Tuning the Architecture of Cellulose Nanocrystal–Poly(allylamine

Article pubs.acs.org/Langmuir

Tuning the Architecture of Cellulose Nanocrystal−Poly(allylamine hydrochloride) Multilayered Thin Films: Influence of Dipping Parameters C. Moreau,*,† N. Beury,† N. Delorme,‡ and B. Cathala† †

UR1268 Biopolymères Interactions Assemblages, INRA, F-44316 Nantes, France Institut des Molécules et Matériaux du Mans − UMR 6283 − Université du Maine, le Mans, France

‡

ABSTRACT: Multilayered thin films consisting of alternating cationic polyelectrolyte, poly(allylamine hydrochloride) (PAH), and anionic cellulose nanocrystals (CNs) were constructed using the dipping procedure by screening different experimental parameters: the drying step between each layer adsorption, the dipping time, the ionic strength of the PAH solution, and the concentration of CNs dispersion. We showed that the drying process and the ionic strength of PAH solution were crucial parameters for the successful construction of 8bilayer films. Film thickness is mainly influenced by dipping time and CN concentration when using the dipping procedure without drying. Two architectures of adsorbed CN layersa single or a double layer of CNswere revealed on the basis of the thickness increment per bilayer, depending on experimental conditions. The layer adsorption process was investigated in realtime using quartz crystal microbalance with dissipation (QCM-D) experiments in an aqueous environment or by incorporating a drying step. On the basis of in situ construction of PAH−CN films in wet media, QCM-D data were indicative of highly hydrated films for which the progressive layer stacking is disturbed or prevented. QCM-D monitoring of CNs and PAH layer adsorption was monitored by incorporating a drying process. The impact of experimental parameters on PAH−CN multilayered construction and on CN layer configuration is discussed. This study offers new opportunities for tailoring the architecture of CNbased multilayer films.

■

obtained by a harsh hydrolysis treatment of fibers.22,23 Depending on the source, CNs are available in a wide variety of aspect ratios, e.g., 200 nm long and 5 nm wide, and up to several micrometers long and 15 nm wide, for cotton and tunicate, respectively.24 Another interest of hydrolysis is to confer negatively charged sulfate groups to the CN surface through the hydrolysis of cellulose, using concentrated sulfuric acid and leading to anionic sulfated CNs.24,25 On one hand, the presence of charges on the CN surface stabilizes the rod-like particles in solution that would otherwise agglomerate and then precipitate. On the other hand, the presence of negative charges promotes interactions with cationic polyelectrolytes and, consequently, their incorporation in PEMs from various cellulose sources with synthetic polyelectrolytes.24,26−28 Although LbL self-assembly is based on electrostatic attraction between oppositely charged species, other types of attractive interactions may take part in the sequential adsorption such as H-bonds, hydrophobic interactions as suggested from films obtained with biopolymer such as collagen12 or bionanoobject such as chitin microfibril.29 From the Jean et al. study on

INTRODUCTION Polyelectrolyte multilayer (PEM) thin films may be typically assembled using the layer-by-layer (LbL) technique through the electrostatic attraction between oppositely charged polyelectrolytes.1,2 In practice, PEMs can be obtained by alternating the dipping of solid charged substrates into diluted solutions of a polycation and a polyanion. The resulting self-assembled film is formed by the charge reversal on the film surface at each layer deposition, with final film thicknesses ranging from nanometers to micrometers. For a few decades now, PEMs have demonstrated their potential for many applications including biomedical field such as biocompatible3 or bioactive coatings4−6 or drug delivery7,8 as well as in material science for the fabrication of sensors9 or optically active surfaces.10,11 Due to increasing environmental concerns, this versatile, efficient, and easy-to-use process has been applied to a wide variety of biological structures derived from renewable resources in the past decade. PEMs based on biocomponents, including protein,12,13 mineral clays,14 and polysaccharides15−20 have been successfully incorporated through electrostatic attraction with charged synthetic polymers. Cellulose is the most widespread member of the polysaccharide family, and fabrication of PEMs from nanocrystals (CNs) was first reported by the Kotov’s group.21 CNs are the crystalline part of cellulose © 2012 American Chemical Society

Received: March 28, 2012 Revised: June 8, 2012 Published: June 10, 2012 10425

dx.doi.org/10.1021/la301293r | Langmuir 2012, 28, 10425−10436

Langmuir

Article

structure of CN layer adsorption depending on the experimental parameters used. This work also provides a deeper understanding on the adsorption mechanism of multilayers incorporating these CN bio-objects.

poly(allylamine hydrochloride)−CN (PAH−CN) multilayered films,30 one driving force could be the entropy gain that derives from the liberation of counterions and water molecules. Multilayered thin films based on CNs have also been reported by taking advantage of the interaction arising from the affinity between cellulose surface and xyloglucan.31−33 These assemblies are based on hydrogen and van der Waals bonds emphasizing the capacity of the CN surface to interact by other types of interactions than electrostatic ones. As in numerous studies on PEMs based on synthetic polyelectrolytes,34−40 the film characteristics (e.g., thickness, roughness, surface properties) and the architecture of LbL assembly incorporating CNs could be controlled by adjusting the processing parameters, such as the pH and ionic strength of polymer solutions or the nature of adsorbing species used for their construction. For instance, Wagberg et al.27 showed that film thicknesses can vary depending on the chemical structure of the polyelectrolyte and on the ionic strength of the adsorption media that have an impact on the conformation of adsorbed polyelectrolytes. Podsiadlo et al.21 suggest that the biological origin can also take a part in the merging of new properties as reported during the dipping fabrication of CNbased LbL films. These films show strong antireflective property with CNs originated from tunicate, while this effect did not exist when cellulose cotton was used as the CN source. This was attributed to the high aspect ratio of the tunicate CNs that enables the creation of the necessary porosity into the whole film to obtain the antireflective effect. Recently, we took advantage of the high affinity between cellulose and hemicellulose to constructed multilayer films with optical properties for the designing of enzymatic biosensors.33 All of these works demonstrated that incorporation of CNs in LbL assemblies offers news potential functionalities due to their unique physical (high aspect ratio, rigid rod-like nanoparticles) and environmental properties for the design of innovative biobased nanomaterials. Thus, manipulating these interactions through appropriate conditions of the LbL process method and parameters of polymer solutions will provide a fine control of the PEM architecture and properties. Even so, all these elements strongly suggest that the understanding of the influence of processing parameters is a key point for the accurate control of the final architecture and functionalities of CN multilayered films, and is thus required to develop innovative applications based on these attractive nano-objects. To date, studies reported in the literature on polyelectrolyte− CN multilayered films do not depict systematic examination of the different processing parameters. Thus, this report aims to address the influence of the adsorption dipping time, the occurrence of a drying step between each adsorption step, the ionic strength of the PAH solution, and the concentration of CN suspension. We demonstrated that the application of a drying step has a more striking effect than the other parameters on overall film construction and that the ionic strength of the PAH solution influences the molecular architecture of the films, especially the configuration of the CN layers when they are deposited. The effects of layer hydration and the ionic strength of the PAH solution were also explored using the quartz crystal microbalance with dissipation (QCM-D) technique and atomic force microscopy (AFM) analyses of film surfaces. CN adsorption was monitored in situ by using the QCM-D in aqueous medium and by introducing a drying process during the QCM experiment. We showed that PAH−CN multilayers can be easily and quickly constructed with a final controlled

■

EXPERIMENTAL SECTION

Materials. PAH (Mw = 120−200 000 g/mol, Sigma Aldrich) was used as a polycation. CNs were prepared as previously described30,33 according to the method of by Revol et al.41 Briefly, cotton linters were hydrolyzed with 65 wt % H2SO4 at 65 °C for 35 min. The suspension was washed by centrifugation and dialyzed again with distilled water until neutrality, after which mixed-bed research-grade resin (Sigma TMD-8) was added to the suspension for 48 h. Aqueous suspension of CN with a charge density of 0.54 e/nm2 was used at 5, 20, and 35 g/L. The average crystal dimensions (length and cross-section) were 100− 200 nm × 6−8 nm. Multilayer Film Preparation. Multilayer films were deposited on silicon wafers as solid substrate, cleaned beforehand for 30 min in a mixture of H2O2/H2SO4 (70%/30% v/v), and then thoroughly rinsed in Millipore water and finally dried under a nitrogen stream. PAH−CN films were built on the basis of the conventional LbL dipping method.1,42 The silicon wafers were alternately immersed in the respective solutions of PAH (4 g/L, pH 5.0, containing 0 or 1 M NaCl) and CN suspension (5, 20, and 35 g/L) for 1, 10, and 20 min. After each immersion step, the substrates were rinsed in pure water (three manual baths) and, the surfaces were then dried under a nitrogen stream. All films were dried under nitrogen flow when the final construction was completed. The dipping sequence was repeated until an eight-bilayer film was formed, with one bilayer defined as a single deposit of a PAH and CN layer. Mechanical Profilometry. Film thickness was measured by mechanical profilometry (Dektak 8, Veeco). The scratched surfaces were scanned along multiple straight lines using a 2.5-μm radius hemispherical tip carbide stylus and a contact force of 5 mg. Accuracy was ensured by averaging six to eight measurements on each film. QCM-D Measurement. The adsorption process of PAH and CN were investigated using a QCM-D (Q-Sense E4 instrument, AB, Sweden), with gold-coated quartz crystals (QSX301, Q-Sense). Briefly, in the QCM experiment, the crystal was excited so that it would oscillate at its fundamental resonance frequency ( f 0) through a driving voltage applied across the gold electrodes. Any material adsorbed (desorbed) on the crystal surface induces a decrease (increase) of the resonance frequency (Δf = f − f 0).43,44 If the adsorbed mass is evenly distributed, rigidly attached, and small compared to the mass of the crystal, Δf is directly related to the adsorbed mass per unit area (mg m−2), Δm, by the Sauerbrey equation:45 Δm = − C

Δf n

(1) −2

−1

where C (= 17.7 ng cm Hz at f 0 ≈ 5 MHz) is the constant for the mass sensitivity, and n is the overtone number. With QCM-D, the dissipation factor, ΔD, is simultaneously monitored, corresponding to frictional losses due to viscoelastic properties of the adsorbed layer.43 High dissipation values reflect a thick and loosely adsorbed layer, while a thin and rigid layer vibrates with the crystal, indicating a low dissipation factor. For adsorbed films in aqueous medium where the dissipation factor is generally large, the Sauerbrey relation is not valid, and the use of the Voigt model is required to determine the hydrodynamic mass of the adsorbed layers.46 All measurements were carried out at 20 °C using the QCM flow cell modules or the open QCM cell module by monitoring changes in Δf and ΔD at 5, 15, 25, 35, 45, and 55 MHz (overtones n = 3, 5, 7, 9, and 11, respectively) and were performed in duplicate. For experiments in the flow mode, PAH and CN solutions were continuously injected over the quartz crystal surface at 100 μL/min. For experiments done with the open cell, multilayer films of 3 and 3.5 bilayers were assembled beforehand on quartz crystal surfaces before mounting them on the open cell module. Solutions were then 10426

dx.doi.org/10.1021/la301293r | Langmuir 2012, 28, 10425−10436

Langmuir

Article

Table 1. Thickness (nm) Values for Multilayered (PAH−CN)8 Films and the Corresponding Thickness Increment Per Bilayer, Depending on Experimental Conditions: without or with Drying between Each Layer Adsorption Using Different Adsorption Dipping Times (20, 10, and 1 min)a

a The ionic strength of the PAH solution (4 g/L) was either 0 or 1M NaCl, and different CN concentrations were used (5, 20, or 35 g/L). nd: film thickness not determined.

manually added and removed by suction from the quartz crystal surface, and the film was dried under an air stream. For both experimental setups, the quartz crystal surface was first allowed to swell in pure water to establish a stable baseline (Δf and ΔD are kept constant), and water rinsing took place each time a polymer was deposited. PAH solution at 0.1 g/L and CN suspension at 0.5 g/L were used for dipped films under the dipping conditions described above (5 min adsorption, rinsing, and drying steps) and for injected and manual deposition of solutions. Salt-free and salt-added (25 mM NaCl) PAH solutions and a 25 mM NaCl aqueous solution were used. Frequency values with standard deviations were established after two to three repetitions of the experiments. AFM. AFM measurements were made on films dipped on silicon wafers with an Agilent 5500 AFM at IMMM (Le Mans). All topography images were obtained in tapping mode using the same PPP-NCHR-W tip (nanosensor; spring constant: ∼40 N/m; frequency: 297.5 kHz). Image processing, coverage determination, and roughness analysis were performed with Gwyddion freeware.

dispersion followed by identical rinsing steps. Four different parameters that can influence this film assembly were investigated: • The presence or lack of a drying step between each CN or PAH layer adsorption, i.e., after the rinsing step of the dipping procedure. • The adsorption dipping time during the stage (i or iii), including a very short one (1 min) and two longer ones (10 and 20 min). • The ionic strength of the PAH solution (0 or 1 M NaCl). • The CN concentrations (5, 20, and 35 g/L) The examination of all these conditions was first performed by determining the thickness of films consisting of eight bilayers, (PAH−CN)8, as summarized in Table 1. When film growth succeeded up to eight bilayers, the average thickness deposited per bilayer (bold values in Table 1) was then estimated from the corresponding total thickness value to provide a comparison between film growth from the samples studied here and those from the literature. As reported in Table 1, film construction up to eight bilayers is effective only under some define conditions that will be discussed in detail in the next subsections; otherwise, film construction failed since adsorbed bilayers became unstable and finally broke away from the solid substrate before the eight-bilayer deposition (i.e., any layer

■

RESULTS Multilayered films consisting of alternating layers of PAH and CN polymers were constructed using the LbL dipping procedure as follows: (i) a charged substrate of a silicon wafer is first immersed in the polycationic PAH solution; (ii) the deposited layer is then rinsed in water to remove all loosely bound polymers and to prevent any polymer contamination during (iii) the subsequent immersion in the anionic CNs 10427

dx.doi.org/10.1021/la301293r | Langmuir 2012, 28, 10425−10436

Langmuir

Article

deposition can be detected). When film construction was completed, total thickness values considerably varied from approximately 50−70 nm to 120−160 nm, corresponding to thickness increments per bilayer ranging roughly from about 7−9 nm and 14−19 nm, respectively, highlighting the significance of the tested parameters. Furthermore, under these conditions, colored films are observed due to the semireflective effect of these multilayered structures.26,27,33 The visual homogeneity of the film color is also a criterion for assessing film growth. Effect of the drying step. As illustrated in Figure 1 for (PAH−CN20)8 films, a drastic change in the appearance of the

importance in accordance with previous works dedicated to PEM thin films.8,47 Effect of Ionic Strength of the PAH Solution. If electrostatic interactions are a driving force of polyelectrolyte self-assembly, ionic strength is therefore a key parameter since the charge screening along the polymer chain may occur and accordingly modify the polymer conformation in solution and upon adsorption onto solid substrate. Many reports have focused on the ionic strength effect on polyelectrolyte PEM construction.27,28,35 The formation of thicker films is usually achieved using polymer solutions with increased ionic strength due to lower charge repulsion between the polymer chains, leading to a more coiled conformation of the polyelectrolyte. However, this approach is somewhat limited to a certain extent, as adding salt at high concentration makes a poorer solvent for polyelectrolyte solutions, and polymer aggregation or total disruption may limit the growth of the film. Since CNs bear a very low amount of charge (charge density of 0.54 e/nm2 in this study), the addition of salt in the CN suspension has to be prohibited since even salt addition at low concentration induces their precipitation. Accordingly, for the CN suspension, we performed all the adsorption steps in pure water with the ionic strength of the PAH solution fixed at 0 (PAH0) or 1 M NaCl (PAH1). The thickness values presented in Table 1 demonstrate a real effect of the ionic strength of the PAH solution. Indeed, when the drying process is not applied between each adsorption step, film construction from the PAH0 solution was achieved with a sufficient adsorption time (10−20 min), whereas it failed at high ionic strength under all of the conditions tested. In contrast, with drying, films were successfully constructed even with short adsorption times and high ionic strengths. Moreover, in the latter case, (PAH1−CN)8 film thicknesses are greater than those of films built using a PAH0 solution. As an example, the measured thickness of (PAH−CN20)8 films is nearly 2-fold higher with PAH1 (132.5 nm) than with PAH0 (62.3 nm). Under these conditions in which films are built, the thickness increments per bilayer range from 7 to 9 nm when the PAH solution is devoid of salt (PAH0), and from 15 to 20 nm when the ionic strength is increased in the PAH solution (PAH1). Considering that the lateral dimensions of the cotton CNs used in this study were 6−8 nm (with a PAH layer thickness ≤1 nm42), this would mean that either a single layer of CNs, when the PAH solutions do not contain any salt, or a double layer of CNs (∼120−160 nm for eight bilayers) when the PAH solutions contain high levels of NaCl, is adsorbed at each deposition step. PAH solution with an intermediate NaCl concentration (0.5M) has been used to construct (PAH−CN)8 multilayers films with drying steps and 1 min adsorption time. No significant difference in film thickness between 0.5 M (116.4 nm at CN5, 138.9 nm at CN20 and 152.9 nm at CN35) and 1 M NaCl was observed, revealing that charge screening by salt ions has a negligible effect on the double-layer construction. This would means that, beside electrostatic interactions between these two charged species, other types of interactions should be take part in the building and the stability of the films. Apart from the salt effect, the total thickness of (PAH−CN)8 films without drying appears to be mainly influenced by the concentration of the CNs and, to a lesser extent, by the time of adsorption. Influence of CN Suspension Concentrations. It can be observed that when film construction was successful, the

Figure 1. Photos of (PAH−CN20)8 films from (a−d) PAH0 (0 M NaCl) and (e−h) PAH1 (1 M NaCl) solutions constructed from dipping at (c,g) 1 min, (a,e) 10 min, and (b,d,f,h) 20 min adsorption dipping times, without or with a drying process.

film is observed, whether or not the films are dried between each adsorption step. Indeed, when a drying step was applied, highly homogeneous films were obtained in all cases with the appearance of intense colors, even for the lowest CN concentration and adsorption time conditions. On the contrary, without the drying step, results are more contrasted, and effective film construction seems to be highly dependent on the other parameters examined, i.e., dipping time, ionic strength of the PAH solution, and CN concentration (Table 1). In the case of a short adsorption time (1 min) and/or at high ionic strength of the PAH solution, no films were obtained when drying was omitted, regardless of the concentration of the CN suspension. Under these conditions, a deposit was observed after five to six dipping steps. All of the material was then fully removed with the subsequent washing steps and film growth failed, suggesting an effect of the substrate instead of a highly cohesive interaction between the CN and PAH layers. When film growth takes place without the drying step, film construction is only obtained with a long adsorption time (10 and 20 min) and without the addition of ionic strength to the PAH solution. No representative or reliable thickness can be measured from highly inhomogeneous and bumpy surfaces such as those observed in Figure 1(e,f). Moreover, film colors are rather nonuniform over the entire surface compared to the corresponding film surface with drying steps (e.g., comparison of photos b and d at 20 min adsorption time; Figure 1). Thus, the drying process appears to be a parameter of prime 10428

dx.doi.org/10.1021/la301293r | Langmuir 2012, 28, 10425−10436

Langmuir

Article

Figure 2. Changes in (a,b) normalized frequency, Δf n/n, and (c,d) dissipation, ΔDn/n (×10−6) signals (at n = 3, 5 and 7) monitored as a function of time from alternated deposition of PAH and CN layers from 0.5 g/L CN suspension and 0.1 g/L PAH solution (a,c) without or (b,d) with 25 mM NaCl.

the adsorption process, we evaluated three adsorption times, including a very fast one. Results reported in Table 1 indicate that dipping time does not influence the formation and the total thickness of the films when a drying step is applied, confirming Kotov’s results on the building up of CN-based multilayered film with antireflective properties.21 In the case where the films were kept wet, the situation is even more contrasted. Our film construction was possible only when dipping time was increased, i.e., 10 and 20 min, and when the PAH solution was devoid of salt. In this case, blue colored films were obtained. Nevertheless, as illustrated in Figure 1(a,b), these films are not as homogeneous over the entire surface as those obtained with a drying step. When salt is present, no films are formed. Consequently, without drying, our results support the hypothesis that the CN adsorption process is diffusion-limited, and that the stabilization of the films requires internal rearrangements that could be enhanced by applying a drying step between each layer adsorption. To summarize these results, the drying step and the ionic strength of the PAH solution appear to be of prime importance for the control of PAH−CN film construction, whereas the effect of adsorption time only occurs when no drying is applied. When multilayer films are successfully constructed, the parameters of polymer solutions, i.e., the ionic strength of the PAH solution or, to a lesser extent, CN concentration, influence the final thickness of the films and, consequently, their final architecture. Therefore, in order to develop a better understanding of the effect of hydration and salt content on film construction, the layer adsorption process was investigated in situ by QCM-D.

increase of the CN concentration from 5 to 35 g/L induced the formation of thicker films. This effect was much more obvious when the drying step was omitted. In this case, the average increased thickness per bilayer changed from 7 to 9 nm, suggesting the deposition of a single CN layer, to 15−19 nm, corresponding perhaps to the deposit of a double layer of CN. However, the intermediate value obtained for the 20 g/L concentration (11.75 and 14.6 nm) suggests that the architecture of the film can be more complex than a single or double layer of the CN separated by one PAH layer. By comparison, when a drying step is applied, the increase of the CN concentration only has a weak effect. The increase of film thickness with increasing CN concentration is still observable, but the overall behavior of the film growth as measured by the thickness increment does not seem to be affected at a given NaCl concentration. However, the fine internal architecture of these films might be affected, and this will be the subject of a future study. Effect of Adsorption Time. In most studies dealing with the formation of PEMs, adsorption time is empirically fixed, and a time range from 10 to 30 min is commonly applied. However, when these parameters are questioned, it is supposed that adsorption occurs within a very short time. Adsorption of polyelectrolytes is usually described by a two-step process where the first adsorption step on the substrate, corresponding to a diffusion process, is very fast, while the second one may span a few minutes and is attributed to a rearrangement process of polymer chains at the surface.47−49 Similar trends can be expected for CN despite the different dynamic characteristics of these rigid nanoparticles compared to flexible segments of polyelectrolytes. In order to acquire a better understanding of 10429

dx.doi.org/10.1021/la301293r | Langmuir 2012, 28, 10425−10436

Langmuir

Article

Figure 3. Variation of (a) the normalized frequency (Δ(−Δf/n)) (filled symbols) and (b) dissipation (ΔD × 10−6) (open symbols) as a function of the number of CN layers for the PAH−CN films in the absence of salt and never dried during construction. Signals originate from the frequency recording of the third (squares), fifth (circles), and seventh (triangles) harmonics.

adsorption steps, leading to a more rigid film. Nevertheless, as revealed by the nonconvergent signals from the different overtones and high dissipation values, it is more likely that viscous film is formed, at least for the first layers. In this hydrated procedure, two explanations can be proposed to explain this QCM-D profile: a smaller amount of adsorbed CNs with associated water hydration occurred and/or additional hydrated layers could not fully couple to the quartz crystal surface. The first suggestion would be in accordance with observations of inhomogeneous dry films from the dipping experiments and suggests an irregular stacking of the eight bilayers. Without salt, despite the fact that the adsorption process seems to take approximately 5 min (i.e., corresponding to the frequency decrease), the continuous flow of polymer solutions for 10 min over the surface did not reveal any additional organization of the deposited layers in terms of steady-state frequency and dissipation signals. Furthermore, the final film construction remained constant throughout the night (>12 h) in an aqueous environment. When the same experiment was carried out with a PAH solution containing 0.25 mM NaCl (Figure 2b,d), film construction consisted of the deposition of up to four to five bilayers, as indicated by the decrease of the frequency signal and the increase of the dissipation signal. It should be mentioned that frequency and dissipation values were higher than those observed from films built with salt-free PAH solution (Figure 2a,c). Furthermore, the frequency decrease for the first three to four bilayers is very irregular compared to that of the film without salt. After the deposition of four bilayers, no additional frequency decrease was observed before increasing at the sixth bilayer. In addition, ΔD gradually increased as more material was deposited. This QCM-D profile obviously indicated that even though polymers could be adsorbed for the first four to five bilayers, a much softer film was obtained with a large water content, as revealed by the high dissipation values. After that, considerable film swelling occurred with material desorption. These experiments are in accordance with observations of unstable hydrogel-like films during the dipping procedure at high ionic strength of the PAH solution and low CN concentration. The stability of the film in a wet environment is thus weakened and affected by the presence of salt in the PAH solution. QCMD Monitoring of Layer Adsorption with a Drying Step: Ionic Strength Effect. In order to gain insight into the impact of the drying process on the subsequent layer deposition

QCMD Monitoring of Film Construction in an Aqueous Environment. The PAH solution and CN suspension were sequentially injected onto the quartz crystal surface in a continuous flow, allowing polymer adsorption to reach equilibrium, i.e., when the stabilization of the frequency (Δf) and dissipation (ΔD) signals was obtained. Polymer solution was replaced with pure water between each polymer adsorption step in order to remove all polymers that were loosely bound to the surface. Changes in normalized frequency and dissipation signals as a function of time with increasing ionic strength, 0 (PAH0) or 25 mM NaCl (PAH025) of the PAH solution, are provided in Figure 2a,c and b,d, respectively. As can be seen from Figure 2a, the frequency signal progressively decreased at each polymer injection, meaning that material adsorption (or mass increase) occurred on the quartz crystal up to eight bilayers. The frequency decrease is simultaneously accompanied by an increase of the dissipation signal (Figure 2c). No desorption was observed for PAH and CNs during the water rinsing steps. These signal variations were mainly determined by the CN adsorption process since PAH adsorption induces very low changes in frequency and dissipation signals, compared to those of CN adsorption. Figure 3 shows the corresponding signal increments from normalized Δf and ΔD changes at each CN adsorption step for the third, fifth, and seventh harmonics. Surprisingly, the changes in frequency at each step were not constant throughout the deposition process (Figure 3a). Initially, frequency increments increase up to five bilayers and then start to decrease, suggesting that less material is adsorbed from the crystal surface. By taking the fact that the mass sensed by QCM-D includes the mass of the adsorbed polymer and the hydration water (the “wet” mass) into account, the increase in mass observed for the first five deposits is linked to CN adsorption with associated water molecules that may be entrapped in the film or associated with polymer chains during construction. This is in accordance with the increase of the ΔD increment (Figure 2d and Figure 3b) that is indicative of viscous film, probably due to film swelling through water incorporation. After five to six deposition steps, frequency increments decrease, even though, as seen from Figure 2a, the frequency signal continues to decrease. This would indicate that a smaller amount of CNs is adsorbed on the surface and/or that water is ejected from the film. Concomitantly, the stabilization of the dissipation signal at a plateau value after the deposition of five to six CN layers could suggest that all water molecules entrapped in the film are released during the subsequent 10430

dx.doi.org/10.1021/la301293r | Langmuir 2012, 28, 10425−10436

Langmuir

Article

factor does not appreciably change over the process (