Surfactant Distribution in Waterborne Acrylic Films. 1. Bulk Investigation

Page 1

S2

CPa

b

a

ARRAA

KLSSSACXA

1

acalfiEaqbas[

Cola

0d

Colloids and Surfaces A: Physicochem. Eng. Aspects 374 (2011) 58–68

Contents lists available at ScienceDirect

Colloids and Surfaces A: Physicochemical andEngineering Aspects

journa l homepage: www.e lsev ier .com/ locate /co lsur fa

urfactant distribution in waterborne acrylic films. Surface investigation

éline Arnolda, Guillaume Kleina, Mounir Maalouma, Marie Ernstssonb, Anders Larssonb,ascal Mariea, Yves Holl a,∗

Institut Charles Sadron, Université de Strasbourg, CNRS UPR 22, 23 rue du Loess, BP 84047, 67034 Strasbourg Cedex 2, FranceInstitute for Surface Chemistry, Box 5607, SE-114 86 Stockholm, Sweden

r t i c l e i n f o

rticle history:eceived 14 September 2010eceived in revised form 28 October 2010ccepted 2 November 2010vailable online 12 November 2010

eywords:

a b s t r a c t

This paper presents results on sodium dodecyl sulfate (SDS) enrichment at the surface of pure acrylicor acrylic/laponite composite latex films. Surface concentrations were measured by Confocal RamanSpectroscopy leading to higher values than the nominal concentration of 6 wt%. X-ray PhotoelectronSpectroscopy (XPS) analyses showed uppermost surface layers saturated with SDS in most cases. Highresolution Atomic Force Microscopy (AFM) revealed a variety of morphologies for these surfactant toplayers, highlighting the occurrence of SDS bilayers in different configurations. In an attempt to check

atex filmurfactant distributionurface enrichmentDSdsorption isothermonfocal Raman Spectroscopy

for a correlation between the surface concentration of the surfactant in dry films and the concentrationof free surfactant in water in the initial latex, this latter concentration was determined from the levelof the plateau in adsorption isotherms. Adsorption studies by conductimetry showed an unexpectedincrease of the amount of adsorbed SDS with pH. The proposed interpretation is that, upon acrylic acidneutralization, the chains at the surface become more hydrophilic and spread out in water, revealing

rb onat the

PS

FM

more sites for SDS to adsoestablished, indicating th

. Introduction

Distribution of surfactants in latex films has been recognizeds a topic of major importance in colloid and polymer physicalhemistry for many decades. Several authors, for instance Routhnd co-workers [1], quote the article by Wagner and Fischer pub-ished in 1936 [2] as the first dealing with surfactants in latexlms. Hundreds of papers have appeared since this pioneer work.xcellent review articles can be found in the literature [3–5]. Inddition, the introductions of several papers [1,6] provide good anduickly read overviews on the topic. Non-uniform surfactant distri-utions affect film properties in many ways [5], for examples: glossnd appearance [7], aesthetic qualities and dirt pick-up [8], adhe-ion and viscoelasticity [9], barrier properties and water whitening10,11], minimum film formation temperature (MFFT) [12].

Recently, we have addressed this issue again [13] mainly using

onfocal Raman Spectroscopy in order to establish the distributionf sodium dodecyl sulfate (SDS) in the bulk of acrylic and compositeatex films, focusing on the effects of particle composition and size,nd pH of the latex. The shapes of the concentration profiles were

∗ Corresponding author. Tel.: +33 3 88 41 41 16; fax: +33 3 88 41 40 99.E-mail address: [email protected] (Y. Holl).

927-7757/$ – see front matter © 2010 Elsevier B.V. All rights reserved.oi:10.1016/j.colsurfa.2010.11.008

. No correlation between free surfactant and surface enrichment could beenrichment process is more complex than expected.

© 2010 Elsevier B.V. All rights reserved.

discussed in terms of latex stability at short interparticle distancesand a model was developed based on the diffusion of the surfactantand its transport by the drying front to account for the formation ornot of aggregates in the bulk of the film. Then, we more specificallyraised the question of surfactant migration toward the film surfacein the same systems and this is the subject of the present paper.

Surfactant enrichment at the surface of waterborne polymerfilms has already been reported many times in the literature[14–18], although depletion, more rarely mentioned, is also pos-sible [19]. Mechanisms for surfactant migration to the surface oflatex films are still in debate. Most of the accumulation occurs inthe wet stage, during drying, even if further enrichment in the dryfilm might be observed [4], due to phase separation between theincompatible polymer and surfactant species. Convection, diffu-sion, surfactant adsorption on particles characterized by adsorptionisotherms, all these phenomena play a role in a complex interplay.Undoubtedly, the problem deserves further research effort.

Our contribution reported in this paper deals with surfactantcharacterization at the film surface: concentration measurementsby Confocal Raman Spectroscopy and X-ray Photoelectron Spec-

troscopy (XPS) and topographic observation by Atomic ForceMicroscopy (AFM). We also wanted to test an often encounteredsimple idea that surfactant enrichment might be directly correlatedwith the concentration of free surfactant in the latex serum. For thatpurpose, the distribution of the surfactant in the latex was studied

Page 2

C. Arnold et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 374 (2011) 58–68 59

epared

tip

2

2

mpt1mAreuscdhiiC5s[

AtdaabcT

TC

Fig. 1. AFM image of a film pr

hrough adsorption isotherms established by conductimetry allow-ng us to quantitatively distinguish the surfactant adsorbed at thearticle water interface and the surfactant free in water.

. Experimental

.1. Latices and films

Two kinds of latices were used in this study: pure poly-er and organic/inorganic composite systems. For the pure

olymer latices, the particles are composed of various propor-ions of butyl acrylate (BuA), methyl methacrylate (MMA) and% acrylic acid (AA). The surfactant was SDS and the poly-erization was initiated by ammonium persulfate ((NH4)2S2O8).ll reagents (from Merck; Darmstadt, Germany) were used aseceived. The syntheses were performed by a semicontinuousmulsion polymerization process, in a double-wall glass reactor,nder inert gas atmosphere. Two core-shell latices were synthe-ized, differing by the composition and the Tg of the particleore: either n-butyl acrylate (Tg = −41 ◦C), or a BuA/MMA ran-om copolymer (59.5% BuA, 39.5% MMA, Tg = 7 ◦C). All particlesad a diameter of 110 nm and contained 1% of acrylic acid, which

s mainly found in the shell of the particles [20]. By adapt-ng the synthesis process and using Abex 2005 (Rhodia, Inc.,ranbury, NJ, USA) as a surfactant, a nanolatex (BuA/MMA/AA9.5/39.5/1) with particles of 30 nm in diameter could be synthe-ized. More details about the synthesis can be found in reference21].

The composite latices are composed of 50% BuA and 50% MMA.n inorganic filler, laponite, is grafted at the surface of the latex par-

icles. Laponite is a synthetic silicate clay made of platelets with aiameter of 30 nm and a thickness of 1 nm. The faces of the platelets

re negatively charged whereas the edges are positively chargedt a pH lower than 9. The laponite content varied from 0% to 7%ased on monomer weight, leading to particles with a diameteromprised between 98 and 165 nm depending on the clay content.he presence of laponite platelets at the surface of the latex parti-

able 1omposition and characteristics of the latices used to form the films analyzed in this stud

Latex Composition (% BuA/% MMA/% AA)

BuA/MMA1 110 pH 10 59.5/39.5/1BuA/MMA1 110 pH 3 59.5/39.5/1BuA/MMA1 30 pH 10 59.5/39.5/1BuA1 110 pH 10 99/0/1BuA1 110 pH 3 99/0/1LCPP 0% 50/50/0LCPP 3% 50/50/0 + 3% laponiteLCPP 5% 50/50/0 + 5% laponiteLCPP 7% 50/50/0 + 7% laponite

from LCPP 7% dialysed latex.

cles was checked by AFM on a dry film made from a dialysed latexcontaining 7% laponite (Fig. 1). More details about the synthesis canbe found in reference [22].

The latices were purified by dialysis using a Millipore membraneuntil the conductivity of the water in contact with the latex wasless than 3 �S/cm. Purification allows elimination of water solubleimpurities (residual salts, oligomers, and surfactant). After dialysisthe solids content of the latices was adjusted to around 25%, exceptfor the nanolatex whose solids content was close to 15%. 6% of SDSbased on total solids content was added. The pH of the pure poly-mer latices was then adjusted to 3 or 10 by adding either dilutedhydrochloric acid or sodium hydroxide aqueous solutions. The pHof the composite latices was not adjusted and remained between7 and 8 after dialysis. The characteristics of the different latices aregathered in Table 1.

2.2. Adsorption isotherms

The adsorption isotherms of sodium dodecyl sulfate on latexparticles were determined by conductimetry. In a first step, theconductivity of solutions containing various amounts of SDS wasmeasured. A plot of the conductivity as a function of SDS concen-tration was obtained. The curves could be divided into two straightlines, below and above CMC (7.4 × 10−3 mol/L). The calibration wasmade for each studied pH.

The free surfactant in latices was determined using the calibra-tion curve: knowing the quantity of surfactant added to the latexand measuring the amount of free surfactant from the calibration,it was possible to deduce the quantity of adsorbed surfactant. Na+

counterions from the adsorbed surfactant were not considered inour calculation although they participate to the total conductivity.It was checked by another method (surface tension measurement)

[21] that neglecting the sodium counterions had no significanteffect on the final results.

Measurements were performed with 100 mL of dialysed latex at4% solids content under magnetic stirring. 1 mL of a 0.1 mol L−1 SDSaqueous solution was added every 3 min. The pH of the latex was

y.

Solids content Mean particle diameter (nm) pH

24.9% 114 1024.4% 112 315.1% 32 1024.8% 116 1023.6% 113 323.5% 98 724.9% 128 725.0% 148 823.4% 165 8

Page 3

6 Physicochem. Eng. Aspects 374 (2011) 58–68

maCa

2

aviflxc

Hotlsnatcipt

ootH

2

cem(1a

sdsp

tcd

2

ozcwwkok

0.0150.0100.0050.000

0.0

0.5

1.0

1.5

2.0

2.5

Γ (µ

mo

l/m

2)

C surfactant solution (mol.L-1

)

pH 9 and 2 �mol/m2 at pH 3) than for BuA/MMA1 (2 �mol/m2 atpH 9 and 1.3 �mol/m2 at pH 3).

The adsorption plateaus obtained with the two previous sys-tems containing acrylic acid are compared to values found in the

0.0150.0100.0050.000

0

1

2

3

4

5

Γ (µ

mo

l/m

2)

0 C. Arnold et al. / Colloids and Surfaces A:

easured with a Titrilab 80 pHmeter (Radiometer Copenhagen)nd the conductivity with a CDM 210 conductimeter (Radiometeropenhagen). The pHs of the latex and of the SDS solution weredjusted before titration.

.3. Confocal Raman Spectroscopy

This technique (also called Confocal Raman Microscopy) allowslocal Raman analysis of a transparent sample. The investigated

olume (typically 1 �m3) can be moved at will along the three axes:n the plane x, y of the surface and along the z axis from the sur-ace to the substrate. For the surface analysis presented here, theaser was focalised at the surface of the films and moved along theand y directions in order to follow the evolution of the surface

omposition from the centre to the edges of the sample.Raman measurements were performed with a Jobin Yvon

oriba spectrometer (LABRAM BX40). Excitation at a wavelengthf 632.81 cm−1 was provided by a He–Ne laser. The diameter ofhe confocal pinhole was 100 �m and a ×100 objective was used,eading to a spatial resolution of about 2 �m. For each point, 8–10pectra were accumulated during 120 s, providing a satisfying sig-al to noise ratio. Raman spectroscopy can be used for quantitativenalysis since the Raman signal is proportional to the concentra-ion. Thus, the surfactant concentration relative to the total solidsan be measured in the films, after calibration. By plotting thentensity of one SDS peak (S O peak at 1086 cm−1) relative to oneolymer peak (C C at 1063 cm−1) as a function of the SDS concen-ration, a calibration curve was obtained.

The films were prepared on quartz plates by depositing 300 �Lf latex with a micropipette, leading to films with a mean thicknessf about 60 �m. The films were dried at least 10 days, but not morehan 15 days under controlled conditions (22 ± 2 ◦C and 55 ± 5%R).

.4. X-ray Photoelectron Spectroscopy (XPS)

XPS, also known as ESCA (Electron Spectroscopy for Chemi-al Analysis), was used to get the chemical surface compositionxpressed in atm% for a 10 nm thick surface layer. Measure-ents were performed with a Kratos AXIS UltraDLD spectrometer

Kratos Analytical, Manchester, UK), at a pressure of 10−9 to0−8 Torr during analysis (ultra high vacuum conditions) and usingmonochromatic Al K� X-ray source.

XPS small spot analysis using the 55 �m aperture (with mostignal from an area with diameter 55 �m), was used to analyzeifferent points at the surface of the films. The points are numberedtarting from 0, at the centre of the film. The distance between theoints, from 1 to 3.5 mm, is indicated in the result tables.

In the analysis, low-resolution wide (survey) spectra were runo detect elements present in the surface layer. The relative surfaceompositions were obtained from quantification of low-resolutionetail spectra run for each element.

.5. AFM

Atomic Force Microscopy was used to visualize the morphologyf the top surface of the films with a special focus on the organi-ation of the surfactant. A multimode AFM with a Nanoscope IVontroller from Veeco (Digital Instrument, Santa Barbara, CA, USA)

as used. The measurements were performed in tapping modeith a silicon tip with a maximal radius of curvature of 10 nm. Two

inds of cantilever were used, either with a resonant frequency f0f 75 kHz and a spring constant k of 5 N/m or with f0 = 300 kHz and= 40 N/m.

Fig. 2. Adsorption isotherms of SDS on BuA/MMA1 latex particles for different pHvalues (�: pH 9; �: pH 4; �: pH 3).

3. Results and discussion

3.1. Adsorption isotherms

Adsorption isotherms were measured in order to quantify therepartition of the surfactant between the serum and the surface ofthe particles. In Fig. 2, the results obtained with the BuA/MMA1particles are presented for different pH values. The graphs clearlyshow that once the particles are saturated with surfactant, a plateauis reached on the isotherms. The transition between the growingpart of the curve and the plateau corresponds to the CMC of SDS(7.4 × 10−3 mol L−1). The first part of the isotherms is not muchinfluenced by pH whereas the level of the plateau increases withthe pH value.

With BuA1 particles (Fig. 3), the level of the plateau alsoincreases with pH. The difference between basic (pH 9) and acidic(pH 3 and 4) pHs is more pronounced than with the BuA/MMA1particles. The plateau is always higher for BuA1 (4.5 �mol/m2 at

C surfactant solution (mol.L-1

)

Fig. 3. Adsorption isotherms of SDS on BuA1 latex particles for different pH values(�: pH 9; �: pH 4; �: pH 3).

Page 4

C. Arnold et al. / Colloids and Surfaces A: Physi

0.0150.0100.0050.000

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0Γ

(µm

ol/

m2)

C surfactant solution (mol.L-1

)

F0

lotfiss[3

etmrhwptsBtp

mwtlcntal

increasing the softness of the particles, from BuA/MMA (LCPP 0%)

TD

ig. 4. Adsorption isotherms of SDS on laponite (�), acrylic latex particles (� LCPP%), and composite latex particles (� LCPP 5%).

iterature. Lam et al. [23] also measured SDS adsorption isothermsn particles containing acrylic acid. In their case, the core ofhe particles was composed of butyl acrylate and styrene. Theyound a maximum adsorption of 2 mg/m2 for SDS, correspond-ng to 10 �mol/m2. This value is higher than ours but still in theame order. Results obtained on particles with a different compo-ition give values similar to ours, for example Vale and McKenna24] found the adsorption plateau of SDS on PVC particles at.5 �mol/m2 and Lin et al. [25] found 3 �mol/m2 on polystyrene.

In a classical study of SDS adsorption on latex particles of differ-nt natures [26], Vijayendran stated that SDS adsorbed even morehan the surface was more hydrophobic (less polar). For example,

ore SDS adsorbed on PVC than on PVAc. This rule is obeyed by ouresults when going from BuA/MMA1 to BuA1, as the latter, moreydrophobic, adsorbs more SDS than the former. However, it is nothen pH effects are considered: adsorption always increases withH. Our interpretation is that, by neutralizing the acrylic acid func-ions, the chains at the surface spread out in water, revealing moreites for SDS to adsorb on. The pH effect is more pronounced foruA1 than for BuA/MMA1 because the acrylic acid (AA) concentra-ion at the surface of BuA1 is higher: 89% of the total AA is in thearticle shell for BuA1 and only 63% for BuA/MMA1 [20].

Adsorption isotherms of SDS on LCPP systems were also deter-ined. In addition, the adsorption isotherm of SDS on pure laponiteas measured on a 4 wt% laponite suspension. It is clear from Fig. 4

hat SDS hardly adsorbs on laponite, the level of the plateau is veryow (0.75 �mol/m2). Anionic SDS does not adsorb on negativelyharged surfaces and the positively charged edges of laponite do

ot offer a great number of adsorption sites. Adsorbing SDS onhe BuA/MMA pure polymer particles leads to a saturation plateaut a value of 3.4 �mol/m2. Since adsorption of SDS on the clay isow compared to the adsorption on the copolymer, it is expected

able 2istribution of the surfactant in the initial acrylic latices, calculated from the adsorption

Latex � plateau (�mol/m2)

BuA/MMA1 pH 10 110 nm 2.0BuA/MMA1 pH 10 30 nm 2.0BuA/MMA1 pH 3 110 nm 1.3BuA1 pH 10 110 nm 4.5BuA1 pH 3 110 nm 2.0LCPP 0% 3.4LCPP 5% 2.8

cochem. Eng. Aspects 374 (2011) 58–68 61

that grafting laponite at the surface of the particles will reduce theadsorbed amount. Indeed, the saturation plateau of the compositelatex LCPP 5% is at 2.8 �mol/m2.

From the level of the saturation plateau and knowing the ini-tial particle volume fraction, the mean particle diameter and theSDS content (6% based on total solids weight), it was possible tocalculate the repartition of the SDS in the initial latices. The valuesare gathered in Table 2. There is a logical relationship between thevalue of the plateau and the surfactant concentration in solution:the higher the plateau, the lower the free surfactant concentration.

3.2. Confocal Raman Spectroscopy

The evolution of the surfactant concentration at the surface of aBuA/MMA1 110 nm pH 10 6% SDS film is given in Fig. 5. The surfaceof the film presents a surfactant enrichment. This enrichment ismore pronounced along the Y than along the X axis and increasesfrom the centre to the edge of the film.

The higher concentration at the edge may be related to the wellknown “coffee stain effect” [27,28] where particles are driven to theedges by water flowing from the centre to compensate for higherevaporation rates in areas where the film is thinner. In this case,the surfactant would be conducted outwards with the particles.One can argue that the lateral drying front might be able to bringthe surfactant back to the centre [29] unless it is trapped in a tooconcentrated area. Therefore, the interpretation of our observationis not straightforward.

The film prepared from the BuA/MMA1 nanolatex (Fig. 6)presents a low surface enrichment compared to the total SDS con-centration in the latex (6%).

Films made from softer latex particles (BuA1 pH 10 6% SDS)present a strong surface enrichment (Fig. 7) compared to the pre-vious cases obtained at the same pH but with more rigid particles.

The effect of the pH on the surface enrichment has been checkedby decreasing its value from 10 to 3. A film made from BuA/MMA16% SDS pH 3 (Fig. 8) latex presents a strong surface enrichmentwhereas the softer particles lead to a low enrichment at pH 3 (Fig. 9).

The SDS concentration at the surface of films prepared fromthe LCPP series was also investigated. The blank LCPP reference,BuA/MMA 50/50 without laponite, presents a slight and almostconstant enrichment (Fig. 10). Grafting laponite at the surfaceof the particles leads to an increase in the surface enrichment(Figs. 11 and 12). For LCPP with 7% laponite the SDS enrichmentalong the X axis increases from the centre to the edge of the film(Fig. 12).

The mean concentration at the surface of the different filmswas calculated, the values are gathered in Table 3. The surfactantconcentration at the surface of the films is systematically higherthan the 6% SDS in the initial latices. It seems that at basic pH

to BuA/MMA1 and finally BuA1 leads to an increase of the sur-face enrichment. At constant particle composition, decreasing theirsize decreases the SDS surface concentration. Decreasing the pHleads to an increase or a decrease of the surfactant concentration

isotherms.

%adsorbed/%solution CSDS solution

42%/58% 40 mM = 5 cmc86%/14% 6 mM = 0.75 cmc27%/73% 50 mM = 6.25 cmc83%/17% 12 mM = 1.5 cmc40%/60% 41 mM = 5.1 cmc80%/20% 13 mM = 1.6 cmc37%/63% 39 mM = 4.9 cmc

Page 5

62 C. Arnold et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 374 (2011) 58–68

Fig. 5. SDS concentration at the surface of a BuA/MMA1 110 nm pH 10 + 6% SDS film as a function of the position.

SDS along X axis

5

7

9

11

13

2520151050

X (mm)

% SDS

SDS along Y axis

0

2

4

6

8

1311975

% SDS

Y (

mm

)

Fig. 6. SDS concentration at the surface of a BuA/MMA1 30 nm pH 10 + 6% SDS film, along the X and Y axes (see Fig. 5 for the position of the axes).

SDS along X axis

5

10

15

20

25

2824201612840

X (mm)

% SDS

SDS along Y axis

-10

-5

0

5

10

252015105

% SDS

Y (

mm

)

Fig. 7. SDS concentration at the surface of a BuA1 110 nm pH 10 + 6% SDS film, along the X and Y axes.

SDS along X axis

5

10

15

20

25

24201612840

X (mm)

% SDS

SDS along Y axis

0

1

2

3

4

5

6

7

252015105

% SDS

Y (

mm

)

Fig. 8. SDS concentration at the surface of a BuA/MMA1 110 nm pH 3 + 6% SDS film, along the X and Y axes (see Fig. 5 for the position of the axes).

Page 6

C. Arnold et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 374 (2011) 58–68 63

SDS along X axis

5

7

9

11

13

302520151050

X (mm)

% SDS

SDS along Y axis

0

2

4

6

8

10

1311975

% SDS

Y (

mm

)

Fig. 9. SDS concentration at the surface of a BuA1 110 nm pH 3 + 6% SDS film, along the X and Y axes.

SDS along X axis

5

7

9

11

13

25155-5-15-25

X (mm)

% SDS

SDS along Y axis

-7

-5

-3

-1

1

3

5

7

1311975

% SDS

Y (

mm

)

Fig. 10. SDS concentration at the surface of a LCPP 0% pH 7 + 6% SDS film, along the X and Y axes.

SDS along X axis

5

7

9

11

13

24201612840

X (mm)

% SDS

SDS along Y axis

-5

0

5

10

1311975

% SDS

Y (

mm

)

Fig. 11. SDS concentration at the surface of a LCPP

SDS along X axis

5

10

15

20

262218141062-2-6-10-14-18-22-26

X (mm)

% SDS

Fig. 12. SDS concentration at the surface of a LCPP 7% pH 8 + 6% SDS film, along theX axis.

Table 3Mean SDS concentration at the surface of the dry waterborne polymer films.

Latex CSDS surface

BuA/MMA1 110 nm pH 10 8.1%BuA/MMA1 30 nm pH 10 6.6%BuA/MMA1 110 nm pH 3 13.2%BuA1 110 nm pH 10 15.2%BuA1 110 nm pH 3 6.5%LCPP 0% pH 7 6.9%LCPP 3% pH 7 8.1%LCPP 7% pH 8 12.8%

-10

3% pH 7 + 6% SDS film, along the X and Y axes.

depending on the composition of the particles. Finally, increasingthe laponite content at the surface of the particles leads to highersurface enrichment.

3.3. XPS analysis

The uppermost surface composition of the dry films was deter-mined by XPS. The results at the surface of the BuA/MMA1 110 nmpH 10 + 6% SDS film are presented in Table 4. Since hydrogen

is not detected, the expected atomic composition for pure SDS(C12H25OSO3Na) measured by XPS is: 66.7% C, 22.2% O, 5.6% S and5.6% Na. The carbon and oxygen contents at the surface correspondto those of pure SDS. The S content is slightly lower whereas slightly

Table 4Atomic composition, determined by XPS, of the surface of a BuA/MMA1 110 nm pH10 + 6% SDS film.

55 �m selected area – position: Relative surface composition (atm%)

Number (and distance between) C(1s) O(1s) Na(1s) S(2p)

0 centre of the film (1.6 mm) 66.0 22.6 6.4 4.91 (1.6 mm) 66.1 22.5 6.4 5.02 (3.5 mm) 65.7 22.6 6.7 5.03 (1.0 mm) 65.8 22.8 6.4 5.04 on edge of the film 65.6 22.7 6.7 5.0Pure SDS (theoretical values) 66.7 22.2 5.6 5.6

Page 7

64 C. Arnold et al. / Colloids and Surfaces A: Physi

Table 5Atomic composition, determined by XPS, of the surface of a LCPP 0% pH 7 + 6% SDSfilm.

55 �m selected area – position: Relative surface composition (atm%)

Number (and distance between) C(1s) O(1s) Na(1s) S(2p)

0 centre of the film (2.3 mm) 70.6 21.6 4.4 3.41 (1.5 mm) 71.5 20.8 4.6 3.22 (1.7 mm) 70.7 21.4 4.6 3.3

mbhepi

1

icTrTas

ibwfN1taic

octB

The phase image of a LCPP 0% pH 7 + 6% SDS film is presented in

TA

TA

3 (1.2 mm) 69.8 22.0 4.8 3.44 (1.7 mm) 69.5 21.6 5.0 3.75 on edge of latex film 69.4 22.0 4.9 3.7

ore Na than expected is found. The low S content can be explainedy the orientation of the surfactant molecule, the sulfate head beingidden under the hydrocarbon tail. The reason for the slight Nanrichment is not known, but could possibly be from adjusting theH with NaOH. The conclusion is that the 10 nm thick surface layer

s essentially composed of pure SDS, whatever the position.Similar results (not shown) were observed for BuA/MMA1

10 nm at pH 3.The carbon content measured at the surface of the LCPP 0% film

s higher than in pure SDS (Table 5) and the sulphur and sodiumontents are lower. The layer analyzed by XPS is thus not pure SDS.he average S and Na contents are 3.5% and 4.7% respectively, cor-esponding to slightly more than half of their proportions in SDS.he 10 nm thick layer analyzed by XPS seems to be composed ofbout 50% SDS, or slightly more. This correlates well with a lowurface enrichment as measured by confocal Raman (6.9%, Table 3)

A film containing 7% laponite was analyzed before and after rins-ng with water. Before rinsing (Table 6), only SDS was detectedy XPS. Silicon and magnesium, two elements present in laponite,ere not detected. Rinsing with deionised water removed the sur-

actant and revealed the clay (Table 7). The laponite formula is:a0.7

+[(Si8Mg5.5Li0.3)O20(OH)4]0.7−. The Si/Mg ratio is theoretically

.45, and slightly higher ratios of 1.9–2.0 are obtained (when usingwo decimals from raw data in calculations). Note that with lowmounts of Si and Mg (<1 atm%), and with a fairly high noise leveln detail spectra, the base line setting will strongly influence the cal-ulated values. Therefore, exact theoretic ratios cannot be expected.

XPS analyses confirmed the surfactant enrichment at the surface

f the films already established over a larger thickness by confo-al Raman. It was not possible with this highly surface sensitiveechnique to make a difference in the SDS enrichment between theuA/MMA1 (pH 3 and 10) and LCPP 7% films because only pure SDS

able 6tomic composition, determined by XPS, of the surface of a LCPP 7% pH 8 + 6% SDS film.

55 �m selected area–position: Relative surface composition (atm%)

Number (and distance between) C(1s) O(1s)

0 centre of the film (1.5 mm) 65.1 23.21 (1.0 mm) 66.5 22.22 (1.0 mm) 64.9 23.73 (1.0 mm) 66.1 22.34 (1.0 mm) 65.7 22.85 on edge of the film 65.1 23.3

able 7tomic composition, determined by XPS, of the surface of a LCPP 7% pH 8 + 6% SDS film af

55 �m selected area–position: Relative surface composition (atm%)

Number (and distance between) C(1s) O(1s)

0 centre of latex film (2.2 mm) 75.9 22.41 (2.0 mm) 77.7 21.32 (1.5 mm) 77.6 21.53 close to edge of latex film 78.1 21.0

cochem. Eng. Aspects 374 (2011) 58–68

in the outermost 10 nm surface layer was detected in all cases. Suchsaturated surfactant layers were already observed by Zhao et al.[30] in the XPS analysis of films made from BuA/MMA particlesstabilized by SDS. One sample, LCPP 0%, presented a lower surfaceenrichment since a part of the signal came from the polymer. Rins-ing the composite films removed the surfactant and revealed thepolymer and the mineral filler.

3.4. AFM images

AFM was used in order to gain information about the organiza-tion of the surfactant layers at the surface of the films. The surface ofa BuA/MMA1 110 nm pH 10 + 6% SDS film (Fig. 13) presents elevatedzones separated by valleys. The valleys are horizontally separatedby more than 2 �m and the height difference between the high-est and lowest zones of the film is small since the Z-scale coversonly 20 nm. The phase image (Fig. 13 right) shows that a uniformlayer covers the surface of the film. According to Raman and XPSresults, no doubt that this layer is SDS. It is not totally uniformbecause the image shows different zones separated by dotted bor-ders. There is no clear explanation concerning the nature of theborders, whether they are due to the roughness of the sample orto a real phase contrast arising from the presence of the polymersublayer.

Similar almost continuous SDS layers were observed at the sur-face of BuA/MMA1 30 nm pH 10 + 6% SDS and BuA1 + 6% SDS pH 3and pH 10 films (images not shown).

The surface of a BuA/MMA1 110 nm pH 3 + 6% SDS was alsoinvestigated. The topography image in Fig. 14(left) presents a wavybackground supporting flat layers. The phase image (right) showsthat there is no phase contrast between the background and thelayers, indicating that both are surfactant. The profile of the topog-raphy image was drawn in order to determine the thickness of thelayers, and a value of 4 nm was found (Fig. 15). The carbon chain ofthe fully extended SDS molecule has a length of 1.7 nm according toTanford [31]. Adding the radius of sulfate ion (0.149 nm), we obtainan SDS chain length of 1.85 nm. An SDS bilayer would thus have athickness of about 3.7 nm, which is close to the value found at thesurface of the BuA/MMA1 110 nm pH 3 film.

Fig. 16. Uncoalesced latex particles are clearly visible in the dark-est zones. The particles are covered with an incomplete surfactantlayer corresponding to the brightest zones of the image. This find-ing agrees with the previously mentioned XPS data, indicating that

Na(1s) S(2p) Mg(2s) Si(2p)

6.5 5.2 – –6.2 5.1 – –6.3 5.1 – –6.4 5.2 – –6.4 5.1 – –6.5 5.1 – –

ter rinsing with deionised water.

Na(1s) S(2p) Mg(2s) Si(2p)

0.1 0.1 0.5 1.0– – 0.3 0.7– – 0.3 0.6– – 0.3 0.6

Page 8

C. Arnold et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 374 (2011) 58–68 65

Fig. 13. Topography (left) and phase (right) images of the surface of a BuA/MMA1 110 nm pH 10 + 6% SDS film.

of the

t5

oeb

pitUi

Fig. 14. Topography (left) and phase (right) images

he 10 nm thick layer analyzed by XPS consists of SDS only to about0% (or slightly more), and not a 100% fully covering SDS layer.

In order to determine the thickness of this SDS layer, the profilef a topography image was drawn (not shown). The height differ-nce was 4.3 nm, which again is close to the thickness of a SDSilayer.

Topography and phase images of a film made from the LCPP 7%

H 8 + 6% SDS latex are presented in Fig. 17. The topography image

ndicates an important height difference between the highest andhe lowest zones of the film, since the vertical scale is 133.5 nm.ncoalesced latex particles are visible in the lowest zones. The

mages suggest that SDS is organized in multilayers. The profile

Fig. 15. Profile of the surface of a BuA/M

surface of a BuA/MMA1 110 nm pH 3 + 6% SDS film.

of the topography image was drawn (Fig. 18). The surfactant formslayers since the profile of the brightest zone forms stairs. The heightof one stair is 4 nm, corresponding to a SDS bilayer.

In summary, several surfactant features can be observed at thesurface of the films studied in this work: an almost flat, continuoussurfactant layer (Fig. 13); a flat surfactant layer with an incompletebilayer on top (Fig. 14); an incomplete bilayer directly sitting on

latex particles (Fig. 16); an ensemble of “stepped pyramids” eachstep being one or a multiple of bilayers (Fig. 17).

Surfactants can be visualized on latex film surfaces either byelectron microscopy or by AFM. As early as in 1966, Bradford andVanderhoff [32] published high quality electron micrographs of

MA1 110 nm pH 3 + 6% SDS film.

Page 9

66 C. Arnold et al. / Colloids and Surfaces A: Physi

naMtmmhafv

Fig. 16. Phase image of the surface of a LCPP 0% pH 7 + 6% SDS film.

on ionic surfactants (ethoxylated nonylphenols) blobs (“blisters”,s they called them) on top of styrene/butadiene copolymer films.uch more recently, Environmental Scanning Transmission Elec-

ron Microscopy [33] was used for distribution studies, althoughore suited for bulk investigation. However, AFM provides the

ost valuable images of surfactant features at film surfaces. People

ave observed rather flat surfactant layers [23,34] or blobs (islands)nd finger like patterns [6,35]. For instance, Aramendia et al. [34]ound a massive SDS exudation to the surface of an acrylic filmery similar to our copolymer (BuA/MMA1) in the form of a con-

Fig. 17. Topography (left) and phase (right) images

Fig. 18. Profile of the surface of a

cochem. Eng. Aspects 374 (2011) 58–68

tinuous, wavy SDS layer resembling our image in Fig. 13 (althoughdirect comparison is not easy because they used a 3D representa-tion). These layers can be easily washed off with water, revealingrougher surfaces underneath [23]. In some instances, the surfactantat the surface can prevent particles from coalescing and it has to beremoved to allow particle deformation [36]. On the other hand, avariety of features were observed by Tzitzinou et al. [6], referred toas “blobs”, “finger-like strips”, “flat or rounded surfactant blobs”.Due to improvements in AFM resolution, more details, like SDSbilayers and multilayers (Figs. 14, 16 and 17), can now be imaged atthe surface of latex films. Similar lamellar structures aligned paral-lel to the substrate were observed by AFM on mica after drying ofSDS solutions [37].

3.5. Surface enrichment versus free surfactant in water

In order to facilitate the following discussion, the most rele-vant results presented above: concentration of free SDS in waterin the initial latex (from adsorption studies), surface concentrationof SDS in the dry films (Confocal Raman Spectroscopy), XPS andAFM results are gathered in Table 8.

More or less explicitly, the idea that there is a correlationbetween the surfactant free in water in the latex and the sur-

factant enrichment at the surface of the dry film exists in somearticles [19,38]. More free surfactant would lead to increased sur-face enrichment. The rational for that is the vision of a water fluxfrom bottom (substrate side) to top (air side) during drying, drivingthe surfactant to the surface by convection. Even if one suspects

of the surface of a LCPP 7% pH 8 + 6% SDS film.

LCPP 7% pH 8 + 6% SDS film.

Page 10

C. Arnold et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 374 (2011) 58–68 67

Table 8Summary of the main results presented so far.

Latex CSDS solution (cmc) CSDS surface (%) XPS AFM

BuA/MMA1 pH 10 110 nm 5 8.1 Surface layer saturated withSDS

Flat, uniform layer

BuA/MMA1 pH 10 30 nm 0.75 6.6 Flat, uniform layerBuA/MMA1 pH 3 110 nm 6.25 13.2 Surface layer saturated with

SDSFlat, uniform layer with partialcoverage of SDS bilayers

BuA1 pH 10 110 nm 1.5 15.2BuA1 pH 3 110 nm 5.1 6.5 Flat, uniform layerLCPP 0% 1.6 6.9 SDS covering ∼50% of surface SDS bilayer with holes showing

latex particles underneathLCPP 3% 8.1LCPP 5% 4.9LCPP 7% 12.8 Surface layer saturated with Rough surface. Stairlike peaks,

tcc(ostirsdpssisi

4

bcetssftmoTf

eflt

iSpmw

oim

[

[

[

hat reality is probably more complex, we found it interesting toheck for this correlation. Using data in Table 8 to plot the surfaceoncentration versus the concentration of free surfactant in waternot shown) immediately rules out the correlation. Especially outf scale are the points corresponding to BuA1, indicating that theoftness of the particle probably also plays a role. In fact, thinkinghat drying proceeds through a bottom to top motion of water isncorrect as there is no water reservoir on the substrate side. Theight vision is that of the water air interface receding toward theubstrate. The only way to “push” water upwards is by particlesiffusing back from the surface region. In most cases, this latterrocess is probably not efficient enough to significantly drive freeurfactant up. However, there is one example of clear evidence ofurfactant convection by water but in very special circumstances,n the case of a highly concentrated PDMS emulsion where dropletstrongly repel each other and coalesce immediately as they comento contact [39].

. Conclusion

This paper reports results about surfactant distribution in laticesefore film formation through adsorption isotherms measured byonductimetry. The composition of the particle shell has an influ-nce on the quantity of adsorbed surfactant. The BuA1 particleshat have more acrylic acid in the shell are able to adsorb moreurfactant than the BuA/MMA1 particles whatever the pH. In ourystems, increasing the pH leads to an increased adsorption of sur-actant. This was tentatively attributed to a partial solubilization ofhe hydrophilic chains and their spreading out in water, revealing

ore sites for SDS to adsorb on. Grafting laponite at the surfacef BuA/MMA particles leads to a decrease of surfactant adsorption.he quantity of adsorbed surfactant is inversely proportional to theree surfactant concentration in the aqueous phase.

In some cases, a high free SDS concentration led to a strong SDSnrichment at the surface of the films. But, the opposite was foundor the BuA1 system. Therefore, we had to rule out a general corre-ation between free surfactant and surface enrichment, indicatinghat the enrichment process is more complex than expected.

The surfactant organization at the surface of the films was stud-ed by AFM. Four features were observed: unorganized SDS layers;DS bilayers on an unorganized SDS layer; SDS bilayers on latexarticles and SDS multilayers on latex particles. SDS bilayers andultilayers have never been observed before at the surface of

aterborne polymer films.

As stated in the companion paper of this one [13], distributionf surfactants in latex films, in the bulk as well as at the surface,s a complex problem. No simple interpretation for surface enrich-

ent could be proposed. Pursuing efforts for a better understanding

[

[

SDS, removable by rinsing steps = SDS bilayers. Latexparticles visible in few holes

of the problem will lead to the possibility of designing systemswith improved properties. This is urgent because it is obvious thatwater based films, especially paints, still do not match the qual-ity of solventborne ones which tend to disappear from the marketfor regulation reasons. Research has to be continued in the threedirections of experimental work, theory and simulation.

Acknowledgments

This work was funded by the EC Framework 6 IntegratedProject, NAnostructured waterborne POLymEr films with Out-staNding properties (NAPOLEON) under Contract No. IP 011844-2.We thank Dr E. Bourgeat Lami and Dr V. Mellon, University of Lyon,France, for the synthesis of the composite latices (LCPP), Mr S. Greeat IS2M, University of Mulhouse for assistance in Raman analyses,Mr C. Contal, ICS Strasbourg, for assistance in AFM imaging. TheInstitute for Surface Chemistry in Stockholm, Sweden, gratefullyacknowledges the grant from Knut and Alice Wallenberg founda-tion for the XPS instrument.

References

[1] V.R. Gundabala, W.B. Zimmerman, A.F. Routh, A model for surfactant distribu-tion in latex coatings, Langmuir 20 (2004) 8721–8727.

[2] H. Wagner, G. Fischer, Kolloid A 77 (1936) 12.[3] J.L. Keddie, Film formation of latex, Mater. Sci. Eng. R 21 (1997) 101.[4] A.C. Hellgren, P. Weissenborn, K. Holmberg, Surfactants in water-borne paints,

Prog. Org. Coat. 35 (1999) 79–87.[5] J.L. Keddie, A.F. Routh, Surfactant distribution in latex films, in: Fundamentals

of Latex Film Formation, Springer, 2010, pp. 185–212.[6] A. Tzitzinou, P.M. Jenneson, A.S. Clough, J.L. Keddie, J.R. Lu, P. Zhdan, K.E.

Treacher, R. Satguru, Surfactant concentration and morphology at the surfaceof acrylic latex films, Prog. Org. Coat. 35 (1999) 89–99.

[7] J.M. Asua, H.A.S. Schoonbrood, Reactive surfactants in heterophase polymer-ization, Acta Polym. 49 (1998) 671–686.

[8] E. Jablonski, J. Hayes, T. Learner, M. Golden, Conservation concerns for acrylicemulsion paints, Rev. Conserv. 4 (2003) 3–12.

[9] Y.K. Yang, H. Li, F. Wang, Studies on the water resistance of acrylic emulsionPSAs, J. Adhes. Sci. Technol. 17 (2003) 1741.

10] P.A. Steward, J. Hearn, M.C. Wilkinson, Studies in permeation through polymerlatex films. 1. Films containing no or only low levels of additives, Polym. Int. 38(1995) 1–12.

11] L.N. Butler, C.M. Fellows, R.G. Gilbert, Effect of systems on the water sensitivityof latex films, J. Appl. Polym. Sci. 92 (2004) 1813–1823;L.N. Butler, C.M. Fellows, R.G. Gilbert, Effect of surfactants used for bindersynthesis on the properties of latex paints, Prog. Organic Coating 53 (2005)112–118.

12] G.A. Vandezande, A. Rudin, Film formation of vinyl acrylic latexes; effects ofsurfactant type, water and latex particle size, J. Coat. Technol. 68 (860) (1996)63–73.

13] C. Arnold, F. Thalmann, C. Marquès, P. Marie, Y. Holl, Surfactant distributionin waterborne acrylic films. 1. Bulk investigation, J. Phys. Chem. B 114 (2010)9135–9147.

14] C.L. Zhao, Y. Holl, T. Pith, M. Lambla, FTIR-ATR spectroscopic determinationof the distribution of surfactants in latex films, Colloid Polym. Sci. 265 (1987)823.

Page 11

6 Physi

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[[

[

8 C. Arnold et al. / Colloids and Surfaces A:

15] K.W. Evanson, M.W. Urban, Surface and interfacial FTIR spectroscopic studies oflatexes. 3. The effect of substrate surface tension and elongation on exudationof Surfactant, J. Appl. Polym. Sci. 42 (1991) 2309.

16] Y. Zhao, M.W. Urban, Mobility of SDOSS powered by ionic interactions in Sty/n-BA/MAA core-shell latex films. 21. A spectroscopic study, Langmuir 16 (2000)9439.

17] W.P. Lee, V.R. Gundabala, B.S. Akpa, M.L. Johns, C. Jeynes, A.F. Routh, Distributionof surfactants in latex films: a Rutherford backscattering study, Langmuir 22(2006) 5314–5320.

18] G.H. Xu, J. Dong, J. Zhang, S.J. Severtson, C.J. Houtman, L.E. Gwin, Character-izing the distribution of nonylphenol ethoxylate surfactants in water-basedpressure-sensitive adhesive films using atomic-force and confocal Ramanmicroscopy, J. Phys. Chem. B 112 (2008) 11907–11914.

19] E. Kientz, Y. Holl, Distribution of surfactants in latex films, Colloid Surf. A 78(1993) 255–270.

20] F. Belaroui, M.P. Hirn, Y. Grohens, P. Marie, Y. Holl, Distribution of water-solubleand surface-active low-molecular-weight species in acrylic latex films, J. Col-loid Interface Sci. 261 (2003) 336–348.

21] C. Arnold, Formation of thin polymer films from composite nanocolloids, PhDThesis, University of Strasbourg, France, 2009.

22] V. Mellon, Synthesis and characterization of waterborne polymer/laponitenanocomposite latexes through miniemulsion polymerization, UniversitéClaude Bernard – Lyon 1, France, 2009.

23] S. Lam, A.C. Hellgren, M. Sjöberg, K. Holmberg, H.A.S. Schoonbrood, M.J. Unzué,J.M. Asua, K. Tauer, D.C. Sherrington, A. Montoya Goni, Surfactants in het-erophase polymerization: a study of film formation using AFM, J. Appl. Polym.Sci. 66 (1997) 187–198.

24] H.M. Vale, T.F. McKenna, Adsorption of sodium dodecyl sulfate and sodiumdodecyl benzenesulfonate on poly(vinyl chloride) latexes, Colloid Surf. A 268

(2005) 68–72.

25] S.Y. Lin, C. Dong, T.J. Hsu, C.T. Hsu, Determination of adsorption of an ionicsurfactant on latex from surface tension measurements, Colloid Surf. A 196(2002) 189–198.

26] B.R. Vijayendran, Polymer polarity and surfactant adsorption, J. Appl. Polym.Sci. 23 (1979) 733–742.

[

[

cochem. Eng. Aspects 374 (2011) 58–68

27] R.D. Deegan, O. Bakijin, T.F. Dupont, G. Huber, S.R. Nagel, T.A. Witten, Capillaryflow as the cause of ring stains from dried liquid drops, Nature 389 (1997)827–829.

28] R.D. Deegan, O. Bakajin, T.F. Dupont, G. Huber, S.R. Nagel, T.A. Witten, Contactline deposits in an evaporating drop, Phys. Rev. E 62 (2000) 756–765.

29] D. Juhué, Y. Wang, J. Lang, O.M. Leung, M.C. Goh, M.A. Winnik, Surfactant exu-dation in the presence of a coalescing aid in latex films studied by atomic forcemicroscopy, J. Polym. Sci. Part B. Polym. Phys. 33 (1995) 1123–1133.

30] C.L. Zhao, F. Dobler, T. Pith, Y. Holl, M. Lambla, Surface composition of coalescedacrylic latex films studied by XPS and SIMS, J. Colloid Interface Sci. 128 (1989)437–449.

31] C. Tanford, The Hydrophobic Effect: Formation of Micelles and Biological Mem-branes, John Wiley & Sons, 1973.

32] E.B. Bradford, J.W. Vanderhoff, Morphological changes in latex films, J. Macro-mol. Chem. 1 (1966) 335–360.

33] J. Faucheu, L. Chazeau, C. Gauthier, J.Y. Cavaillé, M. Goikoetxea, R. Minari,J.M. Asua, Latex imaging by environmental STEM: application to the study ofthe surfactant outcome in hybrid alkyd/acrylate systems, Langmuir 25 (2009)10251–10258.

34] E. Aramendia, J. Mallegol, C. Jeynes, M.J. Barandiaran, J.L. Keddie, J.M. Asua,Distribution of surfactants near acrylic latex film surfaces: a comparisonof conventional and reactive surfactants (Surfmers), Langmuir 19 (2003)3212–3221.

35] H.J. Butt, R. Kuropka, B. Christensen, Colloid Polym. Sci. 272 (1994) 1218–1223.36] J. Mallégol, J.P. Gorce, O. Dupont, C. Jeynes, P.J. McDonald, J.L. Keddie, Origins and

effects of a surfactant excess near the surface of waterborne acrylic pressuresensitive adhesives, Langmuir 18 (2002) 4478–4487.

37] J.S. Bernardes, C.A. Rezende, F. Galembeck, Morphology and self-arraying of SDSand DTAB dried on mica surface, Langmuir 26 (2010) 7824–7832.

38] W.R. Dreher, M.W. Urban, R.S. Porzio, C.L. Zhao, Surface interactions betweenpropylene glycol and sodium dodecyl sulfate during coalescence of methylmethacrylate (MMA)/n-butyl acrylate (nBA)/acrylic acid (AA) colloidal disper-sions; a spectroscopic study, Langmuir 19 (2003) 10254–10259.

39] D. Guigner, C. Fischer, Y. Holl, Film formation from concentrated reactive sili-cone emulsions. 2. Surfactant distribution, Langmuir 17 (2001) 6419–6425.