initial 40 min clearly indicates the stability of the vesicles sprayed in both layers. Again over longer time scales, one observes a slow release of ferrocyanide ions into the film (see inset of Figure 6). It seems that this release is slightly more rapid when two vesicle layers are incorporated compared to the incorporation of one vesicle layer, the characteristic time scale being of the order of 6 h compared to 12 h for one vesicle layer. This could mean that the vesicles are slightly less tight due to the presence of a second vesicle layer. Further studies need to be done to investigate if the number of PGA/PAH bilayers between the two vesicle layers affects this stability, if only the outer layer is

IV. Conclusion We have shown that phopholipid vesicles stabilized by the polycation PDL can be embedded into PGA/PAH polyelectrolyte multilayers without disruption. These vesicles were filled with ferrocyanide ions. However over longer time scales, of the order of a few hours, we still observed ferrocyanide leakage and this leakage is due to the vesicle embedding. Ferrocyanide ions are small compared to polypeptides or proteins. It is thus expected that leakage of such larger molecules could be much smaller or even absent. This would allow polyelectrolyte multilayers functionalization with, for example, biologically active compounds. Our next goals consist in investigating the release of such larger molecules from embedded stabilized vesicles. Further, by applying external stresses such as an electrical field and ultrasounds, we will try to open the vesicles in order to induce a release of their content at a specific moment. Polyelectrolyte multilayers containing embedded vesicles could become a major way to achieve specific multifunctionalizations. Acknowledgment. This work was supported by the ACI “Nanoscience” (NR204) from the Ministe`re Franc¸ ais De´le´gue´ a` la Recherche. Supporting Information Available: Evolution of the thickness increase of a PEI-(PGA/PAH)2-(PGA/SV)-(PGA/ PAH)9 multilayer obtained by the alternated spraying method (Figure 1). Cyclic voltamograms on the surface of a gold electrode covered with a PEI-(PGA/PAH)5 multilayer onto which a 1 mM Fe(CN)4- solution was sprayed (Figure 2). Cyclic voltamograms on the surface of a gold electrode onto which a PEI-(PGA/PAH)2(NSV/PAH)-(PGA/PAH)2 film was sprayed (Figure 3). Cyclic voltamograms of a gold electrode covered with a PEI-(PGA/ PAH)2-PGA film onto which stabilized vesicles (SV) were sprayed and kept uncovered (Figure 4). This material is available free of charge via the Internet at http://pubs.acs.org. LA050497W