Anomalous Thickness Evolution of Multilayer Films Made from Poly-l

Nov 23, 2006 - Bernard Senger,†,‡ and Vincent Ball*,†,‡. Institut National de la Sante´ et de la Recherche Me´dicale, INSERM Unite´ 595, 11...
0 downloads 0 Views 112KB Size
2602

Langmuir 2007, 23, 2602-2607

Anomalous Thickness Evolution of Multilayer Films Made from Poly-L-lysine and Mixtures of Hyaluronic Acid and Polystyrene Sulfonate Gregory Francius,†,‡ Joseph Hemmerle´,†,‡ Jean-Claude Voegel,†,‡ Pierre Schaaf,§ Bernard Senger,†,‡ and Vincent Ball*,†,‡ Institut National de la Sante´ et de la Recherche Me´ dicale, INSERM Unite´ 595, 11 rue Humann, 67085 Strasbourg Cedex, France, Faculte´ de Chirurgie Dentaire, UniVersite´ Louis Pasteur, 1 place de l’Hoˆ pital, 67000 Strasbourg, France, and Centre National de la Recherche Scientifique, Unite´ Propre de Recherche 22, Institut Charles Sadron, 6 rue Boussingault, 67083 Strasbourg Cedex, France ReceiVed October 4, 2006. In Final Form: NoVember 23, 2006 Using a mixture of polyanions or polycations offers a new way to control the properties of polyelectrolyte multilayer (PEM) films. The central issue of PEM films made from blended polyelectrolyte solutions is the relation between the properties of the blended architecture and the properties of the films made from each pure component. Two situations are possible: either (i) the properties of the blended films are intermediate between those corresponding to the single components or (ii) new effects may emerge leading, for instance, to improved mechanical properties. Situation (i) is expected when the chemical natures of both polyelectrolytes from the blended mixture are close, whereas situation (ii) is more probable when the polyelectrolytes from the blend are very different. In this study, we focus on the buildup of PEM films made by the alternate spray deposition of a polyanion blend [a mixture of polystyrene4-sulfonate (PSS) and hyaluronic acid (HA) in different mass fractions] and a polycation solution of poly-L-lysine (PLL). Whereas (HA-PLL) films exhibit a strong exponential growth with the number of deposition steps, the (PSS-PLL) system is only weakly exponential. We find that when the composition of the polyanion blend ranges from pure (HA-PLL) to pure (PSS-PLL), the films can always be constructed. However, the polyanion composition of the films is far from that of the polyanion solutions used for the buildup. One observes a strong preference for the incorporation of PSS over HA into the films. Moreover, the most striking feature is that the film thickness does not evolve monotonously with the polyanion solution composition but passes through a sharp minimum for a polyanion solution containing 90-95% HA. A possible mechanism for this peculiar finding is proposed.

Introduction films1-5

The deposition of polyelectrolyte multilayer (PEM) on charged surfaces is now a very popular and versatile method to control the properties of interfaces and is currently used for applications in electrophoresis,6 microelectronics,7,8 the design of optical materials,9,10 biomaterials science,11-16 drug deliv* To whom correspondence should be addressed. Telephone: +33 (0)3 90 24 32 58. Fax: +33 (0)3 90 24 33 79. E-mail: Vincent.ball@ medecine.u-strasbg.fr. † Institut National de la Sante ´ et de la Recherche Me´dicale, INSERM Unite´ 595. ‡ Universite ´ Louis Pasteur. § Institut Charles Sadron. (1) Decher, G.; Hong, J. D.; Schmitt, J. Thin Solid Films 1992, 210-211, 831. (2) Decher, G.; Hong, J. D. Makromol. Chem., Macromol. Symp. 1991, 46, 321. (3) Decher, G. Science 1997, 277, 1232. (4) Hammond, P. T. Curr. Opin. Colloid Interface Sci. 1999, 4, 430. (5) Scho¨nhoff, M. Curr. Opin. Colloid Interface Sci. 2003, 8, 86. (6) Graul, T. W.; Schlenoff, J. B. Anal. Chem. 1999, 71, 4007. (7) Fou, A. C.; Onitsuka, O.; Ferreira, M.; Rubner, M. F.; Hsieh, B. R. J. Appl. Phys. 1996, 79, 7501. (8) Eckle, M.; Decher, G. Nano Lett. 2001, 1, 45. (9) Nolte, A. J.; Rubner, M. F.; Cohen, R. E. Langmuir 2004, 20, 3304. (10) Lukkari, J.; Saloma¨ki, M.; Viinikanoja, A.; A ¨ a¨ritalo, T.; Paukkunen, J.; Kocharova, N.; Kankare, J. J. Am. Chem. Soc. 2001, 123, 6083. (11) Vautier, D.; Hemmerle´, J.; Vodouheˆ, C.; Koenig, G.; Richert, L.; Picart, C.; Voegel, J.-C.; Debry, C.; Chluba, J.; Ogier, J. Cell Motil. Cytoskelton 2003, 56, 147. (12) Vautier, D.; Karsten, V.; Egles, C.; Chluba, J.; Schaaf, P.; Voegel, J.-C.; Ogier, J. J. Biomater. Sci., Polym. Ed. 2002, 13, 713. (13) Serizawa, T.; Yamaguchi, M.; Matsuyama, T.; Akashi, M. Biomacromolecules 2000, 1, 306. (14) Mendelsohn, J. D.; Yang, S. Y.; Hiller, J.; Hochbaum, A. I.; Rubner, M. F. Biomacromolecules 2003, 4, 96.

ery,17,18 and the design of perm-selective membranes.19 The fine control of the physical properties of these PEM films, e.g., their thicknesses, their permeabilities, and their mechanical properties, is of the highest interest to finely adapt them to their applications. It has been shown that these properties can be modified by varying the physicochemical conditions used during the buildup, such as the ionic strength20 or the solution pH when at least one of the polyelectrolytes used is a weak polyelectrolyte.21 Other investigated parameters are the molecular mass,22 the linear charge density of the polyelectrolytes,23,24 and the temperature.25-27 Recently, it appeared that it is also possible to tailor film properties by using blended polyelectrolyte solutions to construct the PEM (15) Thierry, B.; Winnik, F. M.; Merhi, Y.; Silver, J.; Tabrizian, M. Biomacromolecules 2003, 4, 1564. (16) Etienne, O.; Gasnier, C.; Taddei, C.; Voegel, J.-C.; Aunis, D.; Schaaf, P.; Metz-Boutigue, M.-H.; Bolcato-Bellemin, A.-L.; Egles, C Biomaterials 2005, 26, 6704. (17) Jewell, C. M.; Zhang, J.; Fredin, N. J.; Lynn, D. M. J. Controlled Release 2005, 106, 214. (18) Thierry, B.; Kujawa, P.; Tkaczyk, C.; Winnik, F. M.; Bilodeau, L.; Tabrizian, M. J. Am. Chem. Soc. 2005, 127, 1626. (19) Tieke, B.; van Ackern, F.; Krasemann, L.; Toutianoush, A. Eur. Phys. J. E 2001, 5, 29. (20) Hoogeven, N. G.; Cohen Stuart, M. A.; Fleer, G. J. Langmuir 1996, 12, 3675. (21) Yoo, D.; Shiratori, S. S.; Rubner, M. F. Macromolecules 1998, 31, 4309. (22) Sui, Z.; Salloum, D.; Schlenoff, J. B. Langmuir 2003, 19, 2491. (23) Steitz, R.; Jaeger, W.; von Klitzing, R. Langmuir 2001, 17, 4471. (24) Kolarik, L.; Furlong, D. N.; Joy, H.; Struijk, C.; Rowe, R. Langmuir 1999, 15, 8265. (25) Bu¨scher, K.; Graf, K.; Ahrens, H.; Helm, C. A. Langmuir 2002, 18, 3585. (26) Tan, H. L.; McMurdo, M. J.; Pan, G.; Van Patten, P. G. Langmuir 2003, 19, 9311. (27) Saloma¨ki, M.; Vinokurov, I. A.; Kankare, J. Langmuir 2005, 21, 11232.

10.1021/la062910l CCC: $37.00 © 2007 American Chemical Society Published on Web 01/23/2007

Multilayers Made from PSS and HA Blends with PLL

films. These films are built up from mixtures of different (in the present work, only two) polyanions28-32 adsorbed alternatively with a given polycation or from mixtures of different polycations adsorbed in alternation with a given polyanion.33-35 In these blended solutions, the proportion of each polyanion (or polycation) in the mixture as well as the solution pH can be changed during the spraying or rinsing process after the film has been deposited. In many cases,28,30,31,35 but not in all cases,32 it appeared that one of the polyelectrolytes from the mixture was preferentially incorporated into the film; that is, its molar fraction was higher in the film than in the solution used to build up the film. For instance, we showed by means of Fourier transform infrared spectroscopy in the attenuated total reflection mode (FTIR-ATR) that, by using a mixture of poly-L-glutamic acid (PGA) and polyL-aspartic acid (PAsp) of close molecular masses, deposited in alternation with poly-L-lysine (PLL), the incorporation of PAsp was favored with respect to that of PGA.28 However, the average conformation of the film was closer to that of a pure (PGAPLL)n film (rich in β sheets) than that of a pure (PAsp-PLL)n film (rich in R helices). These findings were pretty surprising because the lateral chains of PGA and PAsp differ by only one methylene group and have only slightly different pKa values. Caruso et al. also found, in the case of solutions containing a mixture of poly-4-styrene sulfonate (PSS) and deoxyribonucleic acid (DNA), that PSS was preferentially incorporated over DNA when the alternated deposition with poly(allylamine) (PAH) was performed from solutions containing sodium chloride. However, the addition of ethanol in the adsorption medium favored the preferential incorporation of DNA.31 It has also been shown that when using a mixture of poly(4-vinylpyridine) (P4VP) and PAH in conjunction with poly(acrylic acid) (PAA) as the polyanion, P4VP is preferentially incorporated over PAH.35 These PEM films were cross-linked by means of carbodiimide chemistry before being subjected to a pH increase (up to 10); this pH modification triggered the release of P4VP from the PEM film, which was bound to the whole architecture mostly by means of hydrogen bonds. As a result of the quantitative P4VP release, it was possible to produce mesoporous films from the initially flat and compact films.35 Up to now, in all these studies, whether the preferential incorporation of one of the polyelectrolytes from the mixture occurred or not, it was found that the film thickness or the deposited mass per unit of surface area (as obtained from UV-vis spectroscopy) was a monotonous function of the mass (or molar) fraction of one of the polyelectrolyte mixture components. It is the aim of this study to report on the evolution of the film thickness, the surface morphology, and the internal composition of PEM films in which the polyanion solution is a mixture of hyaluronic acid (HA) and PSS, whereas the polycation solution is made from PLL, as a function of the mass fraction, x, of HA. We have already investigated the buildup of (HA-PLL)n36,37 and (PSS-PLL)n films.38 At an ionic strength of 0.15 M in NaCl (28) Debreczeny, M.; Ball, V.; Boulmedais, F.; Szalontai, B.; Voegel, J.-C.; Schaaf, P. J. Phys. Chem. B 2003, 107, 12734. (29) Hu¨bsch, E.; Ball, V.; Senger, B.; Decher, G.; Voegel, J.-C.; Schaaf, P. Langmuir 2004, 20, 1980. (30) Cho, J.; Quinn, J. F.; Caruso, F. J. Am. Chem. Soc. 2004, 126, 2270. (31) Quinn, J. F.; Yeo, J. C. C.; Caruso, F. Macromolecules 2004, 37, 6537. (32) Sun, J.; Wang, L.; Gao, J.; Wang, Z. J. Colloid Interface Sci. 2005, 287, 207. (33) Leporatti, S.; Gao, C.; Voigt, A.; Donath, E.; Mo¨hwald, H. Eur. Phys. J. E 2001, 5, 13. (34) Sui, Z.; Schlenoff, J. B. Langmuir 2003, 19, 7829. (35) Li, Q.; Quinn, J. F.; Caruso, F. AdV. Mater. 2005, 17, 2058. (36) Picart, C.; Mutterer, J.; Richert, L.; Luo, Y.; Prestwich, G. D.; Schaaf, P.; Voegel, J.-C.; Lavalle, P. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 12531. (37) Picart, C.; Lavalle, P.; Hubert, P.; Cuisinier, F. J. G.; Decher, G.; Schaaf, P.; Voegel, J.-C. Langmuir 2001, 17, 7414.

Langmuir, Vol. 23, No. 5, 2007 2603

and at pH 7.4, both combinations of polyelectrolytes lead to an exponential evolution of the film thickness with the number of deposited pairs of layers. However, the thickness reached after the deposition of n bilayers is the highest in the case of the HA-PLL combination.38 To construct the PEM films, we used the spraying deposition method,39-41 which offers the advantage of being fast: typically 1 min is needed to build one pair of layers instead of 10-20 min, which is usually needed with the dipping method. In addition, it has been shown that the morphology of the sprayed films is comparable to that obtained upon dipping.39 The obtained architectures from the HA-PSS mixtures will be denoted as PEI-[(HAx-PSS1-x)-PLL]n, where PEI is polyethyleneimine and acts as a primer layer and x denotes the HA mass fraction in the mixed solution. Throughout this study, the whole polyanion concentration will be kept constant at 0.5 mg/mL. The main result of the present investigations is that the ellipsometric thickness (obtained after spray deposition, film drying, and ellipsometry measurements) displays a minimum when the PEM films are built up from solutions with x ≈ 0.9. FTIR-ATR experiments show a preferential incorporation of PSS in the blended films. A possible mechanism for this anomalous thickness evolution will be proposed. Materials and Methods Substrates. Silicon wafers, used as substrates for spray deposition as well as for topographical characterization with AFM, cut in the form of 4 × 1 cm2 rectangular wafers (Wafernet, GmbH, Germany) were cleaned just before each spraying experiment within an ethanol bath, within a Hellmanex bath (at 2% v/v and at 70 °C for 0.5 h), rinsed with ultrapure water, immersed in a hot hydrochloric bath (1.0 M at 70 °C for 0.5 h), and finally intensively rinsed with ultrapure water. A trapezoidal zinc selenide (ZnSe) crystal was used as the substrate for FTIR-ATR measurements. It was cleaned by using a solution of Hellmanex at 2% (v/v) for 15 min, rinsed with ultrapure water, and put in contact with a 0.1 M HCl solution for 15 min. Finally, the ZnSe crystal was rinsed again with ultrapure water. Polyelectrolyte Solutions and Preparation of the Films. All the solutions were prepared by using ultrapure water (MilliQ-Plus system, Millipore, F ) 18.2 MΩ cm). The polyelectrolytes used for the PEM film buildup were polyethyleneimine (PEI, 750 kg/mol, Sigma), poly-L-lysine (PLL, viscosimetric molecular weight 28.2 kg/mol, Sigma, France), polystyrene sulfonate (PSS, 70 kg/mol, Sigma, France), and hyaluronan (HA, 400 kg/mol, Bioiberica, Spain). The HA solutions were always prepared the day before the spray deposition experiment and stored overnight at 4 °C to allow for a complete dissolution of the initially formed gel. We found that the nondialyzed HAx-PSS1-x mixtures, where x denotes the mass fraction of HA in the blended solutions containing polyanions at a total mass concentration of 0.5 mg/mL, displayed an almost linear variation of their pH which decreases from 8.9 down to 6.2 when x increases from 0 up to 1. The pH of the PLL solution (equal to 6.5) was identical to that of the NaCl solution used to dissolve it within the experimental uncertainty of 0.05 pH units. Hence, to check if the change in pH with x had an influence on the buildup regime, we first performed some experiments with all the polyelectrolytes (single components or mixed solutions) dissolved in a 20 mM Tris/0.15 M NaCl buffer with the pH adjusted to 7.4 (with HCl addition). This was sufficient to stabilize the pH of all (38) Tezcaner, A.; Hicks, D.; Boulmedais, F.; Sahel, J.; Schaaf, P.; Voegel, J.-C.; Lavalle, Ph. Biomacromolecules 2006, 7, 86. (39) Schlenoff, J. B.; Dubas, S. T.; Fahrat, T. R. Langmuir 2000, 16, 9968. (40) Izquierdo, A.; Ono, S. S.; Voegel, J.-C.; Schaaf, P.; Decher, G. Langmuir 2005, 21, 7558. (41) Porcel, C.; Lavalle, Ph.; Ball, V.; Decher, G.; Senger, B.; Voegel, J.-C.; Schaaf, P. Langmuir 2006, 22, 4376.

2604 Langmuir, Vol. 23, No. 5, 2007 polyanion mixtures within 0.05 pH units. The film thicknesses were the same, within an experimental accuracy of 10%, for the buildup performed from unbuffered or from buffered media (data not shown). This may originate from the fact that HA (with pKa ) 2.9) and PSS are fully ionized in the 6.2-8.9 pH range. All the data will hence be given for buffered solutions containing 20 mM Tris and 0.15 M NaCl. Each polyelectrolyte solution, either the HAx-PSS1-x mixture or the PLL solution, was then sprayed on the surface of the vertically held substrate from manually pressurized spray cans.40 The distance between the aperture of the spraying can and the substrate was fixed at 20 ( 3 cm, and the spraying time was held at 5 s. The substrate was maintained in a vertical setting ((10°) to allow for liquid drainage along its surface. Each “layer” of the PEM film was deposited by the following method: the polyanion mixture or the PLL solution was sprayed for 5 s, and the surface was then allowed to equilibrate for 15 s before spraying the rinsing buffer for 5 s followed by an additional equilibration for 15 s. The spraying cans were regularly pressurized by manual pumping to allow depositions at a nearly constant pressure. For the morphologic characterization of the obtained PEM films by AFM, we used the same films as those characterized by ellipsometry.41 For the calibration experiments performed by means of FTIRATR experiments, it was not possible to dissolve the high molecular mass HA at a high enough concentration without the formation of a viscous gel. Hence, to relate the area under the vibration peaks to the polyelectrolyte concentration, we used HA with a low molecular mass (6550 g/mol, Lifecore Biomedical, Chaska, MN). To that aim, we assume that the characteristic vibration frequencies and the molar absorption coefficients of HA are molecular mass independent. For the PEM films prepared by dipping, in the case of the FTIRATR experiments, all the polyelectrolytes were dissolved at a total concentration of 0.5 mg/mL in a 20 mM Tris/0.15 M NaCl solution. To build up PLL-[(HAx-PSS1-x)]n films (where n corresponds to the number of deposited pairs of layers and x represents the mass fraction of HA in the mixture containing a total polyanion concentration of 0.5 mg/mL), the HA and PSS solutions were mixed just before the beginning of the experiment. For all experiments, spray deposition or dipping deposition, the PEM film buildup was initiated by the deposition of a PEI primer layer. Indeed, we found that this was mandatory to observe a regular film buildup on the surface of the ZnSe crystal. In the case where the deposition was performed by spraying, the primer layer was nevertheless adsorbed by dipping the silicon wafer in a PEI solution for 5 min. For the infrared experiments, great care was taken to regularly refresh the solutions containing the polyanion mixtures to avoid depletion of the polyelectrolyte present at the smallest concentration. Atomic Force Microscopy (AFM). Atomic force images were obtained in the contact mode in air with the aid of a Multimode Nanoscope IV (Veeco, Santa Barbara, CA) on samples prepared by the spray deposition method on silicon wafers. The contact mode was used to have an overview on the sample morphology as well as for the estimation of the film thickness at different steps of the buildup. Before image acquisition, the imaged zone was scanned several times to ensure that the tip did not damage the PEM film. Before imaging, the films were needle scratched to have access to their thickness, with the flat glass substrate along the scratched line being the reference. For this study, we used cantilevers with a constant spring stiffness of 0.01 N m-1 (Veeco, Santa Barbara, CA). Ellipsometry. The thickness of the PEI-[(HAx-PSS1-x)-PLL]n films obtained by the spraying method was determined not only by AFM as explained above but also by ellipsometry (Jobin Yvon, model PZ200, France). The measurements were performed in the dry state after rapid nitrogen blowing to remove the aqueous solution on top of the film and to avoid the deposition of salt crystals. The measurements were performed at the wavelength of 632.8 nm (HeNe laser) and at an incidence angle of 70°. The real part of the film refractive index was fixed to 1.465, and its imaginary part was fixed to zero, whatever the composition of the polyanion mixture used to

Francius et al. build up the film. This could introduce a systematic bias in the thickness calculation, which was performed from the measured ellipsometric angles using a Fresnel model in which the PEM film was considered as homogeneous and isotropic. However, this bias should be minimal because the optical thicknesses obtained from the ellipsometry measurements matched closely with the profilometric heights obtained by AFM in the dry state. The optical thicknesses are the average values obtained from five measurements carried out at regularly spaced points along a line of the silicon wafer. To check the possible influence of the drying process needed to measure the optical thickness of the sprayed films on the multilayer buildup, we not only measured their optical thickness after the deposition of every second pair of layers but we also constructed films made of 14 pairs of layers without intermediate drying. The average thickness of these films was compared to that of PEM films that were also made from 14 pairs of layers but that underwent seven drying-rehydration steps. Infrared Spectroscopy in the Attenuated Total Reflection Mode. Fourier transform infrared spectra in the attenuated total reflection mode (FTIR-ATR spectra) were acquired on a trapezoidal ZnSe crystal during the multilayer buildup by adding 512 interferograms at 2 cm-1 spectral resolution. To this aim, we used a liquid nitrogen-cooled mercury cadmium detector on a Bruker Equinox 55 spectrometer (Bruker, Wissembourg, France). All the polyelectrolytes were dissolved at a concentration of 0.5 mg/mL in D2O (99.9% isotopic purity, Aldrich). We used D2O to study the evolution of the amide I band due to PLL (between 1600 and 1700 cm-1) as well as the evolution of the band due to the carboxylic groups of HA (at 1600-1610 cm-1) during the multilayer buildup without the interference of the strong water band at ∼1643 cm-1 (O-H bending mode). Indeed, the corresponding band in D2O (O-D bending mode) appears centered at 1209 cm-1 and becomes negative during the multilayer film buildup, meaning that some water is replaced by polymer chains in the region sensed by the evanescent wave on top of the ZnSe crystal. This crystal constituted the lower part of a 500 µL flow cell (Graseby Specac), which was connected to a peristaltic pump by means of Teflon tubings. The polyelectrolyte solutions as well as the rinsing buffer solutions were circulated for ∼10 min over the crystal surface. The infrared spectra were collected after each rinsing step and compared to the spectrum obtained after the deposition of the PEI precursor layer, which served as the reference spectrum to study the evolution of the absorbance due to the film during its LBL deposition. To estimate the relative concentration of HA and PSS in the PEM film and to compare these values with the solution concentrations (0.5x and 0.5(1-x) mg/mL for HA and PSS, respectively), we performed calibration experiments. The HA (of lower molecular weight than that used for the buildup experiments) and PSS were dissolved in the 20 mM Tris/150 mM NaCl buffer. These HA solutions used for calibration were flushed over a PEI-PSS film. The use of this negatively charged film deposited on the ZnSe crystal was necessary to avoid adsorption of HA or PSS on the surface of the bare crystal, which would result in a systematic overestimation of the peak heights or peak areas attributed to both polyelectrolytes.

Results and Discussion It has been previously demonstrated that (HA-PLL)n and (PSS-PLL)n multilayer films grow exponentially with n when the films are built up by dipping.36-38 It has been found that (HA-PLL)n films built up by spray deposition also grow exponentially for n values up to 5-6, before the occurrence of a transition to a linear growth regime.41 We demonstrate in this study that PSS-PLL films prepared by the spray method grow exponentially with the number of deposited bilayers (Figure 1). Hence, we expect that PEM films prepared from mixtures of HA and PSS will also grow exponentially (before the occurrence of an eventual transition to a linear growth regime) for all relative compositions in HA and in PSS. This is not as simple as expected; the growth regime of PEM films made from mixtures containing

Multilayers Made from PSS and HA Blends with PLL

Figure 1. Evolution of the film thickness obtained by the spray deposition method for HA-PSS mixtures of different weight fractions in HA as a function of the number, n, of deposited pairs of layers: x ) 0 (open circles), x ) 0.5 (closed squares), x ) 0.9 (open triangles), x ) 0.99 (closed diamonds), x ) 1 (dotted diamonds). The polyanion mixture was sprayed alternately with PLL from a 20 mM Tris/150 mM NaCl solution at pH 7.4. The data points correspond to the average of five thickness measurements performed on each PEM film, and the error bars correspond to the standard deviation of those five measurements.

a mass fraction x (with 0 e x e 1) of HA and (1-x) of PSS is exponential (Figure 1) for almost all compositions except for those with x values in a narrow composition range (0.90-0.95), where the growth regime appears virtually linear. This is surprising because the solution used for the deposition is rich in HA, so it was expected that the growth regime would be very close to that obtained for a pure HA polyanion solution. Alternatively, one could suspect preferential incorporation of PSS over HA (which indeed occurs, as will be demonstrated by means of FTIR-ATR spectroscopy). In this case, as long as there is enough PSS in the mixed solution, one would expect a buildup regime that would be very close to that of a pure (PSS-PLL) film. This is, however, not the case. For x values of ∼0.90-0.95, the thickness is neither close to that of a pure (HA-PLL) multilayer film nor close to that of a pure (PSS-PLL) multilayer film. It appears that the presence of even a small concentration of PSS in the mixed solution is sufficient to slow down the film growth even if HA is then available in a large excess. To stress this point, Figure 2 represents the evolution of the ellipsometric thickness of films containing a PEI primer layer and 12 additional sprayed bilayers. A pronounced minimum appears when the PEM films are made from solutions with x values close to 0.90-0.95. Figure 2 shows the evolution of the thickness of PEI-(PSSPLL)12 films (open triangles) as a function of the PSS concentration, upper abscissa scale, where PLL was always sprayed at a constant concentration of 0.5 mg/mL. Note that the PSS concentration increases from the right to the left to correlate these experiments with those performed with the (HAx-PSS1-x) mixtures. These experiments were performed to investigate the effect of reducing the PSS concentration on the film thickness in the absence of HA. More specifically, the ellipsometric thickness is about two times lower at x ) 0.90-0.95 than at x ) 0 for a PEI-(PSSPLL)12 film. We carefully checked that the regular dryingrehydration cycles needed before the ellipsometric thickness measurements did not significantly influence the results. We

Langmuir, Vol. 23, No. 5, 2007 2605

Figure 2. Evolution of the thickness of the PEI-[(HAx-PSS1-x)PLL]14 multilayer films as a function of x, lower abscissa scale, as measured by ellipsometry (open circles) and AFM in the dry state and in the contact mode after a needle scratch (closed circles). In one experiment (open squares), 12 bilayers from the polyanion mixture (x ) 0.71) were sprayed alternatively with PLL without intermediate drying stages. The whole concentration of the polyanion mixed solution was constant and equal to 0.5 mg/mL. Evolution of the thickness of PEI-(PSL-PLL)12 films (open triangles) as a function of the PSS concentration, upper abscissa scale, where PLL was always sprayed at a constant concentration of 0.5 mg/mL. Note that the PSS concentration increases from right to left to correlate these experiments with those performed with the HAx-PSS1-x mixtures. These experiments were performed to investigate the effect of reducing the PSS concentration on the film thickness in the absence of HA (see text for additional information).

also verified that the thicknesses measured by ellipsometry, using a constant refractive index of 1.465 for the film, were not affected by systematic errors. This is important because we expect the composition of the film to change with a change in the relative amount of polyanions in the mixed solution. Hence, the refractive index of the film should also change. To check this point, most of the silicon wafers used to perform the spray deposition were needle scratched and imaged by AFM, which allowed for the measurement of the film thicknesses and the observation of their respective morphologies as a function of x. It appears that the film thicknesses obtained by AFM in the dry state are very close to the ellipsometric thicknesses (closed circles in Figure 2) and that there is no systematic trend in the observed differences. In addition, we plotted the height profiles along the lines perpendicular to the scratch direction for all the investigated HA-PSS blended solutions. It appears that a continuous film, without pinholes reaching the silicon oxide substrate, is formed for all values of x (Figure 3). Before investigating the relative composition of HA and PSS inside the film as a function of their relative solution concentration, we have to determine whether the occurrence of a minimum in the thickness versus solution composition curve is not specifically related to the drying of the film before the thickness measurement and subsequent rehydration before the continuation of the spray deposition. To that aim, we deposited 14 pairs of layers from a blended solution (x ) 0.71) without intermediate dryingrehydration stages and the ellipsometric thickness in the dry state was measured only once. It appears that the obtained film has the same thickness, within the experimental uncertainty, as the films obtained with the intermediate drying-rehydration stages (Figure 2).

2606 Langmuir, Vol. 23, No. 5, 2007

Francius et al.

Figure 4. Evolution of the FTIR-ATR spectra of a PEI-[(HA0.9PSS0.1)-PLL]n film as a function of the wavenumber. Each spectrum corresponds to a given number of deposited bilayers, n, as indicated in the inset.

Figure 3. Height profiles obtained by AFM in the dry state and in the contact mode for samples prepared by spraying PEI-[(HAxPSS1-x)-PLL]14 onto a clean silicon wafer (also used for the ellipsometric thickness determinations in Figure 2).

Let us now investigate a possible mechanism for the observed anomalous thickness evolution with the solution composition for the PEI-[(HAx-PSS1-x)-PLL]n films. First, we wanted to have an insight into the film composition to check if one of the two constituting polyanions is preferentially incorporated in the film. To that aim, we performed in situ buildup experiments by the traditional dipping method because it was very difficult to spray films in a reproducible manner on ZnSe crystals. All the films, except the PEI-(HA-PLL)n films, contain PSS, as evidenced by the presence of its characteristic vibrational bands at 1007 and 1035 cm-1, and also HA (with a characteristic band attributed to the elongation of its carboxylates at 1600 cm-1).42 A typical example of ATR-FTIR spectra is given for a PEI[(HA0.9-PSS0.1)-PLL]n film in Figure 4. We have demonstrated that PEM films made from a mixture of HA and PSS, with the polycation being PLL, display a pronounced minimum in their thickness when the dipping solution contains ∼0.475 mg/mL HA (and hence 0.025 mg/mL PSS, corresponding to x ) 0.95). Therefore, it is interesting to give a quantitative estimate of the film composition in HA and PSS with respect to the solution composition. To this end, we performed calibration curves aimed at correlating the absorbance to the solution concentration in both HA and PSS. At a wavenumber of 1600 cm-1, corresponding to the stretching vibration of the HA carboxylates, we obtained an apparent extinction coefficient of (190 ( 10) M-1. This value is given in monomer unit concentrations. In the case of PSS, we obtained apparent extinction coefficients of (105 ( 10) and (141 ( 10) (42) Richert, L.; Boulmedais, F.; Lavalle, Ph.; Mutterer, J.; Ferreux, E.; Decher, G.; Schaaf, P.; Voegel, J.-C.; Picart, C. Biomacromolecules 2004, 5, 284.

M-1 at the wavelengths of 1007 and 1035 cm-1, respectively, which are characteristic of PSS. By measuring, after baseline subtraction, the absorbance of HA and PSS in the films, we calculated the molar concentrations of HA and PSS in the whole region sensed by the evanescent wave (∼600 nm at 1000 cm-1). It has to be noted that, for the experiments performed at x ) 0.5, 0.9, 0.95, and 0.97, the films made from 12 bilayers reached a thickness of