Journal of Colloid and Interface Science 366 (2012) 96–104

Contents lists available at SciVerse ScienceDirect

Journal of Colloid and Interface Science

www.elsevier .com/locate / jc is

Influence of the nature of the polycation on the adsorption kineticsand on exchange processes in polyelectrolyte multilayer films

Vincent Ball ⇑, Abdelghani Laachachi, Valérie Toniazzo, David RuchAdvanced Materials and Structures, Centre de Recherche Public Henri Tudor, 66 rue de Luxembourg, L-4002 Esch-sur-Alzette, Luxembourg

a r t i c l e i n f o

Article history:Received 15 July 2011Accepted 19 September 2011Available online 29 September 2011

Keywords:Polyelectrolyte multilayer filmsExchange processesDeposition kinetics

0021-9797/$ - see front matter � 2011 Elsevier Inc. Adoi:10.1016/j.jcis.2011.09.045

⇑ Corresponding author.E-mail address: vincent.ball@tudor.lu (V. Ball).

a b s t r a c t

The deposition of polyelectrolyte multilayer films (PEMs) appears more and more as a versatile tool tofunctionalize a broad range of materials with coatings having controlled thicknesses and properties. Toincrease the control over the properties of such coatings, a good knowledge of their deposition mecha-nism is required. Since Cohen Stuart et al. (Langmuir 18 (2002) 5607–5612) showed that the adsorptionof one polyelectrolyte could induce desorption of polyelectrolyte complexes instead of regular deposition,more and more findings highlight peculiarities in the deposition of such films. Herein we demonstratethat the association of sodium polyphosphate (PSP) as the polyanion and either poly(-L-lysine hydrobro-mide) (PLL) or poly(allylamine chloride) (PAH) as the polycations may lead to non-monotonous filmdeposition as a function of time. Complementary, films containing PSP and PLL can be obtained from a(PLL–HA)n template films after the exchange of HA (hyaluronic acid) from the sacrificial template byPSP from the solution. This exchange is accompanied by pronounced film erosion. However, when start-ing from a (PAH–HA)n template, the film erosion and exchange due to the contact with PSP is by far lesspronounced, nevertheless the film morphology changes. These findings show that the nature of the poly-cation used to deposit the PEM film may have a profound influence of the film’s response to a competingpolyanion.

� 2011 Elsevier Inc. All rights reserved.

1. Introduction

For a long-time, functionalization of solid–liquid interfaces wasrestricted to the adsorption or chemisorption of essentially mono-layers with some specific deposition methods and techniquesrelated to the chemistry of both the adsorbates and the substrates.

The deposition of layers of amphiphiles by transfer from theliquid–gas interface to solid substrates, the more robust depositionof thiols on the surface of noble metals [1,2] or of alkyl silanes onthe surfaces of oxides [3] belong to these specific methods. Electro-static interactions are by far less specific since almost all surfacescarry surface charges. Hence, it is surprising to observe, a posteriori,how long the adsorption of polyelectrolytes at solid–liquid inter-faces was restricted to the adsorption of ‘‘monolayers’’. It wascertainly not realized that in certain conditions where the polyelec-trolytes adopt coil conformations (i.e., at weak charge density or inthe presence of strong electrostatic screening from the electrolytesolution) their adsorption leads to a charge overcompensation, i.e.,the adsorption of the polyelectrolyte leads not only to an accuratecompensation of the charges carried by the substrate to yield aperfectly neutral interface but to an excess of charges due to poly-electrolyte loops not in contact with the substrate. Hence, the

ll rights reserved.

adsorption of a polyelectrolyte in such conditions should allow forthe adsorption of an oppositely charged polyelectrolyte or an oppo-sitely charged colloidal particle. The feasibility of this idea was firstdemonstrated with rigid and oppositely charged silica particles byIler in 1966 [4]. The lack of efficient surface characterization toolscertainly precluded the extension of this concept to polyelectrolytesand it is only 20 years ago that Decher and coworkers demonstratedthe possibility to deposit polyelectrolyte multilayer (PEM) films [5].This first investigation showed that films made from at least 100alternately deposited ‘‘layers’’ can be obtained. The occurrence ofcharge reversal after the deposition of each ‘‘layer’’ was also sug-gested in this article even if the first demonstration of surface poten-tial reversal was only shown in 1996, to our knowledge [6]. Manyother investigators have then shown that alternated reversal ofthe surface potential upon adsorption of oppositely charged poly-electrolytes is a signature of PEM deposition [7,8] yielding to theconcept that ‘‘surface charge reversal is a driving force’’ for PEMdeposition. However, 20 years of research in the field of layer-by-layer (LBL) deposition showed that the facts are by far much morecomplicated than predicted by the basic proposed model of slightlyinterpenetrating chains [9].

(i) First of all, Kotov [10] proposed some numerical estimates ofthe free energy changes upon LBL deposition of oppositely

V. Ball et al. / Journal of Colloid and Interface Science 366 (2012) 96–104 97

charged polyelectrolytes and arrived to the conclusion thatother weak interactions than pure electrostatics in a mediumof high dielectric constant, like water, should play a major rolein the energy balance.

(ii) Isothermal titration microcalorimetry experiments aimed tomeasure the enthalpy changes upon the interaction betweenoppositely charged polyelectrolytes [11] and the correlationof the obtained values with the film growth regime of the filmsmade from the same polyelectrolytes [12] highlighted theimportance of entropy changes in the deposition of PEM films.The finding that in many situations the film deposition canonly occur through an increase in entropy is in perfect agree-ment with the early assessment of Michaels in the field of poly-electrolyte complexation [13]. The entropy increase may wellbe due to a release of counter-ions and from water of hydrationfrom the polyelectrolyte chains upon their complexation. Theadsorption of the chains decreases their degrees of freedomand has hence a negative contribution on the entropy change.The entropy balance during the deposition of PEM films ishence complicated because it contains terms that can mutuallycompensate.

(iii)The observation of supralinear growth regimes [14], in whichthe amount of polyelectrolytes per unit area of the substrateincreases much rapidly than linearly as expected for a regulargrowth, showed that the inter-diffusion of polyelectrolytesthrough a large part, but not necessarily through the wholethickness of the film, is possible. In this case, regular chargeovercompensation was also observed even after the first depo-sition steps where the substrate was not yet homogenouslycovered with a film but with islands instead [15]. The existenceof non-compensated charges on the surface of such filmsallows to explain why diffusion of the polyelectrolytes out ofthe film and back in the solution is most of the time not quan-titative, otherwise the film would decompose. A reservoir ofmobile molecules in such films is certainly also at the originof their Donnan potential [16], meaning that there is not onlyan excess of charges at the film–solution interface (compen-sated by the counterions from the double layer in solution)but also in the film.

(iv) Some recent investigations showed that some combinationsof polyelectrolytes [17] or some peculiar physicochemicalconditions [18] allow to deposit films that grow monoto-nously with the number of deposition steps but without alter-nated inversion of the surface potential. Indeed after about100 alternated deposition steps of poly (sodium phosphate)(PSP) and poly(allylamine hydrochloride) (PAH), the deposit,made from non-coalescent islands, has a constant negative fpotential, when the deposit is made with polyelectrolytes dis-solved at 10�4 M in monomer units and in the presence of0.15 M NaCl. This means that PSP adsorbs on the alreadydeposited islands or in vacancies in between and in conditionsof repulsive electrostatic interactions. Note that these PSP andPAH containing deposits were prepared through alternatedspray deposition. These findings point to the need to furtherinvestigate about the origin of the driving forces leading tosuch self-assembled films.

(v) Finally, it has been found that in certain circumstances, thealternated deposition of polyelectrolytes leads to very smalldeposition owing to the competition between adsorption anddesorption phenomena [19,20]. This finding has been rational-ized by the phase diagrams of polycation–polyanion mixturesin which there exists a glassy state and a soluble statedepending on the physicochemical conditions (salt concentra-tion, temperature). The occurrence of regular film deposition

corresponding to the ‘‘glassy’’ phase of the phase diagramwhereas the adsorption–desorption regime corresponds tothe ‘‘soluble’’ phase.

The findings of the Cohen Stuart’s group [19,20] are of majorimportance because they provide a general framework for thedeposition of PEM films, as well as of all kinds of ‘‘LBL’’ films(whose relying on polymers interacting through hydrogen bonds[21–23] or through donor–acceptor interactions [24]). This frame-work relies on the assumption that the knowledge of the phasediagram of a polycation–polyanion mixture could be a predictivetool for the occurrence of film deposition (or its absence) depend-ing on the occurrence (or not) of phase separation when PE aremixed in different ratios even different from 1:1 in monomer ratios[25,26].

In the case of the adsorption–desorption regime, a closeinspection of the adsorption kinetics performed by means of stag-nation point optical reflectometry showed that the surface con-centration of deposited polyelectrolyte increases up to a certainpoint in time before desorption starts leading eventually to a com-plete film erosion [19,20]. This finding suggests that the filmdeposition may be possible by allowing adsorption to occur onlyfor a very short time duration. This could make alternated depo-sition of polyelectrolytes through dipping in polyelectrolyte solu-tions a very time efficient deposition method, competitive withalternated spin coating [27] or alternated spraying [28,29] inwhich each deposition step lasts over only a few seconds. Thismay also change the classical idea that LBL deposition throughalternated dipping in solutions containing oppositely chargedpolyelectrolytes is time consuming. Indeed, Kleinfeld and Fergu-son showed that PEM films made from poly(diallyldimethylammonium chloride) and sodium montmorilonite reach 95% oftheir maximal thickness when each adsorption step was per-formed during only 5 s of dipping the substrate in the clay orpolycation containing solution [30]. Of course the concentrationof each polyelectrolyte solution is of major importance in the filmgrowth regime and kinetics [17,31].

Another important point in the dynamic aspects of PEM films isthe possible occurrence of exchange phenomena in which a depos-ited PEM film can undergo compositional and structural changesupon exposure to a solution containing a polyelectrolyte differentfrom its constituent chains [32–35].

Up to now, the influence of adsorption time and the possibilityto produce films with modified composition through an exchangeprocess has never been demonstrated simultaneously. We willshow in this article some peculiar aspects of LBL deposition:namely that for some polycation–polyanion combinations, thickerfilms can be obtained by using shorter deposition times and thatfilms of similar composition can be obtained from sacrificial PEMtemplates in which one kind of polyelectrolyte already present inthe template can be almost quantitatively replaced by the poly-electrolyte of interest. However, such an exchange process de-pends dramatically on the polyelectrolytes used to build up thesacrificial template: for instance, PSP–PLL containing films can bebuild from a (PLL–HA)n template but not from a (PAH–HA)n film.HA represents sodium hyaluronate. Our data hence complementsome suggestions proposed in previous work [36] about the impor-tance of the interpolyelectrolyte interaction strength on thedynamics of PEM films. The interaction strength between thechains can be changed by playing not only on the structure ofone of the polyelectrolytes, as in this work, but also by playingon physicochemical parameters as the pH [37], ionic strength[38], and temperature [39]. The present investigation aims at stim-ulating the need to investigate not only the deposition kinetics of

n 0 2 4 6 8 10 12

d / n

m

0

20

40

60

80

100

120

140

t / min0 1 2 3 4 5 6

d / n

m

101520253035404550

Fig. 1. Influence of dipping time on the deposition of (PLL–PSP)n deposits: ( ) 30 s,(d) 1 min, and (s) 5 min per adsorption step. The inset represents the average filmthickness determined by ellipsometry for films made form n = 8 deposition cycles.The error bars correspond to one standard deviation over five measurementsperformed on an individual coated silicon slide.

n0 2 4 6 8 10 12

d / n

m

0

50

100

150

200

t / min2 30 1 4 5 6

d / n

m

0

40

80

120

160

200

Fig. 2. Influence of dipping time on the deposition of (PAH–PSP)n deposits: ( ) 30 s,(d) 1 min, and (s) 5 min per adsorption step. The inset represents the average filmthickness determined by ellipsometry for films made form n = 8 deposition cycles.The error bars correspond to one standard deviation over five measurementsperformed on an individual coated silicon slide.

98 V. Ball et al. / Journal of Colloid and Interface Science 366 (2012) 96–104

polyelectrolytes during the deposition of PEM films but also toshow that exchange processes are an important issue for the stabil-ity of PEM films.

2. Materials and methods

2.1. Chemicals

All solutions were prepared from double distilled water (Milli-pore Simplicity system, q = 18.2 MX cm). The polyelectrolytes

were solubilized in 0.15 M NaCl solutions whose pH was not ad-justed. The pH of the solutions was checked with a pH meter (Met-tler Toledo) and was found equal to 6.7, 6.5, 7.5, and 8.0 for the PSP,HA, PLL, and PAH solutions, respectively. When not specified, allpolyelectrolyte solutions were prepared at a concentration of1 mg mL�1.

2.2. Substrates

The PEM films were deposited on P-doped silicon wafers (Sil-tronix, Archamps, France) that were previously cut in the form ofrectangles having a size of 4 � 1 cm2. These wafers were cleanedby successive immersion in hot (�60 �C) Hellmanex solutions (2%v:v) (Hellma GmbH, Müllheim, Germany) during half an hour, dis-tilled water (Millipore Simplicity System, q = 18.2 MX cm), hydro-chloric acid at 0.1 M during 10 min, and finally with distilled water.Freshly cleaned silicon wafers were used for each new depositionexperiment.

A trapezoidal ZnSe crystal (Graseby-Specac, Orpington, UK) wasused as a substrate for the infrared spectroscopy experiments inthe totally attenuated reflection mode (ATR-FTIR spectroscopy).The crystal was cleaned with a methanol wiped optical paperbefore the deposition experiments.

2.3. Film deposition

As long as we have not characterized the obtained coatings bymeans of atomic force microscopy, we will call them ‘‘deposits.’’These ‘‘deposits’’ will be called films if they display continuousdeposition, i.e., in the absence of channels going from the depos-it–solution interface down to the substrate. To prepare the (PLL–PSP)n and (PAH–PSP)n PEM deposits on the freshly cleaned siliconwafers, the samples were hold with a cleaned tweezer and im-mersed manually in the polycation, in the sodium chloride rinsingsolution, in the polyanion and again in the rinsing solution. Such adipping cycle consists in the deposition of one polycation and onepolyanion (defined as one layer pair in most articles in the ‘‘LBL’’field). Each (polycation–polyanion)n deposit was prepared individ-ually without intermediate drying and rehydration in order toavoid possible artifacts due to such a partial drying step which isnecessary for the characterization of the deposit. Hence, each pointin the figures corresponds to an individually prepared deposit. Theimmersion times in the polycation and polyanion solutions werethe same and changed from one experiment to the other. Thedeposition experiments were performed at (23 ± 2) �C in a roomfitted with an air conditioner. The importance of controlling thetemperature in the deposition of PEM films is of major importanceparticularly for those displaying a supralinear increase in theirthickness with the number of deposition steps [39].

2.4. Characterization methods

2.4.1. EllipsometryThe average thickness of the deposits was measured by means

of single wavelength ellipsometry at k = 632.8 nm (He–Ne laser)and at an angle of incidence of 70� (PZ 2000 Horiba, Longjumeau,France). To convert the ellipsometric angles W and D, into thick-ness values, we had to assume a value for the refractive index ofthe film. We choose a value of 1.465 as in our previous investiga-tions on (PLL–HA)n and (PLL–PGA)n films which is reasonable fora film made from polymer and containing a certain volume fractionof water [40] and counterions. Since the structure and hydration ofPSP containing films are not yet known and since phosphorous is asomewhat more polarizable element than carbon, nitrogen, andoxygen, the fact to fix the refractive index of the film at 1.465 couldinduce some systematic over or underestimation of the film

Fig. 3. AFM surface topographies of (PAH–PSP)n deposits for different dipping times and for films prepared from n = 4 and n = 8 deposition cycles. Each sample has undergonesome needle scratching to distinguish the deposit from the silicon substrate (right part of each image). The image sizes were of 20 � 20 lm2.

V. Ball et al. / Journal of Colloid and Interface Science 366 (2012) 96–104 99

thickness. To that aim, we used atomic force microscopy (AFM)profiling across sections made in the films deposited on silicon tocheck the consistency of the thickness determination made byellipsometry. In addition, ellipsometry allows only to determinethe average film thickness over an area close to 1 mm2, corre-sponding to the size of the light spot shined on the substrate.One is hence not able to detect whether a true film (without chan-nels reaching the substrate) or islands are deposited. Many LBLdeposits grow in a continuous manner but without forming filmsbut rather droplets that may [15] or not [17] coalesce upon an in-crease in the number of deposition steps. It is hence extremelydangerous to rely on ellipsometry or UV–vis spectroscopy aloneto ascertain the deposition of a film [41]. One or both of these tech-niques have to be combined with local imaging by AFM. In all thefigures, the thickness of the deposits will be denoted by ‘‘d’’.

2.4.2. Atomic force microscopyThe AFM topographies of the dried films (maintained in a vac-

uum chamber during a few hours before image acquisition) wereacquired in the tapping mode with a Pico SPM microscope (Molec-ular Imaging) at a frequency of 1 Hz. Each image was acquired witha new pyramidal silicon tip. The images were acquired on depositsthat were needle scratched just before image acquisition in orderto have access to the morphology of the deposits as well as theirthickness. The line profiles of the obtained sections were averagesover 30 line scans.

The images were acquired over squares 20 � 20 or 30 � 30 lmin area.

2.4.3. Infrared spectroscopy in the attenuated total reflexion mode(ATR-FTIR)

The ATR-FTIR spectra were acquired by the accumulation of 512interferograms with a Equinox 55 spectrophotometer (Bruker,Germany) at a spectral resolution of 4 cm�1 and at wavenumbersbetween 700 and 4000 cm�1.

The PEM films were deposited from NaCl solutions prepared inD2O. Additional details can be found in previous articles [35,36,42].

The IR spectrum of the PSP powder was also acquired in thetransmission mode with a Tensor 27 spectrometer (Bruker,Germany).

3. Results and discussion

3.1. Deposition kinetics of the (PLL–PSP)n and (PAH–PSP)n films

We first investigated the influence of the dipping time perdeposition step on the evolution of the average thickness of thedeposits calculated from ellipsometry measurements for two kindsof polyelectrolyte combinations having a common polyanion, PSP,but differing in the deposited polycation, PLL (Fig. 1) versus PAH(Fig. 2). In the presence of 0.15 M NaCl as the supporting electro-lyte, it appears that for all number of deposition steps, the thick-ness of the (PLL–PSP)n deposits passes through a maximum whenthe deposition time is of 1 min. The inset of Fig. 1 exemplifiesthe situation in the case of deposits made from n = 8 depositioncycles. When the silicon wafers are immersed in the PLL or PSP

Fig. 4. Profile scans (average over 30 scans) over the AFM images acquired in Fig. 3.

Table 1Mean root squared roughness of the (PAH–PSP)n deposits imaged in Fig. 3.

30 s (nm) 1 min (mm) 5 min (mm)

n = 4 21.6 37.6 46.8n = 8 51.9 71.6 134

t / min0 5 10 15 20 25 30

d / n

m

0

100

200

300

400n=5n=10n=15n=20

Fig. 5. Kinetics of film thickness change for different (PLL–HA)n films put in contactwith a PSP containing solution at 1 g mL�1. The average film thickness wasmeasured by means of ellipsometry.

100 V. Ball et al. / Journal of Colloid and Interface Science 366 (2012) 96–104

solutions for 5 min, the average thickness of the deposits is onlyslightly higher than when the deposition is performed during30 s. In these conditions of NaCl concentration, the PLL–PSP combi-nation of polyelectrolytes behaves as the systems described by Co-hen Stuart et al. [19,20], namely as a system in which desorption ofpolyelectrolyte complexes occurs in concurrence to polyelectrolyteadsorption and resulting in a situation in which maximal deposi-tion occurs for an optimal adsorption time. We show here thatthe adsorption kinetics passes through an ‘‘overshoot.’’ For a depo-sition time corresponding to the maximal deposited thickness, asupralinear increase in the thickness is observed (Fig. 1). If wewould not have investigated the influence of the adsorption timeand chosen arbitrarily a long adsorption time (5 or 10 min) as isusual for step-by-step deposition of polyelectrolytes, we wouldhave concluded that the deposition of (PLL–PSP)n yields to onlyvery thin deposits. This is obviously not the case provided theadsorption time is adjusted to reach an optimal deposition. Depo-sition times of 2 and 3 min yield already deposits of smaller thick-ness than those obtained after 1 min (data not shown). Theseresults are similar as those reported by Izumrudov et al. [25].

When replacing PLL by PAH for the alternated dipping processwith PSP, the situation is totally different in the sense that the film

thickness increases monotonously with the dipping time (Fig. 2).The inset in Fig. 2 shows the situation for deposits made fromn = 8 dipping cycles. For this combination of polyelectrolytes dis-playing a supralinear increase in thickness with the number of

n4 6 8 10 12 14 16 18 20 22

d / n

m

10

100

Fig. 6. Film thickness change of (PLL–HA)n films in contact with PSP solutions (at1 mg mL�1 during 25 min) as a function of the number of deposited layer pairs.Open and closed disks correspond to the film thickness before and after contactwith the PSP solution, respectively. Each pair of open disks and black squarescorresponds to an individual experiment.

1/λ (cm-1)80010001200140016001800

Abs

orba

nce

0.00

0.05

0.10

0.15

0.20

0.25

0.30amide I

amide II

polysaccharides

Fig. 7. ATR-FTIR spectra of a (PLL–HA)9 film before ( ) and after (++++) beingput in contact with PSP at 1 mg mL�1 during 25 min. The red line corresponds to thespectrum of the PSP powder. (For the interpretation of the references to color in thisfigure legend, the reader is referred to the web version of this article.)

n0 5 10 15 20

d / n

m

0

50

100

150

200

250

300

350

Fig. 8. Film thickness change of (PAH–HA)n films in contact with PSP solutions (at1 mg mL�1 during 25 min) as a function of the number of deposited layer pairs. Eachpair of open disks and black squares corresponds to an individual experiment. Theblack arrow indicates the thickness change of the as deposited (PAH–HA)n film (s)after 25 min of contact with the PSP solution (j).

V. Ball et al. / Journal of Colloid and Interface Science 366 (2012) 96–104 101

deposition cycles, it appears that the thicker the films are the moretime is required to reach a steady state in the adsorbed amount. In-deed, if one restricts to the deposition of less than n = 4 depositioncycles the thickness reaches a constant value even after 30 s(Fig. 2). This may be in relationship with the mechanism of filmgrowth: in the case of exponentially growing films, the growth oc-curs through the diffusion of one polyelectrolytes from the film tothe solution in order to reach the oppositely charged polyelectro-lyte in solution [43]. The polyelectrolyte complexation then occursat the film solution interface. The time required for the polyelectro-lytes in the film to reach the film–solution interface may depend

on the film thickness and can explain the present finding (Fig. 2)that the thicker the films are, the more time is required to reachan optimal film thickness. In the case where the film grows line-arly, with little interpenetration of the different layers, an increasein the dipping time is not expected to have a major influence of thethickness of the deposit.

In addition, in both systems, (PLL–PSP)n and (PAH–PSP)n, a con-tinuous film is formed after the deposition of n = 4 layer pairs (Figs.3 and 4). For these AFM characterizations, the (PLL–PSP)n and(PAH–PSP)n deposits were produced by changing the dipping time:30 s, 1 min, and 5 min. The (PAH–PSP)n films start to form a contin-uous film after four deposition cycles (Fig. 3). Our findings are not incontradiction with those described for films made from the samepolyelectrolytes sprayed on silicon in the presence of the same sup-porting electrolyte (NaCl at 0.15 M) [17]. Indeed, the spray coatingexperiments were performed at polyelectrolyte concentrations of10�4 M in monomer units whereas the experiments described here-in were performed at a concentration of 1 mg mL�1, namely close to10�2 M in monomer units. For the spray deposition method at poly-electrolyte concentrations of 10�4 M in monomer units, the depos-its never coalesced even after 75 deposition cycles [17].

The mean squared roughness of the (PAH–PSP)n films is given inTable 1. Its value increases with the number of deposition steps atconstant immersion time as well as with the immersion time for adeposit made from a constant number of deposition steps. The(PLL–PSP)n deposits also form continuous films after the depositionof at least four layer pairs as exemplified from the profile scansover the scratched regions in the imaged regions (data not shown).

The main message of this first part of the present article is thatthe optimal adsorption time to reach the thickest possible deposithas to be optimized for each particular film and that some combi-nations of polyelectrolytes may lead to a maximal thickness forvery short adsorption times before the occurrence of film erosion.One other conclusion from this investigation is that the affirmationthat step-by-step deposition by alternately dipping the substratesin oppositely charged polyelectrolyte solutions is a long andcumbersome process, may not always be true. Indeed for the(PLL–PSP)n combination, the thickest films are obtained after only

5 10 15 20

µm

280

200

120

40

nm

0

0 10 20

µm

µm 10

0

0 10 20

µm

0

10

20

µm

0 5 10 15 20

µm

550

400

250

100

nm

0 16 32

µm

µm

0

16

32

________________________0 10

µm

300

150

0

-150

-300

nm

20

3020

Fig. 9. Film morphology (20 � 20 lm) of a (PAH–HA)10 film before (upper row) and after (middle row) exposure to a PSP solution at 1 mg mL�1 during 25 min. The lower rowcorresponds to images (30 � 30 lm) of (PAH–HA)15 film exposed during 25 min to a 1 mg mL�1 PSP containing solution.

102 V. Ball et al. / Journal of Colloid and Interface Science 366 (2012) 96–104

1 min of immersion of the substrate, whatever the number ofdeposition steps.

3.2. Influence of PSP on (HA–PLL)n and on (HA–PAH)n films

In the present part, we ask whether (PLL–PSP)n and (PAH–PSP)n

films whose deposition kinetics are markedly different could notbe obtained through an exchange process in which a common pol-yanion, hyaluronic acid (HA), from (PLL–HA)n and from (PAH–HA)n

films could be exchanged by the polyanion, PSP, put in contactwith the film. Exchange process in PEM films have been widely de-scribed in the literature [32–34], but we aim here to go a bit furtherin this direction to highlight that the nature of the already depos-ited film plays a major role on the issue of the exchange process. By

changing the nature of the polycation, PLL or PAH, of the initiallydeposited and exponentially growing films, the issue of theexchange process is totally different. In the case of (PLL–HA)n films,the average thickness of the deposit decreases rapidly when thefilms are put in contact with a PSP solution at 1 mg mL�1 andreaches a steady state after about 5 min (Fig. 5). Most interestinglyfor (PLL–HA)n deposits with n smaller than 10, the thicker the ini-tially deposited film, the higher is the resulting film thickness(Fig. 6). However, for n larger than 10, the obtained deposit hasan almost constant thickness and is a continuous film (inset ofFig. 6). This result is comparable to the one obtained when(PAH–HA)n films are put in contact with potassium hexacyanofer-rate: such films undergo an important erosion but after the transi-tion from the exponential to the linear growth regime, occurring at

V. Ball et al. / Journal of Colloid and Interface Science 366 (2012) 96–104 103

n = 10–12, the resulting film has an almost constant thickness [36].Such phenomena may well be associated with a change in thecohesion of the films after the exponential to linear growth transi-tion. This point remains to be explored in future investigations.

The inset represents an AFM topography of a (PLL–HA)20 filmafter exposure to a PSP solution during 25 min. The film has beenneedle scratched before image acquisition.

The resulting film not only contains PSP as shown by means ofATR-FTIR spectroscopy (peaks at 875, 1080, and 1270 cm�1) butalmost all HA in the initial (PLL–HA)9 template has been displacedduring the PSP–HA exchange process (Fig. 7). It is highly surprisingthat such an exchange process is so rapid, being achieved in lessthan 10 min. This highlights again the dynamic nature of the(PLL–HA)n films in which a high chain mobility has been found bymeans of fluorescence recovery after photobleaching [44].

However, when one considers the (PAH–HA)n films, it appearsthat the thickness reduction is much less pronounced (Fig. 8) after25 min of contact with PSP solutions (at 1 mg mL�1) than for the(PLL–HA)n films (Fig. 6).

As another marked difference with the (PLL–HA)n deposits, theerosion of the (PAH–HA)n deposits increases continuously when nincreases: there is almost no film erosion for the (PAH–HA)5 filmsbut about 33% decrease in average thickness of the deposit for the(PAH–HA)15 films.

In Fig. 9, we show that the root mean square roughness of the(PAH–HA)10 films increases from 91 nm to 159 nm when put incontact with PSP at 1 mg mL�1. Clearly the deposits are not yetfilms after n = 10 deposition cycles but uncoalesced islands. Thestate of a continuous film is reached only after 15 deposition cycles(Fig. 9).

The ATR-FTIR spectra (data not shown) of the (PAH–HA)9 filmsafter contact with the PSP solution show the incorporation of somePSP in the films but without a complete disappearance of the peaksassigned to HA (at 1610 cm�1) as was observed for the (HA–PLL)9

films (Fig. 7). This shows that the amount of exchangeable HA isby far much lower in (PAH–HA)n than is the (PLL–HA)n films.

We make the assumption that the pronounced difference in thedeposition kinetics of (PLL–PSP)n and (PAH–PSP)n coatings origi-nates from much stronger interactions between PSP and PAH thanbetween PSP and PLL during the deposition. On the other side, weattribute the strong film erosion of (PLL–HA)n in the presence ofPSP versus the weak erosion of (PAH–HA)n films to the strongercohesion of the last kind of films with respect to the former ones.This assumption is in line with Isothermal Titration Calorimetryexperiments which showed that in the conditions of the presentexperiments, the reaction enthalpy between HA and PAH amountsto +640 J mol�1 (of monomer units) whereas the reaction enthalpyof HA and PLL amounts to +535 J mol�1 (of monomer units) [12].

4. Conclusions

In this investigation, we showed that the deposition kineticsleading to (PLL–PSP)n and to (PAH–PSP)n deposits are markedly dif-ferent when the experiments are performed in identical conditions.In the case where PAH is the polycation, the film thickness in-creases monotonously when the adsorption time of each polyelec-trolyte increases. The influence of the adsorption time becomessignificant only when the number of deposition cycles, n, is higherthan 6 in probable relationship with the supralinear (exponential)growth of the deposit. However, in the case where PLL is the poly-cation, the film thickness is maximal, whatever the number ofdeposition cycles, when adsorption is allowed for only 1 minmeaning that the deposition occurs through the interplay of com-petitive adsorption–desorption phenomena. Hence, the change ofthe nature of the polycation has a major influence on the deposi-

tion kinetics. This finding highlights for the need to investigatethe deposition kinetics for each considered pair of polycation/pol-yanion as long as a general ‘‘structure-build-up’’ relationship is notknown for PEM films.

Even more important is the finding that (PLL–HA)n and (PAH–HA)n films react in a different manner when exposed to a solutioncontaining PSP: the first ones undergo a marked dissolution withan almost quantitative displacement of HA in favor of PSP whereasthe second kind of films undergo only a very small thickness reduc-tion with some morphological change, namely an increase in filmroughness. Our investigation shows that PEM film can display fas-cinating dynamic behavior that need to be investigated withmolecular spectroscopy techniques, like NMR methods.

Acknowledgment

The authors thank the FEDER ‘‘Compétitivité régionale etemploi’’ 2007-2013 for financial support of this research Chapto-chem project N� 2009-02-039-35.

References

[1] R.G. Nuzzo, D.L. Allara, J. Am. Chem. Soc. 105 (1983) 4481–4483.[2] J.C. Love, L.A. Estroff, J.K. Kriebel, R.G. Nuzzo, G.M. Whitesides, Chem. Rev. 105

(2005) 1103–1169.[3] I. Haller, J. Am. Chem. Soc. 100 (1978) 8050–8055.[4] R.K. Iler, J. Colloid Interf. Sci. 21 (1966) 569–594.[5] G. Decher, J.D. Hong, J. Schmitt, Thin Solid Films 210–211 (1992) 831–835.[6] N.G. Hoogeveen, M.A. Cohen Stuart, G.J. Fleer, M.R. Böhmer, Langmuir 12

(1996) 3675–3681.[7] F. Caruso, E. Donath, H. Möhwald, J. Phys. Chem. B 102 (1998) 2011–

2016.[8] G. Ladam, P. Schaad, J.-C. Voegel, P. Schaaf, G. Decher, F.J.G. Cuisinier, Langmuir

16 (2000) 1249–1255.[9] G. Decher, Science 277 (1997) 1232–1237.

[10] N.A. Kotov, Nanostruct. Mater. 12 (1999) 789–796.[11] C.B. Bucur, Z. Sui, J.B. Schlenoff, J. Am. Chem. Soc. 128 (2006) 13690–13691.[12] N. Laugel, C. Betscha, M. Winterhalter, J.-C. Voegel, P. Schaaf, V. Ball, J. Phys.

Chem. B 110 (2006) 19443–19449.[13] A.S. Michaels, Ind. Eng. Chem. 57 (1965) 32–36.[14] D.L. Elbert, C.B. Herbert, J.A. Hubbell, Langmuir 15 (1999) 5355–5362.[15] C. Picart, Ph. Lavalle, P. Hubert, J.F.G. Cuisinier, G. Decher, P. Schaaf, J.-C. Voegel,

Langmuir 17 (2001) 7414–7424.[16] E. Hübsch, G. Fleith, J. Fatisson, P. Labbé, J.-C. Voegel, P. Schaaf, V. Ball,

Langmuir 21 (2005) 3664–3669.[17] N. Cini, T. Tulun, G. Decher, V. Ball, J. Am. Chem. Soc. 132 (2010) 8264–

8265.[18] M. Adusumilli, M.L. Bruening, Langmuir 25 (2009) 7478–7485.[19] D. Kovacevic, S. Van der Burgh, A. de Keizer, M.A. Cohen Stuart, Langmuir 18

(2002) 5607–5612.[20] D. Kovacevic, S. Van der Burgh, A. de Keizer, M.A. Cohen Stuart, J. Phys. Chem. B

107 (2003) 7998–8002.[21] A. Laschewsky, E. Wischerhoff, E. Denzinger, H. Ringsdorf, A. Delcorte, P.

Bertrand, Chem. Eur. J. 3 (1997) 34–38.[22] W.B. Stockton, M.F. Rubner, Macromolecules 30 (1997) 2717–2725.[23] S.A. Sukhishvili, S. Granick, J. Am. Chem. Soc. 122 (2000) 9550–9551.[24] Y. Shimazaki, R. Nakamura, S. Ito, M. Yamamoto, Langmuir 17 (2001) 953–

956.[25] S.A. Sukhishvili, E. Kharlampieva, V. Izumrudov, Macromolecules 39 (2006)

8873–8881.[26] H. Mjahed, J.-C. Voegel, A. Chassepot, B. Senger, P. Schaaf, F. Boulmedais, V.

Ball, J. Colloid Int. Sci. 346 (2010) 163–171.[27] P.A. Chiarelli, M.S. Johal, J.L. Casson, J.B. Roberts, J.M. Robinson, H.-L. Wang,

Adv. Mater. 13 (2001) 1167–1169.[28] J.B. Schlenoff, S.T. Dubas, T.R. Fahrat, Langmuir 16 (2000) 9968–9969.[29] A. Izquierdo, S.S. Ono, J.-C. Voegel, P. Schaaf, G. Decher, Langmuir 21 (2005)

7558–7567.[30] E.R. Kleinfeld, G.S. Ferguson, Science 265 (1994) 370–373.[31] L. Richert, Ph. lavalle, E. Payan, X.Z. Shu, G.D. Prestwich, J.-F. Stoltz, P. Schaaf, J.-

C. Voegel, C. Picart, Langmuir 20 (2004) 448–458.[32] H.W. Jomaa, J.B. Schlenoff, Langmuir 21 (2005) 8081–8084.[33] N.S. Zacharia, M. Modestino, P.T. Hammond, Macromolecules 40 (2007) 9523–

9528.[34] V. Ball, E. Hübsch, R. Schweiss, J.C. Voegel, P. Schaaf, W. Knoll, Langmuir 21

(2005) 8526–8531.[35] A.M. Pilbat, V. Ball, P. Schaaf, J.-C. Voegel, B. Szalontai, Langmuir 22 (2006)

5753–5759.[36] C. Betscha, V. Ball, Soft Matter 7 (2011) 1819–1829.[37] D. Yoo, S.S. Shiratori, M.F. Rubner, Macromolecules 31 (1998) 4309–4318.

104 V. Ball et al. / Journal of Colloid and Interface Science 366 (2012) 96–104

[38] H.W. Jomaa, J.B. Schlenoff, Macromolecules 38 (2005) 8473–8480.[39] M. Salomäki, I.A. Vinokurov, J. Kankare, Langmuir 21 (2005) 11232–11240.[40] M. Schönhoff, V. Ball, A.R. Bausch, C. Déjugnat, N. Delorme, K. Glinel,

R. v Klitzing, R. Steitz, Colloids Surf. A. Phys. Eng. Aspects 303 (2007) 14–29.[41] J.-F. Koenig, D. Martel, Thin Solid Films 516 (2008) 3865–3872.

[42] M. Debrezceny, V. Ball, F. Boulmedais, B. Szalontai, J.-C. Voegel, P. Schaaf,J. Phys. Chem. B 107 (2003) 12734–12739.

[43] P. Lavalle, C. Picart, J. Mutterer, C. Gergely, H. Reiss, J.-C. Voegel, B. Senger, P.Schaaf, J. Phys. Chem. B 108 (2004) 635–648.

[44] C. Picart, J. Mutterer, Y. Arntz, J.-C. Voegel, Microsc. Res. Tech. 66 (2005) 43–57.