Tribology International 57 (2013) 257–265

Contents lists available at SciVerse ScienceDirect

Tribology International

0301-67

http://d

n Corr

E-m

journal homepage: www.elsevier.com/locate/triboint

Friction properties of acrylic-carboxylated latex films. 2:Effect of post added surfactant

Guillaume Klein, Vincent Le Houerou, Christian Gauthier, Yves Holl n

Institut Charles Sadron, Universite de Strasbourg, CNRS UPR 22, 23 rue du Loess, B.P. 84047, 67034 Strasbourg Cedex, France

a r t i c l e i n f o

Article history:

Received 11 April 2012

Received in revised form

2 July 2012

Accepted 4 July 2012Available online 20 July 2012

Keywords:

Latex film

Acrylic copolymers

Friction

Sodium dodecyl sulfate

9X/$ - see front matter & 2012 Elsevier Ltd. A

x.doi.org/10.1016/j.triboint.2012.07.011

esponding author. Tel.: þ333 88 41 41 16.

ail address: yves.holl@unistra.fr (Y. Holl).

a b s t r a c t

The second part of this work aiming at investigating how specificities of thin films prepared from

aqueous polymer colloids (latexes) influence their friction properties is devoted to the role of

surfactants. Two acrylic latexes containing either 1 wt% or 4 wt% of acrylic acid were used, to which

0 to 9 wt% of sodium dodecyl sulfate (SDS) was post added before film formation by drying. Bulk

mechanical properties were also studied in order to improve interpretation of friction results.

Increasing the amount of SDS in the films has the effect of increasing Young’s and shear storage

moduli and decreasing tan d. SDS, forming crystallized aggregates in the bulk of the films, behaves like a

reinforcing filler, except at 9 wt% where the SDS phase percolates under stress and allows film

deformation by shear bands at very low stress. Friction coefficients dramatically decrease with

increasing SDS concentration, especially at high strain rate. Surface shear stress is strongly decreased

due to lubrication by SDS having migrated to the film surface. In order to gain more insight in the

lubrication mechanisms, SDS layers were deposited on glass in increasing amounts. This part of the

study seemed to indicate that organization of the SDS surface layers has more impact than their

amount.

& 2012 Elsevier Ltd. All rights reserved.

1. Introduction

In a companion paper to this one [1], we started to investigatehow specificities of latex films influence their tribological proper-ties. Latex films are prepared from polymer aqueous colloids in athree step process: drying, particle deformation and particlecoalescence [2,3]. They are used in many applications like paints,adhesives, papers, cosmetics, biomaterials. Often, they are com-pared to ‘‘solution films’’ which have similar applications butwhere the polymer is dissolved in an organic solvent and notdispersed in water. The obvious advantage of latex films is thatthey are almost Volatile Organic Compounds (VOCs) free. How-ever, their properties can be lower than those of solution films.That is the reason why they were, and still are, so extensivelystudied in industry and academia.

Latex films have specificities, mainly due to their particularstructure and the presence of surfactants [1]. Surfactants partici-pate to the mechanism of latex synthesis by polymerization in adispersed state (in practice, mainly emulsion polymerization) [4],are necessary to stabilize the latex and help latex casting onvarious substrates. They are usually present in the concentrationrange 0.5–3 wt%, even if higher values can be found, for instance

ll rights reserved.

in pressure sensitive adhesives (PSAs) [5] (some practitioners inthis field ironically state that waterborne PSAs are ‘‘surfactantstuff with a few other components inside’’). Even when it is aminor constituent, the surfactant may heavily affect film forma-tion mechanisms [6–8] and properties like gloss and appearance[9], water uptake [10,11], permeability [12], glass transition [13],peel resistance [14,15], tensile strength [16]. It is important tonote that, being largely incompatible with the polymeric matrixand surface-active, surfactants are not homogeneously distribu-ted in latex films but, rather, tend to concentrate at the surfaceand the interface with the substrate and to form aggregates in thebulk of the film. A large number of papers can be found in theliterature dealing with distribution of surfactants in latex films,starting as early as in 1936 [17] (quoted by Routh [18]). Introduc-tions of several papers [17,19] provide good and quickly readoverviews on the topic. In the 1980s and 1990s, most of thepublished work was devoted to the investigation of the surfactantat the interfaces. Enrichment was reported in a vast majority ofcases [20,21] although depletion, more rarely mentioned, is alsopossible [22]. All studies proved that surfactant distribution isinfluenced by several parameters: nature of the system (polymer,substrate, and surfactant), time, total concentration and filmformation conditions. Later on, more complete concentrationprofiles could be established using Raman Confocal Spectroscopy,showing micron sized surfactant aggregates inside of the films[23]. Modelling was attempted in the recent years [17,24,25], only

Fig. 1. One millimeter thick films of BuA/MMA 4 containing increasing amounts

of SDS.

G. Klein et al. / Tribology International 57 (2013) 257–265258

partially supported by experiments because of the complexity ofthe mechanisms behind surfactant distributions. Next step in thisfield probably will be computer simulation but correspondingpapers are still to appear.

After investigating the effects of acrylic acid concentration andpH on tribological behavior of acrylic latex films in our previouspublication [1], we now report results on the influence ofsurfactants on friction properties of similar films. The same butylacrylate/methyl methacrylate random copolymer was used as themain polymeric constituent, also containing 1 or 4 wt% of acrylicacid. The pH was fixed at an intermediate value of 4.5, except inone case where it was raised to 10 in order to improve filmquality to establish friction master curves. Surfactant free latexeswere used as references to which 1.5, 3, 6 or 9 wt% of sodiumdodecyl sulfate was post-added.

2. Experimental

Most details of the experimental section of this paper arecommon with the previous one [1]. They will not all be repeatedhere. Only the main features of the experiments which were carriedon will be recalled, together with characteristics specific to this part.Obvious details on classical and marginally used techniques (AFMand contact angle measurements) will be omitted.

2.1. Latexes and films

Two core–shell latexes were synthesized, differing by the con-centration of acrylic acid in the polymer: 1 wt% or 4 wt%. The corewas a butyl acrylate -co- methyl methacrylate random copolymer(60/40 weight ratio). The two latexes will be designated as BuA/MMA 1 or BuA/MMA 4 in the following text and figures. Aftersynthesis, the latexes were purified by dialysis using a Milliporemembrane until the conductivity of water in contact with the latexwas less than 3 mS/ cm. The pH was subsequently adjusted to 4.5 (to10 in one case) by dropwise addition of 1 M NaOH solutions inwater. Sodium dodecyl sulfate (SDS) (A.C.S. Reagent, Sigma-Aldrich,purity 499%) was post-added in various amounts (1.5 to 9 wt%based on the polymer weight) under stirring. An equilibration timeof 5 h always preceded film casting. Table 1 resumes the maincharacteristics of the latices used in this study.

For friction tests, latexes were cast on glass substrates(75�25 mm) previously cleaned in acetone and water, and driedfor 10 days under controlled conditions (22 1C and 50% RH). Thedry film thickness, measured by optical profilometry, was 90 mm.Thicker films (1 mm) were prepared for large strain tensile anddynamic mechanical measurements. Latexes were filled into aTeflon mold and dried in the same conditions as thin supportedfilms. Samples of adequate shapes were cut out of large films withspecific punches. Fig. 1 shows thick films of the BuA/MMA 4 systemcontaining no or progressively increasing concentrations of SDS.Totally flaw free thick films are difficult to prepare because ofstresses appearing during drying [26]. When the SDS concentrationincreases, films become hazy and almost opaque at 9 wt%. This isdue to surfactant aggregates in the bulk.

Table 1Main characteristics of the latexes used in this study.

Latex Composition Mean particle

diameter

(nm7 1nm)

Size

dispersity

index

(%71%)

Tg

(1C)

(DSC)BuA

(wt%)

MMA

(wt%)

AA

(wt%)

BuA/MMA 1 59.5 39.5 1 102 4.2 4BuA/MMA 4 58 38 4 111 10.4 9

2.2. Bulk mechanical properties

True stress/true strain curves were established in uniaxialtensile tests on bone-shaped thick films, using a device equippedwith a camera allowing the determination of the stressed sampledimensions at any time. Tensile rate was set at 0.1 mm s�1 andtemperature at 25 1C71 1C.

Dynamic mechanical and thermal analysis (DMTA) was per-formed with an ARES rheometer (Advanced Rheometer ExpansionSystem) using rectangular (30�8 mm) thick films in torsionmode, at 1% strain. Curves were recorded in a frequency range0.1–100 rad s�1 at 11 different temperatures ranging from�20 1C to 30 1C. Master curves could be built using the time–temperature equivalence.

2.3. Friction measurements

Friction measurements were performed on thin films sup-ported on glass in a home-made tribometer described in details inreferences [1] and [27]. Measurements were performed at sevendifferent temperatures, namely: �40 1C, �20 1C, 0 1C, 5 1C, 10 1C,20 1C, 40 1C. In this apparatus, the sample was moving whereasthe tip was fixed. A built-in microscope allowed an in-situobservation of the tip/ film contact area and the groove left onthe surface through the transparent sample. The tip was apolished stainless steel bead of radius 12.5 mm. The bead waswashed in water and ethanol and dried in nitrogen before eachmeasurement. The normal force was set at 0.5 N. In a standardexperiment, the sample started moving at the lowest velocity andaccelerated stepwise up to the highest velocity. Our speed rangewas 10�3 to 10 mm/s. At each speed step, it moved over adistance of at least 1 mm in order to stabilize the contact area.Strain rates are obtained by dividing the speed by the width of thecontact area (see [1] for a discussion of the choice of a relevantcharacteristic length). At certain velocities and temperatures,sliding can become unstable, giving rise to Schallamach wavesand stick slip [28]. The consequence is a loss of precision ontangential force and contact area measurements. Almost opaquefilms also much decrease the precision of the contact areameasurements. This explains that friction results may look noisyand show relatively large error bars (see [1] for more detailsabout reproducibility tests).

2.4. Ellipsometry

Measurements of SDS layer thicknesses was carried out with aPLASMOS SD 2100 instrument operating at the single wavelengthof 632.8 nm and a constant angle of 45 1C. The refractive index ofSDS layers was assumed to be constant at n¼1.460 [29] whateverthe number of layers. For each studied concentration, several

G. Klein et al. / Tribology International 57 (2013) 257–265 259

points were measured to obtain the average value for the filmthickness and to determine the film homogeneity.

Fig. 3. Young’s moduli (small strain) versus SDS concentration (wt%) for BuA/

MMA 1 and BuA/MMA 4 latex films, determined from the data in Fig. 2.

0 1.5 3 6 9SDS wt.%

substratesubstrate substrate substrate substrate

Fig. 4. Scheme of SDS (in blue) distribution in latex films with increasing

concentration.

3. Results and discussion

3.1. Bulk mechanical properties of latex films with increasing SDS

concentration

Fig. 2 shows the large strain tensile behavior of BuA/MMA1 and BuA/MMA 4 films containing various amounts of SDS.Increasing the surfactant concentration from 0 to 6 wt% leads toan increase in tensile strength for the two acrylic acid contents (1and 4 wt%). At 9 wt% of SDS, the stress initially rises steeply andthen decreases to almost zero upon film elongation. In all curves,the last points correspond to full stroke in the instrument and notto film rupture.

Young’s moduli increase with SDS concentration in bothsystems (Fig. 3).

The role of surfactants on mechanical properties of polymer filmshas to be discussed in relation to surfactant distribution. Certainkinds of surfactants have a limited solubility in some polymers andcan therefore act as plasticizers. A known example is ethylene oxidecontaining non ionic surfactants in acrylic matrices [30,31]. How-ever, most of them are incompatible with polymers as it is the casefor ionic surfactants in hydrophobic polymers. For SDS, it wasrecognized a long time ago [12] that it phase separates from usualhydrophobic polymers to form ionic clusters with a postulatedstructure looking like a reverse micelle, playing the role of acrystalline filler, increasing Tg and Young’s modulus. This wasconfirmed by Kientz et al. [32] a couple of years later.

In a typical dry latex film, a surfactant like SDS is found at theinterfaces (film–air and film–substrate), in the form of aggregates

Fig. 2. True stress versus true strain for different concentrations (wt%) of SDS in

BuA/MMA 1 (a) and BuA/MMA 4 (b) latex films at a tensile speed of 0.1 mm/ s and

a temperature of 25 1C. Latex pH: 4.5. Last points do not correspond to film

rupture.

in the bulk of the film and also trapped at the interfaces betweenthe particles which formed the film [33]. These trapped moleculesresult from the incomplete desorption of the surfactant whenparticles get in contact upon drying [34]. When the total surfac-tant concentration increases, all these locations develop: surfac-tant layers at the surface increase in thickness, more and largeraggregates are found inside the film and the surfactant located atthe particle interfaces tends toward a continuous phase. For thereader not familiar with latex films, a scheme (Fig. 4), hopefully,makes the picture clearer.

From zero to 6 wt%, discrete SDS aggregates dominate thelarge strain mechanical behavior of films (Fig. 2). These aggre-gates are crystalline at room temperature. They reinforce thefilms like would do ordinary fillers. When the SDS concentrationreaches 9 wt%, aggregates dominate the mechanical behavior atrelatively small strain (10 to 20%), but, upon further elongation,the SDS phase becomes almost continuous, the surfactant trappedat the particle interfaces playing the major role. This layeredsurfactant phase easily deforms by shearing. Consequently, thefilm can be elongated quite a lot at very low stress.

Small strain dynamic mechanical analysis confirms this inter-pretation. Storage shear moduli for the BuA/MMA 4 system arepresented in Fig. 5. Moduli at the rubber plateau monotonicallyincrease with SDS concentration, as expected. The strain beingsmall (1%), the film with 9 wt% SDS does not behave in a peculiarway. Similar results (not shown) were observed for BuA/MMA 1.

Loss factors versus frequency are shown in Fig. 6 over a largefrequency range. As SDS concentration increases, the film as awhole becomes more rigid, progressively preventing chainmotion and therefore decreasing viscoelastic dissipation. Fromzero to 3 wt% SDS, the effect is weak whereas it becomes moremarked for 6 and 9 wt%. There is a slight shift of peaks when BuA/MMA 1 (Fig. 6a) and BuA/MMA 4 (Fig. 6b) are compared. This isdue to the polymeric matrix becoming more rigid when theacrylic acid content increases from 1 to 4 wt%, as was discussedin the previous paper [1].

3.2. Friction properties

A first series of friction results are presented in Fig. 7 to Fig. 9.Fig. 7 shows friction coefficients versus strain rate for BuA/MMA

Fig. 5. Master curves of storage shear moduli versus time for different concentra-

tions of SDS (wt%) in BuA/MMA 4 for 1% strain and a reference temperature of

24 1C. Latex pH: 4.5.

Fig. 6. Master curves of loss factors versus frequency for different concentrations

of SDS (wt%) in BuA/MMA 1 (a) and BuA/MMA 4 (b). Strain: 1%. Reference

temperature: 24 1C. Latex pH: 4.5.

Fig. 7. Friction coefficient versus strain rate for different concentrations of SDS

( wt%) in BuA/MMA 1 (a) and BuA/MMA 4 (b) latex films. Temperature: 24 1C.

Normal force: 0.5 N. Latex pH: 4.5.

G. Klein et al. / Tribology International 57 (2013) 257–265260

1 and BuA/MMA 4 containing increasing concentrations of SDS. Atlow strain rate, the effect of SDS is minimal whereas at high rate itbecomes massive. For instance, at 10 s�1, the friction coefficienton BuA/MMA 4 (Fig. 7b) is divided by a factor around 30 whengoing from 0 to 9 wt% of SDS. The following curves mostobviously will demonstrate that this is due to a decrease in shearresistance and not to a pressure effect.

Pictures of the stainless steel tip – BuA/MMA 4 film contactarea at increasing SDS concentration (Fig. 8) show the evolutionfrom a plastic to an almost purely elastic contact. At 0 and 1.5 wt%of SDS, the film surface is strongly damaged whereas at 3 wt% andabove only a progressively vanishing groove is visible behind thetip. The bead slides more and more easily on the film surface. Inorder to check for a pressure effect, contact pressures versusstrain rate are drawn in Fig. 9a. The usual increase of pressure

with strain rate is observed, but the SDS concentration effect isweak, essentially negligible at high rate (above 10�1 s�1), whereit is so marked for the friction coefficient. If now, the shear stressis plotted as a function of the contact pressure (Fig. 9b), at apressure of 2 MPa (corresponding to a strain rate of 10 s�1 inFig. 9a), the shear stress is again divided by a factor close to 30when the SDS concentration is increased from 0 to 9 wt%. It isstraightforward to explain these results by a lubrication effect ofthe SDS exuded to the film surface upon drying, in increasingamount when the nominal surfactant concentration is increasedin the system (see Fig. 4).

In order to confirm this interpretation, a BuA/MMA 1 film with6 wt% of SDS was rinsed by a quick splash of water from a pipette.Fig. 10 shows that the friction coefficients on the rinsed film tendtoward the ones on the surfactant free film but remain below,probably because rinsing was incomplete.

In the results presented so far, only the speed range availablein our instrument was used, at room temperature. In order to gaina broader view over a wider strain rate scale, measurements wereperformed at different temperatures and master curves could bebuilt for the BuA/MMA 1 system without and with 6 wt% of SDS(Fig. 11 and Fig. 12). For this part of the study, we wanted toensure good film quality. Therefore, the pH of the latex was raisedto 10 before film formation in order to increase its stability. Forthe surfactant free film, the friction coefficient (Fig. 11) goesthrough a maximum corresponding to a peak of dissipation, aswas discussed in the preceding paper [1]. The film containing6 wt% of SDS has much lower friction coefficients especially athigh strain rate. The decrease factor in this case is around 50.Fig. 12 clearly confirms what was stated above: contact pressurecurves almost superpose whereas shear stress is much decreased

Fig. 8. Pictures of the stainless steel tip – BuA/MMA 4 film contact area at increasing SDS concentration corresponding to the circles in Fig. 7b. The tip moves to the right.

Strain rate: 10�1 s�1. Temperature: 24 1C. Normal force: 0.5 N. Latex pH: 4.5. The scale bar is 1 mm.

G. Klein et al. / Tribology International 57 (2013) 257–265 261

for the film containing SDS. The decreased friction coefficientundoubtedly can be attributed to a shear effect due to a surfactantlayer present at the film surface.

At this point, it seemed useful to try to get more informationabout the surfactant layers at the surface of the films. For thatpurpose, contact angles of water were measured and some AFMimages were taken. Fig. 13 shows contact angles of water versustime on the BuA/MMA 4 films with SDS. The contact anglesprogressively decrease when the concentration of SDS increases,as could be expected. A classical problem in a study where awater droplet is deposited on a surface on which a surfactant ispresent is that of the dissolution of the surfactant in the testliquid [35]. Therefore, we checked for a kinetic effect by measur-ing the contact angle versus time over a couple of minutes.Surprisingly, the contact angle remained constant over 3 min,except for the film containing 9 wt% of SDS. This result could beexplained in terms of structure of the surfactant layer sitting ontop of the polymeric film. It is known that the surfactantmigrating to the surface of a latex film may adopt variousstructures [18,36]. When ionic surfactants are concerned, thesestructures are often composed of bilayers exposing the hydro-phobic tail to the exterior and hiding the ionic head inside,arranged in various ways. Such bilayers adsorbed on a hydro-phobic surface on one side and exposed to air on the other sideare favored from a thermodynamic point of view. In the presentcase, between 1.5 and 6 wt% of SDS, the surfactant layer couldbe an incomplete bilayer, with holes showing the polymer

underneath (see Fig. 15 for an illustration). The holes woulddecrease in size with increasing SDS concentration. Dissolution ofSDS in the water droplet and contact angle equilibration would berapid because water would be in direct contact with the ionicheads at the edges of the holes. At a concentration of 9 wt%, thebilayer would be complete. The initial contact angle would thenbe higher because water would sit on a continuous surface ofhydrophobic tails, and the dissolution process would beslowed down.

This interpretation is supported by AFM images (Fig. 14 andFig. 15). Fig. 14 represents a topographic image of the BuA/MMA4 film with 9 wt% of SDS. It shows a certain roughness but it looksessentially continuous. Imaging soft polymer films by AFM is nota straightforward task [37], especially when a mobile surfactantcomplicates the system [38]. We could not get high enoughquality images with BuA/MMA 4 films with amounts of surfa-ctant lower than 9 wt%. It was in a slightly different case (sameconstituents but lightly different monomer ratio: 50/50 instead of60/40 and no acrylic acid) that we could observe the incompleteSDS bilayer mentioned above (Fig. 15). This image is reproducedhere as an illustration because we think that the same structureprobably exists on the system discussed in this paper.

3.3. SDS layer on glass

The next step in this work was to try to better understand themechanism of the lubricating effect of a SDS layer sitting on a

Fig. 9. Contact pressure versus strain rate (a) and shear stress versus contact

pressure (b) for different concentrations of SDS in BuA/MMA 4 latex films.

Temperature: 24 1C. Normal force: 0.5 N. Latex pH: 4.5.

Fig. 10. Friction coefficient versus strain rate for BuA/MMA 1 latex films without

SDS (squares), with 6 wt% SDS (circles) and with 6 wt% SDS after rinsing with

water (triangles). Temperature: 24 1C. Latex pH: 4.5.

Fig. 11. Local friction coefficient versus strain rate for a BuA/MMA 1 latex film

without SDS (squares) and with 6 wt% SDS (circles). Reference temperature: 24 1C.

Latex pH: 10.

Fig. 12. Contact pressure (a) and shear stress (b) versus strain rate (master curves)

for a BuA/MMA 1 latex film without SDS (squares) and with 6 wt% SDS (circles).

Reference temperature: 24 1C. Latex pH: 10.

G. Klein et al. / Tribology International 57 (2013) 257–265262

surface. For that purpose, friction measurements were performedon a model system composed of pure SDS on top of a glasssubstrate.

It has been recognized that surfactants decrease friction for along time [39]. The main questions that have been raised in theliterature concern the effects of the nature of the system (natureof the surfaces in contact and of the surfactant) [40–43], of thelayer thickness [44], of the durability of the lubrication by asurfactant [45,46]. Sahoo and Biswas [39] have shown thatlinoelic acid (unsaturated) has a stronger lubricating effect thanstearic acid (saturated) on steel because of a chemical reaction ofthe unsaturation with iron oxide, assisted by the favorableorientation of the molecule due to the contact pressure andleading to a very low friction coefficient (0.04). On the other

hand, Graca et al. [40] presented results on a series of non ionicsurfactants with different alkyl chain lengths (Tween), suggestingthat while Tween molecules adsorb onto hydrophobic surfaces toform a robust separating layer, the lubricating properties of theselayers are dominated by a highly dissipative slip plane, the samefor all chain lengths. Novotny et al. [41] have studied fatty acidsalts on silicon substrates and ceramic sliders. They demonstrateda surprising absence of thickness effect. In fact, one monolayerwas efficient for friction coefficient reduction and, in case of thepresence of multilayers, all layers except one were swept awayfrom the contact area by the sliding process. In general, durabilitydepends on the system and may vary from almost zero to a fewhundreds or thousands of cycles [42] or even reach more than 104

cycles [43].In our case, SDS was deposited by spin coating on glass

(or silicon wafer for ellipsometry measurements and AFM

Fig. 13. Contact angle of water on BuA/MMA 4 latex films with various concen-

trations of SDS versus time. Time zero corresponds to the first possible measure,

3 to 5 s after depositing the water droplet. Droplet volume: 3 ml. Temperature:

24 1C. Latex pH: 4.5.

Fig. 14. AFM topographic image of a BuA/MMA 4 film with 9 wt% of SDS.

Fig. 15. Phase image of the surface of a BuA/MMA copolymer, slightly different

from the systems presented in this paper (monomer ratio: 50/50 instead of 60/40,

no acrylic acid)þ6 wt% SDS. Polymer particles are visible in the holes. Reproduced

with permission from Ref. [33].

Table 2Thicknesses, measured by ellipsometry, of SDS layers spin coated on silicon wafer

as a function of the initial solution concentration.

Concentration (M) Thickness (nm) Standard deviation (nm)

0 (uncoated glass)

10–4 2.0 0.7

10–3 3.6 1.0

10–2 4.7 1.9

10–1 4.9 1.1

G. Klein et al. / Tribology International 57 (2013) 257–265 263

observation) substrates previously treated in piranha solution(mixture of sulfuric acid and hydrogen peroxide in a 3 to 1 ratio).Different concentrations of SDS solutions ensured different

amounts of surfactant left on the substrates. More details on theexperimental procedures can be found in reference [47]. Table 2shows the average thickness of the surfactant layer as a functionof the concentration (from 10�4 to 10�1 M) of the spin coatedSDS solution. The value at 10�4 M, 2 nm, corresponds well to thetheoretical length of a SDS molecule [48]. Piranha treatmentleading to protonated glass or silicon surfaces, it is likely that acontinuous SDS monolayer is deposited under these conditions.

Furthermore, an AFM topographic image (not shown) onlyrevealed a very flat, uniform surface. As concentration isincreased, the average thickness progressively increases and, onAFM images, heterogeneities appear on top of a flat surface. Theheight of these structures, 3.7 nm, corresponds to the thickness ofa SDS bilayer (Fig. 16).

These fragments of bilayers most probably sit on a monolayer,this is the favorable structure from a thermodynamic point ofview, and, as concentration further increases, they merge in adisordered way. At 10�1 M, a very rough surface is observed (notshown), as also reflected by the high standard deviation of thethickness in Table 2. It is very difficult to get homogeneous layersof SDS with monotonically increasing thicknesses, whatever thesubstrate. Bernardes et al. [49] have tried on cleaved mica andalso obtained rough structures at higher SDS concentration. Apossible way to obtain increasing even numbers of layers mightbe to deposit Newton black films (drained bilayers) in sequence,as was performed by Andreatta at al. [50] with cationic and nonionic surfactants.

SDS was then deposited on glass under the same spin coatingconditions than on silicon. It was postulated that very similarstructures were obtained on both substrates. The tribologicalbehavior of these layers was tested against a glass bead of radius9.82 mm, at a normal force of 0.5 N and a sliding speed of0.1 mm s–1. Results are shown in Fig. 17. As compared to theuncoated glass surface, the friction coefficient is reduced by afactor of 6 when SDS is present. Surprisingly, the friction coeffi-cient does not depend much on the layer thickness. It seems thatwe are in the same situation as Novotny et al. [41] where thefriction behavior was dominated by a surfactant monolayer,whatever the total layer thickness.

It is now possible to compare friction results on the latex filmsand on glass. It was shown that friction on latex films is highlydependent on the nominal SDS concentration and therefore on thequantity of surfactant at the surface. On the other hand, on glass,friction is only slightly dependant on the amount (thickness) of theSDS layer. It is not sure that robust conclusions can be drawn fromthis comparison because substrates are so different (one soft andcontaining surfactant in the bulk, the other rigid with surfactantonly on top), but it suggests that the key characteristic might be thestructure of the surfactant layer rather its thickness. In the case ofthe films, the steel bead is in contact with surfactant and polymer,and the steel–polymer contact progressively decreases with increas-ing SDS concentration. In the case of the glass substrate, the bead isin contact at least with a continuous monolayer and the supple-mentary surfactant only plays a minor role. In our opinion, the

Fig. 16. AFM topographic image of SDS on a silicon wafer surface after spin coating of a 7�10–3 M solution.

Fig. 17. Apparent friction coefficient versus SDS concentration in spin coated

solution on glass. Sliding speed: 10–1 mm s–1. Temperature: 24 1C. Normal force:

0.5 N. The red line corresponds to non coated glass.

G. Klein et al. / Tribology International 57 (2013) 257–265264

problem of the influence of the surfactant layer structure on frictionproperties is interesting and deserves further research work.

4. Conclusion

At the end of this study, the following highlights may beemphasized:

At low deformation and at large deformation in the concentra-tion range 1.5 to 6 wt%, SDS acts like a reinforcing filler, increasingYoung’s and shear moduli at the rubbery plateau, decreasingdissipation peaks and increasing stress in the strain range 0 to120%. This is due to a phase separation of the surfactant inmicrosized crystalline aggregates during film formation. At a highSDS concentration of 9 wt%, the surfactant phase in the filmbecomes continuous under stress, allowing very large deforma-tion at almost zero stress.

As far as friction properties are concerned, SDS largely decreasesfriction coefficients at high strain rate, from around 4 to less than0.2. It was clearly demonstrated that pressure effects are negligibleand that the decrease in friction can be attributed to a decrease inshear resistance. SDS segregated at the film surface definitely acts asa lubricant.

Although the relevance of the comparison with polymericfilms is quite uncertain, friction results on pure SDS on glasssuggested that the lubricating effect of a surfactant at a surface isnot just a matter of amount or thickness but that the structure, i.e.the way the surfactant is arranged (leaving or not holes at thesurface, for instance) might play a major role.

Several questions on the role of surfactants on friction proper-ties of thin polymeric films remain open. One is to better separatecontributions of the surfactant in the bulk and at the surface, indirect contact with the moving body. It could be a way to progressin the general problem in tribology of the respective importanceof dissipative processes specific to the bulk and to the surface.Another is to investigate how the structure of the surfactant at thesurface, in relation to its amount, influences shear resistance. Allthese questions are linked to the general and vast problem of thedistribution of surfactants in latex films, itself dependent ondrying mechanisms.

Acknowledgements

The financial support received from the European Unionproject (Napoleon NMP3-CT-2005–011844) is gratefully appre-ciated. One of us (GK) gratefully acknowledges the grant from theRegional Council of Alsace, France to support his Ph.D. work.

References

[1] Klein G, Le Houerou V, Gauthier C, Holl Y. Friction properties of acrylic-carboxylated latex films. 1. Effects of acrylic acid concentration and pH.Tribology International 2012;53:142–9.

[2] Keddie JL, Routh AF. Fundamentals of Latex Film Formation, Processes andProperties. Springer; 2010.

[3] Keddie JL. Film formation of latex. Materials Science and Engineering1997;21:101–70.

[4] Lovell PA, El-Aasser MS. Emulsion Polymerization and Emulsion Polymers.New York: J Wiley & Sons; 1997.

[5] Satas D. Handbook of Pressure Sensitive Adhesives. New York: Van NostrandReinhold; 1989 p 437.

[6] Lam S, Hellgren AC, Sjoberg M, Holmberg K, Schoonbrood HAS, Unzue MJ,Asua JM, Tauer K, Sherrington DC, MontoyaGoni A. Surfactants in heterophasepolymerization: A study of film formation using force microscopy. Journal ofApplied Polymer Science 1997;66:187–98.

[7] Vandezande GA, Rudin A. Film formation of vinyl acrylic latexes; effects ofsurfactant type, water and latex particle size. Journal of Coatings Technology1996;68:63–73.

[8] Juhue D, Lang J. Effect of surfactant postadded to latex dispersion on filmformation: a study by atomic force microscopy. Langmuir: the ACS Journal ofSurfaces and Colloids 1992;9:792–6.

[9] Asua JM, Schoonbrood HAS. Reactive surfactants in heterophase polymeriza-tion. Acta Polymerica 1998;49:671–86.

[10] Butler LN, Fellows CM, Gilbert RG. Effect of surfactant systems on the watersensitivity of latex films. Journal Of Applied Polymer Science 2004;92:1813–23.

[11] Yang YK, Li H, Wang F. Studies on the water resistance of acrylic emulsionPSAs. Journal of Adhesion Science and Technology 2003;17:1741–50.

[12] Steward PA, Hearn J, Wilkinson MC. Studies on permeation through polymerlatex films. I. Films containing no or only low levels of additives. PolymerInternational 1995;38:1–12.

G. Klein et al. / Tribology International 57 (2013) 257–265 265

[13] Tanaka T, Fujimoto T, Shibayama K. Glass transition behavior of emulsion-polymerized polymer films containing a surface-active agent. Journal ofApplied Polymer Science 1979;23:1131–6.

[14] Charmeau JY, Gerin P, Vovelle L, Schirrer R, Holl Y. Adhesion of Latex Films.III. Surfactant Effects at Various Peel Rates. Journal of Adhesion Science andTechnology 1999;13:203–15.

[15] Xu GH, Dong J, Zhang J, Severtson SJ, Houtman CJ, Gwin LE. Characterizing thedistribution of nonylphenol ethoxylate surfactants in water-based pressure-sensitive adhesive films using atomic-force and confocal raman microscopy.The Journal of Physical Chemistry B 2008;112:11907–14.

[16] Amalvy JI. Colloidal and film properties of carboxylated acrylic latices, effectof surfactant concentration. Pigment and Resin Technology 1998;27:20–7.

[17] Wagner H, Fischer G. Kolloid A 1936 . [17].;77:12.[18] Gundabala VR, Zimmerman WB, Routh AF. A model for surfactant distribu-

tion in latex coatings. Langmuir: the ACS Journal of Surfaces and Colloids2004;20:8721–7.

[19] Tzitzinou A, Jenneson PM, Clough AS, Keddie JL, Lu JR, Zhdan P, Treacher KE,Satguru R. Surfactant concentration and morphology at the surface of acryliclatex films. Progress in Organic Coatings 1999;35:89–99.

[20] Zhao CL, Holl Y, Pith T, Lambla M. FTIR-ATR spectroscopic determination ofthe distribution of surfactants in latex films. Colloid and Polymer Science1987;265:823–9.

[21] Evanson KW, Urban MW. Surface and interfacial FTIR spectroscopic studies oflatexes. 3. The effect of substrate surface tension and elongation on exuda-tion of Surfactant. Journal of Applied Polymer Science 1991;42:2309–20.

[22] Kientz E, Holl Y. Distribution of surfactants in latex films. Colloid Surface A1993;78:255–70.

[23] Belaroui F, Grohens Y, Boyer H, Holl Y. Depth profiling of small molecules indry latex films by confocal Raman spectroscopy. Polymer 2000;41:7641–5.

[24] Gundabala VR, Lei CH, Ouzineb K, Dupont O, Keddie JL, Routh AF. Lateralsurface nonuniformities in drying latex films. AIChE Journal AmericanInstitute of Chemical Engineers 2008;54:3092–105.

[25] Arnold C, Thalmann F, Marques C, Marie P, Holl Y. Surfactant Distribution inWaterborne Acrylic Films. 1. Bulk Investigation. The Journal of PhysicalChemistry B 2010;114:9135–47.

[26] Petersen C, Heldmann C, Johannsmann D. Internal Stresses during FilmFormation of Polymer Latices. Langmuir: the ACS Journal of Surfaces andColloids 1999;15:7745–51.

[27] Gauthier C, Schirrer R. Time and temperature dependence of the scratchproperties of poly(methylmethacrylate) surfaces. Journal of Materials Science2000;35:2121.

[28] Baumberger T, Heslot F, Perrin B. Crossover from creep to inertial motion infriction dynamics. Nature 1994;367:544–6.

[29] Levchenko AA, Argo BP, Vidu R, Talroze RV, Stroeve P. Kinetics of sodiumdodecyl sulfate adsorption on and desorption from self-assembled mono-layers measured by surface plasmon resonance. Langmuir: the ACS Journal ofSurfaces and Colloids 2002;18:8464–71.

[30] Eckersley ST, Rudin A. The effect of plasticization and pH on film formation ofacrylic latexes. Journal of Applied Polymer Science 1993;48:1369–81.

[31] Odrobina E, Feng JR, Kawaguchi S, Winnik MA, Neag M, Meyer EF. Char-acterization of the miscibility of poly(butyl methacrylate) with a nonylphe-nolethoxylated surfactant and with various poly(ethylene glycol) oligomers.Macromolecules 1998;31:7239–47.

[32] Kientz E, Dobler F, Holl Y. Desorption of the surfactant from the particlesurface during latex film formation. Polymer International 1994;34:125–34.

[33] Belaroui F, Hirn MP, Grohens Y, Marie P, Holl Y. Distribution of water-solubleand surface-active low-molecular-weight species in acrylic latex films.Journal of Colloid and Interface Science 2003;261:336–48.

[34] Belaroui F, Cabane B, Dorget M, Grohens Y, Marie P, Holl Y. Small-angleneutron scattering study of particle coalescence and SDS desorption duringfilm formation from carboxylated acrylic latices. Journal of Colloid andInterface Science 2003;262:409–17.

[35] Cipriano BH, Raghavan SR, McGuiggan PM. Surface tension and contact anglemeasurements of a hexadecyl imidazolium surfactant adsorbed on a claysurface. Colloids and Surfaces A 2005;262:8–13.

[36] Arnold C, Klein G, Maaloum M, Ernstsson M, Larsson A, Marie P, Holl Y.Surfactant distribution in waterborne acrylic films 2. Surface investigation.Colloids and Surfaces A 2011;374:58–68.

[37] Mallegol J, Dupont O, Keddie JL. Obtaining and interpreting images of acrylicPSAs by tapping-mode AFM. Langmuir: the ACS Journal of Surfaces andColloids 2001;17:7022–31.

[38] Mallegol J, Gorce JP, Dupont O, Jeynes C, McDonald PJ, Keddie JL. Origins andeffects of a surfactant excess near the surface of waterborne acrylic pressuresensitive adhesives. Langmuir: the ACS Journal of Surfaces and Colloids2002;18:4478–87.

[39] Bowden FP, Tabor D. Friction and lubrication of solids. Oxford: ClarendonPress; 1950.

[40] Xiao X, Hu J, Charych DH, Salmeron M. Chain length dependence of thefrictional properties of alkylsilane molecules self-assembled on mica studiedby atomic force microscopy. Langmuir: the ACS Journal of Surfaces andColloids 1996;12:235.

[41] Boschkova K, Feiler A, Kronberg B, Stalgren JJR. Adsorption and frictionalproperties of gemini surfactants at solid surfaces. Langmuir: the ACS Journalof Surfaces and Colloids 2002;18:7930.

[42] Sahoo RR, Biswas SK. Frictional response of fatty acids on steel. Journal ofColloid and Interface Science 2009;333:707–18.

[43] Graca M, Bongaerts JHH, Stokes JR, Granick S. Friction and adsorption ofaqueous polyoxyethylene (Tween) surfactants at hydrophobic surfaces.Journal of Colloid and Interface Science 2007;315:662–70.

[44] Novotny V, Swalen JD, Rabe JP. Tribology of Langmuir-Blodgett layers.Langmuir: the ACS Journal of Surfaces and Colloids 1989;5:485–9.

[45] Agabekov V, Zhavnerko G. Mono- and multimolecular Langmuir-Blodgettfilms as protective coatings for friction surfaces. Russian Journal of GeneralChemistry 2007;77:343.

[46] Chugg KJ, Chaudhri MM. Boundary lubrication and shear properties of thinsolid films of dioctadecyl dimethyl ammonium chloride (TA 100). Journal ofPhysics D: Applied physics 1993;26:1993–2000.

[47] Klein G. PhD thesis. University of Strasbourg, France 2011.[48] Tanford C. The Hydrophobic Effect: Formation of Micelles and Biological

Membranes. New York: J Wiley& Sons; 1993.[49] Bernardes JS, Rezende CA, Galembeck F. Morphology and self-arraying of SDS

and DTAB dried on mica surface. Langmuir: the ACS Journal of Surfaces andColloids 2010;26:7824–32.

[50] Andreatta G, Benattar JJ, Petkova R, Wang JYJ, Tong P, Polidori A, Pucci B.Deposition of organized surfactant films on solid substrates. Colloids andSurface A 2008;321:211–7.