Do

Ka

b

a

ARRAA

KTPM

1

iarfbsddimtt

0h

Applied Surface Science 261 (2012) 375– 384

Contents lists available at SciVerse ScienceDirect

Applied Surface Science

j our nal ho me p age: www.elsev ier .com/ loc ate /apsusc

iverse 2D structures obtained by adsorption of charged ABA triblock copolymern different surfaces

atri S. Kontturi a,∗, Arja-Helena Vesterinenb, Jukka Seppäläb, Janne Lainea

Department of Forest Products Technology, School of Chemical Technology, Aalto University, P.O. Box 16300, FI-00076 Aalto, FinlandDepartment of Biotechnology and Chemical Technology, School of Chemical Technology, Aalto University, P.O. Box 16100, FI-00076 Aalto, Finland

r t i c l e i n f o

rticle history:eceived 13 May 2012eceived in revised form 29 June 2012ccepted 3 August 2012vailable online 14 August 2012

eywords:riblock copolymerolyelectrolyte adsorptionorphological patterns

a b s t r a c t

In the larger context of 2D polymeric structures, the morphologies obtained by adsorption and subsequentdrying of charged, ABA type amphiphilic triblock copolymer of poly[2-(dimethylamino)ethyl metacry-late] (PDMAEMA) and poly(propylene oxide) (PPO) were investigated with atomic force microscopy andX-ray photoelectron spectroscopy as well as in situ adsorption analysis with quartz crystal microbalancewith dissipation monitoring. Hydrophilic silica and hydrophobic polystyrene (PS) were used as substratesfor adsorption. The structures emerging from the self-assembly of adsorbing polymer were profoundlyinfluenced by composition of the aqueous solution and the choice of substrate. When adsorbed fromdilute polymer solution where the concentration is so low that the polymer does not yet show surface-active behavior, the triblock copolymer unimers associated on hydrophilic silica surface forming large,irregular clustered aggregates, with sizes increasing with electrolyte concentration of the solution. Ona hydrophobic PS substrate, on the other hand, unimers spread much more evenly, forming clear sur-face patterns. The roughness of these patterned structures was tuned with the electrolyte concentrationof the solution. Adsorption from a more concentrated polymer solution, where the surface-activity ofthe polymer is perceptible, resulted in the formation of a smooth film with complete coverage overthe hydrophilic silica substrate when the electrolyte concentration was high. On PS, on the other hand,

nucleation of evenly scattered globular, disk-like micelles was induced. Besides the dry film morphology,the even distribution of the irreversibly adsorbed polymer over the PS surface was likely to serve as anoptimal platform for the build-up of reversible hydrophobically bound multilayers at high electrolyteconcentration. The multilayer formation was reversible because a decrease in the electrolyte concentra-tion of the solution re-introduces strong electrostatic repulsion between the multilayered polymer coils

wn of

which results in breakdo

. Introduction

Supported 2D structures from polymers have gained burgeon-ng interest because of their potential in novel applications suchs electronic sensors [1], antireflective coatings [2], smart mate-ials [3], platforms for controlled drug release [4], and scaffoldsor nanoparticle growth [5]. In case of polymers, there are a num-er of different ways to implant various nano- and microscaletructures on a substrate: spin coating [6], Langmuir–Blodgetteposition [7], various lithographic techniques [8], and controlledewetting [9,10], to name but a few. Adsorption from a solution

s an especially simple and effortless method for building poly-

eric surface structures as the process is completely defined by

he chemical interactions between the polymer, the solvent, andhe surface. As the self-assembly during adsorption is based on

∗ Corresponding author. Fax: +358 9 4514259.E-mail address: Katri.Kontturi@aalto.fi (K.S. Kontturi).

169-4332/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.apsusc.2012.08.018

the layer.© 2012 Elsevier B.V. All rights reserved.

reaching the thermodynamic equilibrium, the obtained structuresare stable. The use of block copolymers instead of homopolymersfor preparation of nanostructured templates for various function-alities typically results in formation of more elaborate structures[11]. However, adsorption of block copolymers and the result-ing 2D structures are increasingly difficult to predict. In thispaper, we have investigated the fundamental dry morphologiesobtained by adsorption and subsequent drying of charged, ABA typeamphiphilic triblock copolymer of poly[2-(dimethylamino)ethylmetacrylate] (PDMAEMA) and poly(propylene oxide) (PPO) on twochemically different surfaces.

There are numerous studies on adsorption of amphiphilicdiblock copolymers [12–14]. Fundamental studies exist also onthe adsorption of neutral, nonionic amphiphilic triblock copoly-mers, mainly composed of poly(ethylene oxide) (PEO) and PPO

(Pluronics) [15]. Neutral ABA triblock copolymers have latelybeen found to form outstandingly diverse morphologies whenphysically forced on a substrate by spin coating [16,17]. From afundamental point of view, it is relevant to understand also films

376 K.S. Kontturi et al. / Applied Surface

*O

O

O

n m *

O

On

N+

O

N+

n=24

m=34

aBAmbDb[bsITfaadu

2

2

qwtmfcst∼2q

bb

F(

Fig. 1. Chemical structure of the qDMAEMA24–PO34–qDMAEMA24.

nd submonolayer structures resulting from chemical deposition.ringing electrostatically charged groups into the structure of anBA block copolymer provides the possibility for tuning the confor-ation and subsequent morphological patterns upon adsorption

y varying the charge density and the electrolyte concentration.espite their potential, systematic research on structures formedy charged amphiphilic ABA triblock copolymers number in few18]. This study concentrates on morphological structures obtainedy adsorbing charged PDMAEMA–PPO–PDMAEMA on hydrophilicilica and hydrophobic polystyrene surfaces from aqueous solution.nfluence of salt and polymer concentrations are also examined.he focus is set on morphologies in the dry state, followed by atomicorce microscopy (AFM), because of their application potential. Inddition, a study of solution properties and in situ monitoring ofdsorption with quartz crystal microbalance (QCM-D) were con-ucted in order to complement the ex situ AFM data and to providenderstanding of polymer self-assembly on a fundamental level.

. Materials

.1. Quaternized PDMAEMA–PPO–PDMAEMA triblock copolymer

An ABA type triblock copolymer qDMAEMA24–PO34–DMAEMA24 (the subscripts indicate the number of repeats)as synthesized by oxyanionic polymerization of PPO macroini-

iator and 2-(dimethylamino)ethyl metacrylate (DMAEMA)onomer and then quaternized (q) into a permanently cationic

orm (PqDMAEMA–PPO–PqDMAEMA). The triblock copolymerhain is terminated at both ends simply with a proton. The detailedynthesis procedure is described in ref. [19] and the quaterniza-ion in ref. [20]. The molar weight of the PDMAEMA block was3.7 kDa (equivalent to 24 DMAEMA units) and the PPO block.0 kDa (equivalent to 34 PO units). The chemical structure of theDMAEMA24–PO34–qDMAEMA24 is presented in Fig. 1.

After synthesis, the polymer was purified from salt residuesy dialyzing with Spectra/Por regenerated cellulose dialysis mem-ranes with a molecular weight cut off value 6.0–8.0 kDa.

ig. 2. AFM height images (scan size 5 �m × 5 �m, with an inset of 1 �m × 1 �m) ofa) silica and (b) polystyrene.

Science 261 (2012) 375– 384

2.1.1. Polymer solutionsAqueous 20 g dm−3 stock solution was prepared by heating and

stirring mixture of polymer and water in ∼70 ◦C for 30 min. Afterthis, stirring was continued overnight at room temperature. The pHof the solutions was ∼6.5.

2.2. Other chemicals

Milli-Q water (resistivity 18.2 M� cm) was used throughout thestudy. NaCl of analytical grade was supplied by Merck (Darmstadt,Germany) and used without further purification.

2.3. Substrates

The QCM-D crystals used as substrates for adsorption wereAT-cut quartz crystals supplied by Q-Sense AB, Gothenburg,Sweden, with thickness 0.3 mm, fundamental resonance frequencyf0 ≈ 5 MHz, and sensitivity constant C = 0.177 mg m−2 Hz−1.

2.3.1. Silica substratesThe AT-cut quartz crystals were coated by the supplier with sil-

ica by means of vapor deposition. Before the QCM-D measurements,crystals were cleaned in a UV/ozonator (Bioforce Nanosciences,Ames, Iowa) with 28 mW cm−2 at 254 nm wavelength for 20 minand rinsed with water to remove any organic contaminants and toinduce the formation of hydroxyl groups on the surface. An AFMimage of a silica surface is presented in Fig. 2a.

2.3.2. Polystyrene substratesAT-cut quartz crystals that were coated with gold by the supplier

were used as substrates for polystyrene (PS) films. PS was depositedon gold surface by spin coating from 10 g dm−3 toluene solution atspinning speed of 4000 rpm. AFM image of a PS surface is presentedin Fig. 2b.

3. Methods

3.1. Solution properties

In order to study the behavior of qDMAEMA24–PO34–qDMAEMA24 in aqueous solution, series of 11 solutions with dif-ferent concentrations (from 0.0001 to 10 g dm−3) and 10 mM NaClwas prepared by diluting the 20 g dm−3 stock solution with waterand 0.1 M NaCl.

3.1.1. Dynamic light scattering (DLS)DLS measurements were conducted with a Malvern Zetasizer

Nano ZS instrument operating at a scattering angle of 173◦ anda wavelength of 633 nm. All DLS sample solutions were filteredwith 0.2 �m filters before the measurement. Temperature of thesolutions was maintained at 25 ◦C. Reported values of light scatter-ing intensities are averages of three measurements. In addition toscattering intensity values, the data on particle sizes in 20 g dm−3

solution with 0.1 M and 0.5 M NaCl were examined.

3.1.2. Surface tension measurementSurface tension was measured with a Kibron AquaPi tensiome-

ter by the Du Nouy maximum pull force method. The samplesolutions were equilibrated in saturated atmospheric humidity

and constant temperature (25 ◦C) for 24 h to ensure equilibriumbetween the air/solution interface layer and the bulk solution. Thesurface tension probe was cleaned by flaming before each measure-ment. Reported values are averages of three measurements.

urface

3

ttf1

3(

lScqao[poofasw

�

wtd

qw∼atttf

3

rfl

3

suicl

SIwt

3

mmsp

K.S. Kontturi et al. / Applied S

.2. Adsorption

Polymer solutions for QCM-D measurements were prepared inwo different concentrations, 0.05 g dm−3 and 0.5 g dm−3, based onhe data received from the solution property studies. Three dif-erent electrolyte concentrations were used: 1 mM, 10 mM, and00 mM NaCl.

.2.1. Quartz crystal microbalance with dissipation monitoringQCM-D)

Polymer adsorption and the characteristics of the adsorbedayer were studied with QCM-D (Q-Sense model E4, Gothenburg,weden) [21]. In this technique, the polymer-adsorbing QCM-Drystal is resonated in pulses, and the changes of resonant fre-uency �f and energy dissipation �D are measured simultaneouslyt the fundamental frequency and six overtones. Mass adsorbedn the crystal surface is directly proportional to the detected �f22]. �D measures the exponential decay in oscillation that is pro-ortional to the energy dissipated and stored during a period ofscillation. These data are attributed to the viscoelastic propertiesf the crystal and the adsorbed layer. If the adsorbed moleculesorm a thin, rigid, and homogenous layer (�D < 10−6) and thedsorbed mass is small compared to that of the resonating sub-trate, the mass of the adsorbed polymer can be calculated simplyith Sauerbrey equation [22]:

m = −C�f

n, (1)

here n is the overtone number and C is the sensitivity constant ofhe resonator. The principle of QCM-D has been described in moreetail elsewhere [23].

In this study, the aqueous solution of qDMAEMA24–PO34–DMAEMA24 was injected into the QCM-D cell as a continuous flowith the rate of 0.1 ml min−1. The shifts in f and D were followed for30 min, followed by rinsing first with aqueous NaCl solution with

concentration corresponding to that of the polymer solution, andhen with pure water. Temperature of 25 ◦C was used throughouthe study. In the reported experiments, results from the third over-one were used since these were more stable than those from theundamental frequency.

.3. Composition of the dry adsorbed films

After the adsorption experiments, the wet QCM-D crystals wereemoved from the QCM-D chamber and dried instantly with N2 gasow.

.3.1. Atomic force microscopy (AFM)A Nanoscope IIIa Multimode scanning probe microscope (ver-

ion V6.13 R1, Digital Instruments Inc., Santa Barbara, CA, USA) wassed to determine the morphology of the dry polymer films. The

mages were scanned in tapping mode in air at 25 ◦C using siliconantilevers. No image processing except flattening was done. Ateast three areas on each sample were measured.

Image analysis was performed with Nanoscope softwares andcanning Probe Image Processor (SPIP) software (version 4.5.3,mage Metrology, Lyngby, Denmark). The surface coverage values

ere determined using the Grain Analysis module in SPIP withhreshold algorithm.

.3.2. X-ray photoelectron spectroscopy (XPS)A Kratos Analytical AXIS 165 electron spectrometer with a

onochromatic A1 K� X-ray source was used to analyze the ele-ental and chemical compositions of the sample surfaces. All

pectra were collected at an electron take-off angle of 90◦ from sam-le areas less than 1 mm in diameter and using 100 W irradiation

Science 261 (2012) 375– 384 377

[24]. The spectra were recorded at least from three different spotson each sample and the analysis vacuum was monitored during theexperiments with an in situ reference sample.

The surface coverage of adsorbed polymer was calculated basedon the intensity of N 1s emission, taking into account the atomicproportion of N of all the XPS-detectable elements in a polymermolecule. The theoretical proportion of N of the detectable atoms(C, O, and N) in the structure of a qDMAEMA24–PO34–qDMAEMA24molecule (Fig. 1) was calculated to be 4.8%. Hence, if the intensity ofN 1s emission (i.e., the percentage of N) detected from the samplesurface would be, e.g., 0.5%, the percentage of polymer coveringthis surface would be 0.5/4.8 ≈ 10.4%. At least three measurementswere taken and averaged from each sample.

4. Results and discussion

First, the concentration dependency of the inter-molecular interactions of ABA type triblock copolymerqDMAEMA24–PO34–qDMAEMA24 in solution was investigated.Second, the adsorption process was monitored with QCM-D in twoconcentrations bearing different levels of surface-activity and alsoin different electrolyte concentrations. Third, the morphology ofthe dried adsorbed polymer films was studied.

4.1. Solution properties

Amphiphilic block copolymers are known to form micellesor less ordered aggregates in aqueous solution. ABA type blockcopolymers – ‘A’ denoting the hydrophilic and ‘B’ the hydropho-bic sequence – typically form a micelle with a dense hydrophobiccore and a corona of swollen hydrophilic chains [25]. Generally,the aggregation tendency of a neutral amphiphilic block copolymerincreases with increasing proportion of the hydrophobic sequencein the polymer chain.

When the block A of an amphiphilic block copolymer is charged,the association gets more complicated as also the electrostatic fac-tors affect the system. Charge reduces the hydrophobic effect andcauses electrostatic repulsion, which weakens the driving force foraggregate formation. Association may be completely hindered ifthe charged block is long compared to the hydrophobic block [26].Besides the chemical structure of the block copolymer, the ionicstrength of solution has a straight correlation on the aggregationbehavior in these kinds of systems [27].

Quaternized poly(dimethylaminoethyl metacrylate) (PqD-MAEMA) carries a strong positive charge with constant chargedensity within the whole pH range. PPO, on the other hand, isthermoresponsive by its solubility [28]. Throughout this study,however, the PPO block has constant hydrophobic nature as thetemperature is maintained at 25 ◦C.

Transitions in light scattering intensity and surface ten-sion values correspond to changes in molecular arrangementsin block copolymer solution. Surface tension data in Fig. 3indicate that the critical micellization concentration (cmc) forqDMAEMA24–PO34–qDMAEMA24 in 10 mM NaCl is ≈10 g dm−3.Above this concentration, the air/solution interface is saturatedwith polymer and polymer unimers in bulk solution start to formdistinct and stable micelles in order to minimize the contactbetween the hydrophobic moieties and water.

The surface excess, � , at the air/solution interface at ≥cmc wasobtained from the surface tension isotherm by the Gibbs adsorption

equation for ionic surfactants

� = − 12RT

(∂�

∂ ln c

)T

, (2)

378 K.S. Kontturi et al. / Applied Surface

Fig. 3. Light scattering intensity and surface tension as a function ofqN

w2cs(hd

ttcoPttcm[p

qtstcs

qcMaTmmDc[

qa≥w

in the charge densities of the adsorbent and the adsorbate indi-

DMAEMA24–PO34–qDMAEMA24 concentration in aqueous solution with 10 mMaCl. The solid lines are added as a guide to the eye.

here � is the surface tension, in the present case measured at98.2 K, R is the universal gas constant, and c is the molar con-entration of the qDMAEMA24–PO34–qDMAEMA24. The calculatedurface excess at maximum packing was 0.910 molecules nm−2

1.10 nm2 molecule−1 or 1420 ng cm−2). These values are notedere as a reference for further comparison with the areal adsorptionetermined at solid/liquid interfaces (see later sections).

Development of light scattering intensity in Fig. 3 indicates thathe size of particles in the solution starts to increase already whenhe polymer concentration exceeds ≈0.1 g dm−3 – a concentrationlearly below the actual cmc. This kind of premicellar aggregationf amphiphilic molecules is an established phenomenon [29,30].remicellar aggregation of block copolymers has been explainedo result from local minimization of free energy in the system: ashe polymer concentration increases, the amphiphilic moleculesollide with one another more and more in the bulk solution, whichay lead to formation of hydrophobically bound small aggregates

31]. Premicellar aggregates are less stable than micelles and inde-endent on the monolayer formation at the air/solution interface.

The relatively high aggregation concentrations of theDMAEMA24–PO34–qDMAEMA24 indicate that the micellisa-ion tendency of the block copolymer is rather moderate. Aubstantially lower cmc value has, indeed, been reported in litera-ure for qDMAEMA15–PO36–qDMAEMA15 [32] – a block copolymerorresponding to ours but having a higher B:A ratio and, thus,tronger hydrophobic effect that drives association.

Obviously, the stable micelles of ourDMAEMA24–PO34–qDMAEMA24 in aqueous solution shouldonsist of hydrophobic PPO cores surrounded by hydrophilic PqD-AEMA shells. Likewise, the possible weak premicellar aggregates

re also based on hydrophobic connection between PPO blocks.he shape of both the unstable aggregates and the distinct micellesay, however, be far from what is typical for regular surfactanticelles. For example, the micelle shape of a block copolymerMAEMA13–PO69–DMAEMA13 in pH 3 (i.e., PMDAEMA positivelyharged as in our case) was reported to be irregular (“worm-like”)33].

Measurement of the hydrodynamic size ofDMAEMA24–PO34–qDMAEMA24 premicellar aggregates or

ctual micelles turned out to be successful only in concentrationscmc where the amount of stable micelles present in the solutionas high enough. Furthermore, due to the constraints imposed

Science 261 (2012) 375– 384

by the measurement technique concerning the analyses of poly-electrolytes [34], high salt concentrations were required. In 20 gdm−3 qDMAEMA24–PO34–qDMAEMA24 solutions with 100 mMand 500 mM NaCl, the micelle sizes were detected to be 5.6 nmand 6.5 nm, respectively. It should be borne in mind that thepresence of salt typically decreases the hydrodynamic size of apolyelectrolyte to some extent [35].

4.2. Adsorption mechanism

The influences of the chemical composition of the substrate, theelectrolyte concentration, and the polymer concentration on theadsorption were investigated.

4.2.1. Effect of electrolyte concentrationIonic strength is widely known to affect solution properties and

adsorption of polyelectrolytes [35]. In the studied systems, screen-ing of the charges by addition of electrolyte affects in the systemin two ways: (1) by decreasing the attraction between the cationicpolyelectrolyte and the anionic surface and (2) by decreasing therepulsion between charges along polyelectrolyte chains.

4.2.1.1. Hydrophilic silica. The QCM-D results in Fig. 4a andc illustrate the dynamics of the process of the cationicqDMAEMA24–PO34–qDMAEMA24 adsorbing on silica from differ-ent NaCl concentrations. The recorded values of frequency shift(�f) are directly proportional to the mass adsorbed on the surface,whereas the dissipation factor (�D) is attributed to the viscoelasticproperties of the adsorbed layer. The |−�f| values increase uponincreased electrolyte concentration (Fig. 4a). This is an expectedphenomenon in polyelectrolyte adsorption: the presence of saltweakens the repulsive intramolecular electrostatic interactionsalong the polyelectrolyte chain, which leads to increased coilingof the molecule and thus enables more molecules to fit the surface[35]. The �D values increase slightly with electrolyte concentra-tion (Fig. 4b), indicating increased softness of the adsorbed layer,but the values are nonetheless so low (<10−6) that the layers canall be considered to be tightly bound and rather rigid. Despite thesubstantial screening of the charges, relatively high and irreversibleadsorption of qDMAEMA24–PO34–qDMAEMA24 on silica indicatesthat the polymer still behaves like polyelectrolyte in 100 mM NaCl.The electrostatic interaction between the cationic PqDMAEMAblock and anionic [36,38] silica surface is likely to be the major driv-ing mechanism behind the adsorption [39,40]. Numerous studiesexist on the charge density of silica surfaces under the influenceof various electrolyte concentrations and pH values [36–38]. How-ever, electrostatically driven polyelectrolyte adsorption neutralizesthe charge because of its irreversible nature, which increases thedissociation level of the yet unoccupied hydroxyl groups [41]. In thetheoretical limiting case where all surface silanols are ionized dueto adsorption of the cationic polyelectrolyte the maximum den-sity of charges is 3.1 sites/nm2 [42], corresponding to an averagedistance of ∼0.6 nm between the adjacent charges.

Rigidity of the adsorbed qDMAEMA24–PO34–qDMAEMA24 lay-ers seems reasonable from the following simple calculations: ifthe PqDMAEMA chain would be like a comb, all qDMAEMA sidechains pointing to the same direction, the distance between adja-cent cationic groups would be ∼0.25 nm. In reality, however, theeffective distance between charges along the polymer is undoubt-edly longer because the qDMAEMA side chains are mobile and thedistance of a cationic quaternized amine from the polymer back-bone is as long as ∼0.6 nm. All in all, the same order of magnitude

cate – neglecting the potential conformational restrictions causedby the lack of affinity between the hydrophobic maximum 12 nmlong PPO middle block and anionic silica – that flat adsorption with

K.S. Kontturi et al. / Applied Surface Science 261 (2012) 375– 384 379

(a) (b)

polymer

polymer

+ salt

salt water100 mM

10 mM

1 mM

80

-60

-50

-40

-30

-20

-10

0

Δf

(Hz)

saltpolymer

+ saltwater

-

100 mM

10 mM

1 mM

60

70

50

40

30

20

10

0

-Δf

(Hz)

(d)(c)1.7

1.5

1.3

1.1

0.9

0.7

0.5

ΔD

(10

-6)

100 mM

10 mM

1 mM

100 mM

10 mM

1 mM

1.7

1.5

1.3

1.1

0.9

0.7

0.5

ΔD

(10

-6)

706050403020100

time (min)706050403020100

time (min)

0.3

0.1

-0.1

time ( min)time ( min)

0.3

0.1

-0.10 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

(min)time(min)time

Fig. 4. The changes in inverse of frequency (−�f, upper figures) and dissipation (�D, lower figures) as a function of time during adsorption of qDMAEMA24–PO34–qDMAEMA24

f hilic sc

rPdat

4qcatcaarwdsnbsbi

atsbwa(tbtN

rom 0.05 g dm−3 solutions with different NaCl concentrations on (a and c) hydrophanging the injected solution.

ather efficient charge neutralization could be possible between theqDMAEMA block and silica. Rinsing the layers with pure wateroes not have a remarkable effect on the �f. This confirms that thedsorption is irreversible, further supporting the assumption thathe process is electrostatically driven.

.2.1.2. Polystyrene. QCM-D data of the adsorption of theDMAEMA24–PO34–qDMAEMA24 on PS from different NaCloncentrations are presented in Fig. 4b and d. Similar to silica,dsorption basically increases with increasing electrolyte concen-ration, as indicated by the �f values (Fig. 4b). Also here, increasedoiling of the polymer enables more polymer molecules to reachnd attach on the surface. The �D values (Fig. 4d) were likewisenalogous to those on silica, i.e., the adsorbed layers appearedelatively rigid. However, the adsorption kinetics differs fromhat was observed with silica, which must be connected with theifferent interaction mechanism between the polymer and theurface. As the PS substrate is hydrophobic and electrostaticallyeutral, ABA type amphiphilic block copolymers with PPO middlelock have been generally assumed to adsorb from aqueousolution onto it via hydrophobic forces between anchoring PPOlock, whereas the solvated hydrophilic head groups extend out

nto the solution as tails and loops [43–45].At 100 mM NaCl concentration, the adsorption on PS does not

ppear to reach a plateau (Fig. 4b). In this electrolyte concentration,he repulsion between the charged polymer coils is screened toome extent, enabling a slow multilayer formation on a hydropho-ic substrate as the attraction between the hydrophobic blocksithin the polymers is emphasized. The fact that the layer formed

t 100 mM NaCl is partially removed by rinsing with pure waterFig. 4b and d), is also indicative of a multilayer formed. By con-

rast, at 10 mM NaCl concentration, the screening of the repulsionetween the charged polymer coils is not intensive enough for mul-ilayer formation to occur. However, the charge screening at 10 mMaCl leads to coiling of the charged blocks, which appears to allow

ilica and (b and d) polystyrene. The arrows are added to point out the moment of

relatively fast adsorption kinetics as opposed to the system in 1 mMNaCl concentration where the polymers are still straight-chainedand the adsorption is slow (Fig. 4b).

Overall, the QCM-D results (Fig. 4) indicate that the mostdistinct difference between the adsorption of the chargedqDMAEMA24–PO34–qDMAEMA24 block copolymer on silica andPS is the reversible multilayer formation driven by intermolecu-lar hydrophobic forces speculated to take place on PS. The reasonwhy the multilayer structure does not occur on silica as well mustbe related to the differences between the structures of the firstadsorbed layers on these substrates.

4.2.2. Effect of polymer concentrationTo investigate the influence of the effective surface-activity on

the morphologies of the adsorbed films, two polymer concentra-tions were selected for the adsorption experiments: one belowthe surface active region (0.05 g dm−3), and another in the sur-face active region (0.5 g dm−3), where the polymer enriches tothe air/liquid interface and also some premicellar aggregation isexpected to take place (based on the study of solution propertiesin Section 4.1). The salt concentration was kept constant (100 mMNaCl). The QCM-D curves −�f and �D of the polymer adsorptionon silica and PS from 0.05 g dm−3 and 0.5 g dm−3 are presented inFig. 5.

Fig. 5a and c shows that the increase in polymer concentra-tion seems to have only a minor effect on the adsorption onsilica; there is only a slightly higher amount of water bound insidethe layer adsorbed from 0.5 g dm−3 than there is in the case ofthe 0.05 g dm−3. At both concentrations, i.e., in the case of bothunimers and the hydrophobically formed premicellar aggregates,the adsorption is electrostatically driven and irreversible. The main

difference between the adsorbed layers is that rinsing with watercauses a leap in −�f and �D values for the layer adsorbed from0.5 g dm−3 (Fig. 5a and c). The leap is attributed to water uptakethat is probably an indication of electrostatically driven expansion

380 K.S. Kontturi et al. / Applied Surface Science 261 (2012) 375– 384

F D, low 05 g dp f chan

omfos

P0w−tsisfawbcfsamoa

tqStcustcw

ig. 5. The changes in inverse of frequency (−�f, upper figures) and dissipation (�ith 100 mM NaCl and polymer concentrations below the surface active region (0.olystyrene as a function of time. The arrows are added to point out the moment o

f the adsorbed polymer coils and hydrophobically formed pre-icellar aggregates caused by a sudden decrease of ionic strength

rom 100 to 0 mM NaCl. The fact that the leap occurs only in the casef 0.5 g dm−3 indicates that there must be differences between thetructural properties of these two adsorbed layers.

As for the adsorption of qDMAEMA24–PO34–qDMAEMA24 onS, polymer concentration has a substantial effect. Increase from.05 to 0.5 g dm−3 results in significant addition in the amount ofater bound inside the layer, as shown by clear increase in both�f and �D in Fig. 5b and d. In both concentrations, the adsorp-

ion still continues after a fast initial phase, and replacing the NaClolution with water leads to substantial desorption. These resultsndicate that a build-up of a hydrophobically bound multilayertructure occurs in both concentrations. The multilayer adsorbedrom 0.5 g dm−3 containing premicellar aggregates that are formedlready in bulk solution is more loose and able to bind moreater. Rinsing with water results in breakdown of these multilayers

y re-introducing the strong electrostatic repulsion between theharges along PqDMAEMA chains. The breakdown effect is greateror the layer adsorbed from 0.5 g dm−3 solution as suggested by aevere decrease in −�f, which is probably an indication of heavyggregates releasing from the layer. This type of monocomponentultilayer formation has been observed also during the adsorption

f a charged, highly hydorphobized copolymer from high polymernd electrolyte concentrations [46].

When the adsorbed layer is rigid (�D < 10−6) – which ishe case in most experiments of this study – the mass ofDMAEMA24–PO34–qDMAEMA24 adsorbed can be estimated withauerbrey equation (Eq. (1)). The maximum masses adsorbed onhe solid/solution interfaces are in the range of 50–280 ng cm−2,orresponding to packing of 0.03–0.18 molecules nm−2. These val-es are much lower than the extent of surface coverage at the

aturated air/solution interface (0.91 molecules nm−2, see Sec-ion 4.1). This kind of denser packing at air/solution interfaceompared to solid/solution has been observed also by othershen adsorbing qDMAEMA24–EO132–PO50–EO132–qDMAEMA24

wer figures) during adsorption of qDMAEMA24–PO34–qDMAEMA24 from solutionsm−3) and on the surface active region (0.5 g dm−3) on (a and c) silica and (b and d)ging the injected solution.

onto polypropylene and silica surfaces: the amount on the satu-rated solid/solution interfaces was ∼0.03 molecules nm−2 whereasthe value for the air/solution was 0.81 molecules nm−2 [40]. Thiskind of denser packing at air/solution interface compared tosolid/solution interface is typically ascribed to the higher config-urational entropic penalty during packing of the copolymers at asolid surface compared to packing at a more diffuse air/solutioninterface [47].

Also, general trends observed in this study for adsorp-tion of the qDMAEMA24–PO34–qDMAEMA24 were similar tothose observed when adsorbing the qDMAEMA24–EO132–PO50–EO132–qDMAEMA24 from the surface-active polymerconcentration region with 100 mM NaCl: on hydrophobic surface,the adsorbed mass was higher than on silica and also the deforma-tion of the polymer layer induced by rinsing with water was moresubstantial [40].

4.3. Morphology of the adsorbed layers

4.3.1. Effect of electrolyte concentrationAFM images of the dried qDMAEMA24–PO34–qDMAEMA24 films

adsorbed on hydrophilic silica from different NaCl concentrationsare presented in Fig. 6. The films consist of aggregates or clus-ters of aggregates that cover the surface only partly. Generally,the aggregate size increases with electrolyte concentration. In caseof 10 mM NaCl, however, the film appeared to be much moreheterogeneous than the films adsorbed from 1 or 100 mM NaCl andthereby could not be represented by a single AFM image. Forma-tion of increasing aggregates with increasing NaCl concentrationis in agreement with the QCM-D analysis in Section 4.2 where theadsorption was found to increase with salt concentration. The pres-ence of salt causes screening of electrostatic repulsion between the

forming aggregates, which enables denser packing and formationof larger clusters on the surface.

At 0.05 g dm−3 concentration, the polymer should be essentiallynon-aggregated in solution. However, the formation of aggregates

K.S. Kontturi et al. / Applied Surface Science 261 (2012) 375– 384 381

Fig. 6. 7 �m × 7 �m AFM height images of dried qDMAEMA24–PO34–qDMAEMA24

films adsorbed on hydrophilic silica from 50 ppm solutions with (a) 1 mM, (b and c)1oi

ociatcibibo

attA

XPSAFM

25

20

15

he s

urf

ace (

%)

10

5

0

Co

vera

ge o

f th

1 mM 10 mM 100 mM

0 mM, and (d) 100 mM NaCl. The h values denote the height scale of the featuresn the surface. The two different images of the film adsorbed from 10 mM NaClllustrate the poorer repeatability of the measurement.

n the surface adsorbed from a solution consisting of free polymerhains is not an unusual phenomenon. There are both theoret-cal [48] and experimental [49] studies indicating that chargedmphiphilic block copolymers or surfactants in solutions belowheir critical micellisation concentration can associate at oppositelyharged surfaces as the local concentration of charged unimersn the vicinity of an oppositely charged surface exceeds theirulk solution concentration. It has been observed before [50] that

ncreasing the salt concentration of a dilute solution of chargedlock copolymer can lead to the formation of aggregate networksr clusters during adsorption onto different substrates.

Because of the detected heterogeneity of the polymer film

dsorbed from 10 mM NaCl on hydrophilic silica (Fig. 6), the inves-igation was complemented with XPS measurements. The XPS dataogether with AFM image analysis data are presented in Fig. 7.s observed with AFM, films formed on hydrophilic silica do not

Fig. 7. qDMAEMA24–PO34–qDMAEMA24 amount on hydrophilic silica as calculatedbased on XPS and AFM data of the polymer films adsorbed on the surface from0.05 g dm−3 solutions with different NaCl concentrations.

fully cover the surface. According to both the XPS data and theAFM image analysis, surface coverage varies between ∼5 and 20%,basically increasing with salt concentration. The dispersion of thedetected polymer amounts on the film adsorbed from 10 mM NaClis especially wide, supporting the conception that the film is par-ticularly heterogeneous.

AFM images of the dried qDMAEMA24–PO34–qDMAEMA24 filmsadsorbed on PS are presented in Fig. 8. Also here, the films consistof aggregates with sizes increasing with salt concentration. In eachof the films, however, aggregates are much more evenly shapedand distributed than those observed on hydrophilic silica (Fig. 6)and thus each film appears rather uniform. The film adsorbed onPS from 1 mM NaCl solution bears actually complete coverage overthe substrate with an appearance very similar to deformed vesiclesof a charged amphiphilic diblock copolymer poly(butadiene)140-poly(2-vinylpyridene)140 adsorbed on mica [51]. This kind oftoroidal shape of a vesicle is explained to form when water escapesfrom the inner core of the vesicle during drying of the sample. Incase of qDMAEMA24–PO34–qDMAEMA24 solution, however, vesi-cle formation seems unlikely [52] because the level of surfaceactivity of the polymer is only moderate, as stated in Section4.1. Instead, the surface structures on PS may have resulted fromqDMAEMA24–PO34–qDMAEMA24 unimers adsorbing individuallyon top and besides each other via hydrophobic forces, with thedegree of molecular coiling and thus the z scale variation of theadsorbed layer increasing with salt concentration.

Formation of multilayer structures in the presence of 100 mMNaCl during the adsorption of qDMAEMA24–PO34–qDMAEMA24was speculated based on the QCM-D results in Section 4.2. Theidea of reversible build-up of multilayer stemmed from the factthat (i) adsorption onto PS did not reach a plateau and (ii) rinsingwith water at the end of the measurement induced substan-tial desorption from the adsorbed layer. The suggested processof multilayer formation is driven by intermolecular hydropho-bic attraction between polymer coils when the charges along thepolyelectrolyte blocks are effectively screened by NaCl. Corre-spondingly, the breakdown of the multilayer is a result of increasedelectrostatic repulsion between the multilayered coils induced bythe drastic decrease in NaCl concentration during rinsing (i.e.,replacing the strong NaCl solution with pure water). The fact that

the reversible multilayer formation seemed to take place on PS butnot on silica can be rationalized by the morphological differences ofthe first layer adsorbed on these substrates: the even distributionof the irreversibly adsorbed polymer over the PS surface (Fig. 8c)

382 K.S. Kontturi et al. / Applied Surface Science 261 (2012) 375– 384

Fa(

il(f

ial

4

a

Fig. 9. AFM height images of qDMAEMA24–PO34–qDMAEMA24 films from solu-tions with 100 mM NaCl and polymer concentration (a) 0.05 g dm−3 adsorbed onhydrophilic silica (scan size 7 �m × 7 �m), (b) 0.5 g dm−3 adsorbed on hydrophilic

ig. 8. 1 �m × 1 �m AFM height images of qDMAEMA24–PO34–qDMAEMA24 filmsdsorbed on polystyrene from 0.05 g dm−3 solutions with (a) 1 mM, (b) 10 mM, andc) 100 mM NaCl. The h values denote the height scale of the features on the surface.

s likely to serve as an optimal platform for the build-up of multi-ayer structures. The silica surface with unevenly scattered clustersFig. 6d), on the other hand, would be too heterogeneous and roughor the purpose.

We emphasize that all AFM images illustrate the morphologyn the dry state. It is the focus of this study because of the largepplication potential of dry polymeric 2D structures. Therefore, theink to the in situ QCM-D data is always indirect.

.3.2. Effect of polymer concentrationAFM images of the dried qDMAEMA24–PO34–qDMAEMA24 films

dsorbed on hydrophilic silica and polystyrene from 0.05 g dm−3

silica (7 �m × 7 �m, with an inset of 1 �m × 1 �m), (c) 0.05 g dm−3 adsorbedon polystyrene (1 �m × 1 �m), and (d) 0.5 g dm−3 adsorbed on polystyrene(1 �m × 1 �m). The h values denote the height scale of the features on the surface.

and 0.5 g dm−3 solutions are presented in Fig. 9. The unimersadsorbing from 0.05 g dm−3 on silica (Fig. 9a) form rather large clus-ters on the surface, as discussed already in the previous section.From 0.5 g dm−3 (Fig. 9b), on the other hand, the film adsorbed isstrikingly smooth (z variation ∼6 nm) and seems to be fully coveredwith the polymer. (Note: z variation of vapor deposited silica surfaceis in the same range (∼6 nm) but the morphology is different, seeFig. 2a.) Hydrophobic attraction between the PPO cores has, thus,overcome the electrostatic repulsion between the PqDMAEMAcoronas of the aggregates as the aggregates have merged into a uni-form film. Based on the QCM-D analysis in Section 4.2, the amountof polymer adsorbed from 0.05 g dm−3 and 0.5 g dm−3 solutions isapproximately the same. The major difference between the twoadsorbed layers detected by QCM-D was that rinsing with waterinduced strong swelling of the layer adsorbed from 0.5 g dm−3 solu-tion, whereas no such effect was seen in the layer adsorbed from0.05 g dm−3. This seems logical from the morphological point ofview: a uniform film (Fig. 9b) is undoubtedly able to bind morewater than the sparsely scattered clusters (Fig. 9a). Appearanceof qDMAEMA24–PO34–qDMAEMA24 in solution in these two con-centrations and the structures adsorbed on silica are schematicallyillustrated in Fig. 10.

It seems based on AFM data that the formations adsorbed fromthe pre-surface-active concentration region on polystyrene (Fig. 9c)have spread over the surface rather evenly although the shapesof the structures are somewhat irregular. As speculated earlier,the formations may consist of qDMAEMA24–PO34–qDMAEMA24unimers adsorbed individually on top and besides each other viahydrophobic forces. The film adsorbed on polystyrene from thesurface-active region (Fig. 9d) consists of globular and disk-likeaggregates that have nucleated already in the solution state. Thevicinity of the surface seems to have induced growing of the weakpremicellar aggregates to more distinct and regular features. This

hypothesis is in agreement with the general fact that the cmc tendsto be lower on a surface than in bulk solution [53]. The assump-tion of micellization is supported not only by the regular shapeof the observed features on the surface, but also by the fact that

K.S. Kontturi et al. / Applied Surface Science 261 (2012) 375– 384 383

F 00 mr er) fro

ttr

ioatwrpmc

tstquoqfibamdhchs

5

qmfacfsl1s

ig. 10. Schematic illustration of qDMAEMA24–PO34–qDMAEMA24 solution with 1egion (0.5 g dm−3), and the structures adsorbed irreversibly (after rinsing with wat

he detected feature height value of ∼5–9 nm is on the range ofhe aggregate sizes measured in solution state with DLS (∼6 nm,eported in Section 4.1).

As discussed earlier in Section 4.2 based on the QCM-D data,t appears that both the unimeric coils (Fig. 9c) and premicellarr micellar aggregates (Fig. 9d) are able to pile up to form a softnd thick layer with a lot of water imbibed. The patterns seen inhe Fig. 9c and d represent the state of the films after rinsing withater, i.e., after the loosely bound multilayering material has been

emoved from the surface. Extremely even distribution of surfaceatterns in both cases explains the ability for acting as platform forultilayers. These irreversibly adsorbed structures are schemati-

ally illustrated in Fig. 10.In literature [54], micelles of nonionic surfactants (i.e., neu-

ral diblock copolymers) adsorbed on hydrophilic and hydrophobicubstrates are reported to form the structures of opposite typeo what we found with the adsorbing (premicellar) aggregates ofDMAEMA24–PO34–qDMAEMA24: nonionic surfactants form glob-lar micelles on hydrophilic surface and laterally continuous layersn hydrophobic surface. It can be speculated that merging of theDMAEMA24–PO34–qDMAEMA24 aggregates into more continuouslm on the hydrophobic substrate is hindered by (i) the repulsionetween the charges of coronal PqDMAEMA and (ii) their lack offfinity toward the hydrophobic surface and also (iii) the limitedobility of the aggregated ABA triblock chains compared to that of

iblock copolymer chains. In the adsorption of ABA aggregates ontoydrophilic surface, on the other hand, such major conformationalhanges are not required for spreading of aggregates because of theigh affinity between the coronal charges of the aggregates and theurface.

. Conclusions

In this study, the cationic ABA type triblock copolymerDMAEMA24–PO34–qDMAEMA24 was found to be a versatileolecule for preparation of tunable 2D morphologies emerging

rom the simple self-assembly upon adsorption. By tuning thedsorption conditions – electrolyte and polymer concentration andhoice of the substrate – diverse morphologies were achieved:rom smooth continuous films to conspicuous aggregates of varying

izes, shapes and regularity in appearance. At times, the adsorptioned to loosely scattered aggregates with diameters on the order of

�m whereas other conditions resulted in more densely occurringurface features with circular cross sections and diameters of tens

M NaCl from pre-surface-active region (0.05 g dm−3) and from the surface-activem both concentrations on silica and PS.

of nanometers. The formation of the morphologies was fundamen-tally explored with the theoretical aspects of polyelectrolyte andblock copolymer adsorption, supported with experimental in situdata on adsorption kinetics.

This study presents fundamental aspects of 2D structure forma-tion from qDMAEMA24–PO34–qDMAEMA24. It serves as a startingpoint for preparing diverse morphologies from similar triblockcopolymers. Since this paper presents essentially dry morpholo-gies, i.e., the state after the films have been removed from theaqueous environment, the possible uses for these materials arelikely to involve functions that employ the structures in their drystate. However, also the wet state morphologies may find valuableapplications with the structural rearrangements in the adsorbedlayers imposed by changing the solution conditions in the sur-rounding aqueous medium. Naturally, tailoring the 2D structuresfor each functionality will require further tuning of the adsorptionconditions.

Acknowledgements

Dr. Joseph Campbell and Dr. Leena-Sisko Johansson are acknowl-edged for their valuable contribution in the XPS analysis. LauraTaajamaa M.Sc. (Tech.) is warmly acknowledged for helpingwith the QCM-D measurements. Dr. Eero Kontturi is generouslyacknowledged for the discussions and insightful comments regard-ing to the manuscript. Financial support for this work was providedby the Finnish Funding Agency for Technology and Innovation(TEKES) and a group of Finnish industries.

References

[1] M.E. Roberts, A.N. Sokolov, Z. Bao, Material and device considerations fororganic thin-film transistor sensors, Journal of Materials Chemistry 19 (2009)3351–3363.

[2] S. Walheim, E. Schaffer, J. Mlynek, U. Steiner, Nanophase-separated polymerfilms as high-performance antireflection coatings, Science 283 (1999) 520–522.

[3] I. Tokarev, M. Motornov, S. Minko, Molecular-engineered stimuli-responsivethin polymer film: a platform for the development of integrated multi-functional intelligent materials, Journal of Materials Chemistry 19 (2009)6932–6948.

[4] A.N. Zelikin, Drug releasing polymer thin films: new era of surface mediateddrug delivery, ACS Nano 4 (2010) 2494–2509.

[5] K.D. Anderson, K. Marczewski, S. Singamaneni, J.M. Slocik, R. Jakubiak, R.R. Naik,

T.J. Bunning, B.B. Tsukruk, Plasma amino acid coatings for a conformal growth oftitanian anoparticles, ACS Applied Materials & Interfaces 2 (2010) 2269–2281.

[6] K. Norrman, A. Ghanbari-Siahkali, N.B. Larsen, Studies of spin-coated poly-mer films, Annual Reports on the Progress of Chemistry Section C 101 (2005)174–201.

3 urface

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

84 K.S. Kontturi et al. / Applied S

[7] M. Pohjakallio, T. Aho, K. Kontturi, E. Kontturi, Morphology of poly(methylmethacrylate) and polystyrene blends upon Langmuir–Schaefer deposition,Soft Matter 7 (2011) 743–748.

[8] G.A. Ozin, A.C. Arsenault, L. Cademartiri, Chemical patterning and lithography,in: Nanochemistry: A Chemical Approach to Nanomaterials, 2nd edn, RoyalSociety of Chemistry, Cambridge, 2008, pp. 57–111.

[9] D.U. Ahn, Z. Wang, R. Yang, Y. Ding, Hierarchical polymer patterns driven bycapillary instabilities at mobile and corrugated polymer–polymer interfaces,Soft Matter 6 (2010) 4900–4907.

10] G.G. Baralia, C. Filiâtre, B. Nysten, A.M. Jonas, Nanodecoding by dewetting,Advanced Materials 19 (2007) 4453–4459.

11] H.-C. Kim, S.-M. Park, W.D. Hinsberg, Block copolymer based nanostructures:materials, processes, and applications to electronics, Chemical Reviews 110(2010) 146–177.

12] C. Amiel, M. Sikka, J.W. Schneider Jr., Y.-H. Tsao, M. Tirrell, J.W. Mays, Adsorptionof hydrophilic–hydrophobic block copolymers on silica from aqueous solu-tions, Macromolecules 28 (1995) 3125–3134.

13] I.W. Hamley, S.D. Connell, S. Collins, In situ atomic force microscopy imagingof adsorbed block copolymer micelles, Macromolecules 37 (2004) 5337–5351.

14] O. Théodoly, M. Jacquin, P. Muller, S. Chhun, Adsorption kinetics of amphiphilicdiblock copolymers: from kinetically frozen colloids to macrosurfactants, Lang-muir 25 (2009) 781–793.

15] X. Liu, D. Wu, S. Turgman-Cohen, J. Genzer, T.W. Theyson, O.J. Rojas, Adsorp-tion of a nonionic symmetric triblock copolymer on surfaces with differenthydrophobicity, Langmuir 26 (2010) 9565–9574.

16] M.D. Rodowogin, C.S. Spanjers, C. Leighton, M.A. Hillmyer,Polylactide–poly(dimethylsiloxane)–polylactide triblock copolymers asmultifunctional materials for nanolithographic applications, ACS Nano 4(2010) 725–732.

17] C. Wang, D.H. Lee, A. Hexemer, M.I. Kim, W. Zhao, H. Hasegawa, H. Ade, T.P.Russell, Defining the nanostructured morphology of triblock copolymers usingresonant soft X-ray scattering, Nano Letters 11 (2011) 3906–3911.

18] E. Rakhmatullina, W. Meier, Solid-supported block copolymer membranesthrough interfacial adsorption of charged block copolymer vesicles, Langmuir24 (2008) 6254–6261.

19] A. Vesterinen, S. Lipponen, J. Rich, J. Seppälä, Effect of block compositionon thermal properties and melt viscosity of poly[2-(dimethylamino)ethylmethacrylate], poly(ethylene oxide) and poly(propylene oxide) block co-polymers, Express Polymer Letters 5 (2011) 754–765.

20] A. Vesterinen, J. Rich, J. Seppälä, Synthesis and solution rheology ofpoly[(stearyl methacrylate)-stat-([2-(methacryloyloxy)ethyl] trimethylammonium iodide)], Journal of Colloid and Interface Science 351 (2010)478–484.

21] M. Rodahl, F. Höök, A. Krozer, P. Brzezinski, B. Kasemo, Quartz crystal microbal-ance setup for frequency and Q-factor measurements in gaseous and liquidenvironments, Review of Scientific Instruments 66 (1995) 3924–3930.

22] G. Sauerbrey, Verwendung von Schwingquarzen zur Wägung dünner Schichtenund Microwägung, Zeitschrift für Physik 155 (1959) 206–222.

23] F. Höök, M. Rodahl, P. Brzezinski, B. Kasemo, Energy dissipation kinetics forprotein and antibody–antigen adsorption under shear oscillation on a quartzcrystal microbalance, Langmuir 14 (1998) 729–734.

24] L.-S. Johansson, Monitoring fibre surfaces with XPS in papermaking processess,Microchimica Acta 138 (2002) 217–223.

25] Y. Morishima, Amphiphilic Polyelectrolytes: Characterization of AssociativeProperties and Self-Assembled Nanostructures in Water, ACS, Washington DC.,2006, pp. 19–46 (Chapter 2).

26] J.F. Marko, Y. Rabin, Microphase separation in charged diblock copolymers: theweak segregation limit, Macromolecules 24 (1991) 2134–2136.

27] I. Astafieva, K. Khougaz, A. Eisenberg, Micellization in block polyelectrolytesolutions. 2. Fluorescence study of the critical micelle concentration as a func-tion of soluble block length and salt concentration, Macromolecules 28 (1995)7127–7134.

28] W.S. Tan, R.E. Cohen, M.F. Rubner, S.A. Sukhishvili, Temperature-induced,reversible swelling transitions in multilayers of a cationic triblock copolymerand a polyacid, Macromolecules 43 (2010) 1950–1957.

29] X. Cui, S. Mao, M. Liu, H. Yuan, Y. Du, Mechanism of surfactant micelle formation,Langmuir 24 (2008) 10771–10775.

30] M. Beija, A. Fedorov, M.-T. Charreyre, J.M.G. Martinho, Fluorescence anisotropyof hydrophobic probes in poly(N-decylacrylamide)-block–poly(N,N-diethylacrylamide) block copolymer aqueous solutions: evidence ofpremicellar aggregates, Journal of Physical Chemistry B 114 (2010)9977–9986.

[

[

Science 261 (2012) 375– 384

31] Y. Zhan, W.L. Mattice, Self-assembly and adsorption of diblock copolymers fromselective solvents. 1. Self-assembly, Macromolecules 27 (1994) 677–682.

32] P.-H. Ni, Q.-S. Pan, L.-S. Zha, C.-C. Wang, A. Elaïssari, S.-K. Fu, Syntheses and char-acterizations of poly[2-(dimethylamino)ethyl methacrylate]–poly(propyleneoxide)–poly[2-(dimethylamino)ethyl methacrylate] ABA triblock copolymers,Journal of Polymer Science Part A: Polymer Chemistry 40 (2002) 624–631.

33] P. Petrov, C.B. Tsvetanov, R. Jérôme, Two-component, onionlike micelles witha PPO core, PDMAEMA shell and a PEO corona: formation and crosslinking,Polymer International 57 (2008) 1258–1264.

34] M. Sedlák, What can be seen by static and dynamic light scattering in polyelec-trolyte solutions and mixtures? Langmuir 15 (1999) 4045–4051.

35] G.J. Fleer, M.A. Cohen Stuart, J.M.H.M. Scheutjens, T. Cosgrove, B. Vincent, Poly-mers at Interfaces, Chapman & Hall, London, 1993 (Chapter 4).

36] G.H. Bolt, Determination of the charge density of silica sols, Journal of PhysicalChemistry 61 (1957) 1166–1169.

37] Th.F. Tadros, J. Lyklema, Adsorption of potential-determining ions at thesilica–aqueous electrolyte interface and the role of some cations, Journalof Electroanalytical Chemistry and Interfacial Electrochemistry 17 (1968)267–275.

38] F.A. Rodrigues, P.J.M. Monteiro, G. Sposito, Surface charge density of silica sus-pended in water–acetone mixtures, Journal of Colloid and Interface Science 211(1999) 408–409.

39] L. Nurmi, S. Holappa, A. Nykänen, J. Laine, J. Ruokolainen, J. Seppälä, Ultra-thin films of cationic amphiphilic poly(2-(dimethylamino)ethyl methacrylate)based block copolymers as surface wettability modifiers, Polymer 50 (2009)5250–5261.

40] X. Liu, A.-H. Vesterinen, J. Genzer, J.V. Seppälä, O.J. Rojas, Adsorption ofPEO–PPO–PEO triblock copolymers with end-capped cationic chains of poly(2-dimethylaminoethyl methacrylate), Langmuir 27 (2011) 9769–9780.

41] B.C. Bonekamp, J. Lyklema, Conductometric and potentiometric monitoring ofpolyelectrolyte adsorption on charged surfaces, Journal of Colloid and InterfaceScience 113 (1986) 67–75.

42] D.E. Yates, T.W. Healy, The structure of the silica/electrolyte interface, Journalof Colloid and Interface Science 55 (1976) 9–19.

43] C.G.P.H. Schroën, M.A. Cohen Stuart, K. van der Voort Maarschalk, A. van derPadt, K. van’t Riet, Influence of preadsorbed block copolymers on proteinadsorption: surface properties, layer thickness, and surface coverage, Langmuir11 (1995) 3068–3074.

44] P. Alexandridis, T.A. Hatton, Poly(ethylene oxide)–poly(propyleneoxide)–poly(ethylene oxide) block copolymer surfactants in aqueoussolutions and at interfaces: thermodynamics, structure, dynamics, andmodeling, Colloids and Surfaces A 96 (1995) 1–46.

45] Y. Li, H. Liu, J. Song, O.J. Rojas, J.P. Hinestroza, Adsorption and association of asymmetric PEO–PPO–PEO triblock copolymer on polypropylene, polyethylene,and cellulose surfaces, ACS Applied Materials & Interfaces 3 (2011) 2349–2357.

46] Y. Samoshina, T. Nylander, P. Claesson, K. Schillén, I. Iliopoulos, B. Lindman,Adsorption and aggregation of cationic amphiphilic polyelectrolytes on silica,Langmuir 21 (2005) 2855–2864.

47] P. Brandani, P. Stroeve, Adsorption and desorption of PEO–PPO–PEO triblockcopolymers on a self-assembled hydrophobic surface, Macromolecules 36(2003) 9492–9501.

48] H. Cheng, M. Olvera de la Cruz, Hydrophobic-charged block copolymer micellesinduced by oppositely charged surfaces: salt and pH dependence, Macro-molecules 39 (2006) 1961–1970.

49] J.-F. Liu, A. Ducker, Surface-induced phase behavior of alkyltrimethylammo-nium bromide surfactants adsorbed to mica, silica, and graphite, Journal ofPhysical Chemistry B 103 (1999) 8558–8567.

50] S. Förster, N. Hermsdorf, W. Leube, H. Schnablegger, M. Regenbrecht, S. Akari,P. Lindner, C. Böttcher, Fusion of charged block copolymer micelles into toroidnetworks, Journal of Physical Chemistry B 103 (1999) 6657–6668.

51] M. Regenbrecht, S. Akari, S. Förster, H. Möhwald, Fusion of micelles ofpoly(butadiene-block-2-vinylpyridene) and derivatives on different sub-strates, Surface and Interface Analysis 27 (1999) 418–421.

52] C. Cai, L. Zhang, J. Lin, L. Wang, Self-assembly behavior of pH- and thermosen-sitive amphiphilic triblock copolymers in solution: experimental studies andself-consistent field theory simulations, Journal of Physical Chemistry B 112(2008) 12666–12673.

53] C. Ligoure, Surface micelles formation by adsorption of block copolymers,Macromolecules 24 (1991) 2968–2972.

54] L.M. Grant, F. Tiberg, W.A. Ducker, Nanometer-scale organization of ethyleneoxide surfactants on graphite, hydrophilic silica, and hydrophobic silica, Journalof Physical Chemistry B 102 (1998) 4288–4294.