Nanoscale Aggregate Structures of Trisiloxane Surfactantsat the Solid-Liquid Interface

Jinping Dong† and Guangzhao Mao*

Department of Chemical Engineering and Materials Science, Wayne State University,5050 Anthony Wayne Drive, Detroit, Michigan 48202

Randal M. Hill

New Ventures Materials Physics, Dow Corning Corporation, 2200 West Salzburg Road,Midland, Michigan 48686

Received October 31, 2003. In Final Form: January 22, 2004

The self-associating structures at the solid-liquid interface of three nonionic trisiloxane surfactants((CH3)3SiO)2Si(CH3)(CH2)3(OCH2CH2)nOH (n ) 6, 8, and 12), or BEn, are studied as a function of substrateproperties by atomic force microscopy (AFM) imaging and force measurement. These trisiloxane surfactantsare known as superwetters, which promote rapid spreading of dilute aqueous solutions on low-energysurfaces. This study also attempts to relate the BEn surface aggregate structures at the solid-liquidinterface to their superwetting behavior. Four substrates are used in the study: muscovite mica, highlyoriented pyrolytic graphite, and oxidized silicon wafer with and without a full monolayer of self-assembledn-octadecyltrichlorosilane (OTS). The concentration of BEn is fixed at 2 times the critical aggregationconcentration (CAC). The BEn surfactants are only weakly attracted to hydrophilic surfaces, more onoxidized silicon than on mica. All three form ordinary planar monolayers on HOPG and OTS-coveredoxidized silicon. The significance of surfactant adsorption on the AFM tip is investigated by comparingthe force curves obtained by tips with and without thiol modification. The surface aggregate structuresof the BEn surfactants correlate with their bulk structures and do not exhibit anomalous adsorptionbehavior. The adsorption behavior of the BEn superwetters is similar to that of the CmEn surfactants. Thus,our results confirm previous work showing that superwetting shares its main features with other classesof surfactants.

Introduction

Siloxane surfactants are amphiphilic materials con-taining a methylated siloxane hydrophobe coupled to oneor more polyoxyethylene (EO) polar groups. Their surfaceand phase properties have been extensively researchedbecause of their use in antifoaming, in enhanced wetting,in bactericides, and in skin, hair, and fabric conditioning.1

Certain trisiloxane polyoxyethylene glycol surfactants(Me3SiO)2Si(Me)(CH2)3(OCH2CH2)nOH (n ) 5, 6, 8, and12), or BEn, have been shown to promote rapid spreadingof dilute aqueous solutions on low-energy surfaces.2,3

Recent research on the superwetters has been reviewedby Hill.4,5 Surfactants that form turbid dispersions arefound to enhance wetting. Other work has specificallylinked the presence of dispersed bilayer particles (vesicles)to enhanced wetting.6-8 Vesicle-containing solutions (BE6or BE8) spread faster than a micelle-containing solution(BE12).9,10 Other studies link the enhanced spreading rateto the formation of a bilayer at the spreading edge.11,12 A

maximum spreading rate at intermediate substratesurface energy has been found for both nonionic and ionicsurfactants,13 though only the trisiloxane surfactants areable to wet extremely hydrophobic surfaces. The linkbetween particles in solution and wetting, which is aninterfacial phenomenon, is not obvious and demandsdeeper investigation. The work reported here is inspiredby this enigma and explores the relationship between thesurfactant aggregate structures in solution and those atinterfaces.

The Soft-Contact atomic force microscopy (AFM) tech-nique images surfactant surface aggregates such ascetyltrimethylammonium bromide (CTAB), sodium dode-cyl sulfate (SDS), and poly(oxyethylene)n-dodecyl ether(C12En) by maintaining an image force as small as 10-12

N.14-16 Using this technique, we have found a surface phasetransition for C12E5 near its cloud point temperature onmica.17 BE12 has been found to exhibit aggregate structuralchanges at the solid-liquid interface that resemble its

* Corresponding author. E-mail: gzmao@eng.wayne.edu.† Current address: University of Minnesota, 350 Shepherd

Laboratories, 100 Union St. SE, Minneapolis, MN 55455.(1) Silicone Surfactants; Hill, R. M., Ed.; Surfactant Science Series,

Vol. 86; Marcel Dekker: New York, 1999.(2) Schwarz, E. G.; Reid, W. G. Ind. Eng. Chem. 1964, 56, 26.(3) Ananthapadmanabhan, K. P.; Goddard, E. D.; Chandar, P.

Colloids Surf. 1991, 44, 281.(4) Hill, R. M. Curr. Opin. Colloid Interface Sci. 1998, 3, 247.(5) Hill, R. M. Curr. Opin. Colloid Interface Sci. 2002, 7, 256.(6) Stoebe, T.; Lin Z.; Hill, R. M.; Ward, M. D.; Davis, H. T. Langmuir

1997, 13, 7270.(7) Svitova, T.; Hoffmann, H.; Hill, R. M. Langmuir 1996, 12, 1712.(8) Churaev, N. V.; Esipova, N. E.; Hill, R. M.; Sobolev, V. D.; Starov,

V. M.; Zorin, Z. M. Langmuir 2001, 17, 1338.

(9) Hill, R. M.; He, M.; Davis, H. T.; Scriven, L. E. Langmuir 1994,10, 1724.

(10) He, M.; Hill, R. M.; Lin, Z.; Scriven, L. E.; Davis, H. T. J. Phys.Chem. 1993, 97, 8820.

(11) Tiberg, F.; Cazabat, A. M. Langmuir 1994, 10, 2301.(12) Ruckenstein, E. J. Colloid Interface Sci. 1996, 179, 136.(13) Stoebe, T.; Lin, Z.; Hill, R. M.; Ward, M. D.; Davis, H. D. Langmuir

1996, 12, 337.(14) Manne, S.; Cleveland, J. P.; Gaub, H. E.; Stucky, G. D.; Hansma,

P. K. Langmuir 1994, 10, 4409.(15) Wanless, E. J.; Ducker, W. A. J. Phys. Chem. 1996, 100, 11507.(16) Patrick, H. N.; Warr, G. G.; Manne, S.; Aksay, I. A. Langmuir

1997, 13, 4349.(17) Dong, J.; Mao, G. Langmuir 2000, 16, 6641.

2695Langmuir 2004, 20, 2695-2700

10.1021/la036059b CCC: $27.50 © 2004 American Chemical SocietyPublished on Web 02/28/2004

bulk phase sequence.18 BE12 changes its surface aggregatestructure with increasing surface hydrophobicity, fromspherical micelles, to elongated micelles, to defectedmonolayer, and to continuous monolayer. Direct forcemeasurements have been conducted in polymeric silicone-and hydrocarbon-based surfactant solutions in order todetermine the adsorption behavior of the surfactants asa function of molecular chain structure and solventquality.19,20 The film thickness of the polymeric samplehas been found to be proportional to (number of oxyeth-ylene units)0.62. Here the adsorption behavior of thetrisiloxane surfactants at the solid-liquid interface isstudied by combining the Soft-Contact imaging and directforce measurement. We report the equilibrium aggregatestructures of BE6, BE8, and BE12 on mica, graphite,hydrophilic silica, and hydrophobic silica. We comparethe adsorption behavior of the trisiloxane surfactants tothat of hydrocarbon polyoxyethylene CmEn surfactants.

Experimental Section

Materials and Preparation. The BEn (n ) 6, 8, and 12)surfactants are used as received from Dow Corning Corp. with95% purity. n-Octadecyltrichlorosilane (OTS) is purchased fromUnited Chemical Technologies and distilled just before use.Hexadecane (99%) and carbon tetrachloride (99.9%) are pur-chased from Sigma and used as received. Reagent gradeammonium hydroxide, certified 30% hydrogen peroxide, andreagent grade concentrated nitric acid are purchased from FisherScientific and used as received. Water is deionized to 18 MΩ cmresistivity (Nanopure System, Barnstead). Grade 2, muscovitemica is purchased from Mica New York and hand-cleaved justbefore use. ZYH grade highly oriented pyrolytic graphite (HOPG)(12 × 12 × 2 mm3) is purchased from Advanced Ceramics andhand-cleaved with an adhesive tape until a smooth surface isobtained. Polished N type silicon (111) wafers are purchasedfrom Wafer World with resistivity between 50 and 75 Ω cm. Thesilicon substrate is cleaned following the “RCA clean” proceduresused in the integrated circuit manufacturing.21 OTS (0.007 mol/dm3) is dissolved in 16 cm3 hexadecane and 4 cm3 carbontetrachloride in a poly(tetrafluoroethylene) (PTFE) bottle. OTSdeposition is carried out in a nitrogen-purged glovebox. Afterwardthe substrate is rinsed thoroughly in chloroform. The substratesare stored in water before use. The OTS monolayers have beencharacterized by contact angle goniometer, Fourier transforminfrared spectroscopy in attenuated total reflection mode, andAFM.22

The contact angle values of BEn surfactants at 2 times thecritical aggregation concentration (CAC) and pure water on thevarious substrates are listed in Table 1. The contact angle ismeasured by an NRL contact angle goniometer (model 100, Rame-Hart) in the laboratory atmosphere. A droplet of 20 µL is placedon the substrate, and contact angles are read on both sides of thedroplet. Five droplets are placed at various spots near the center

of the substrate, and contact angles are averaged with an errorof (3°. Contact angles less than 10° cannot be measuredaccurately, and 0° is assigned in such cases. When the solutiondoes not wet the substrate completely, as in the cases of graphiteand the OTS monolayer, air bubbles can be trapped at the surface,which gives rise to unusual images and force curves. It is possibleto eliminate the bubbles by slowly purging the liquid cell withthe solution.

AFM Characterization. A Multimode Nanoscope IIIa AFM(Digital Instruments) is used. The substrate is mounted onto astainless steel disk and scanned in the aqueous solution of thesurfactant. The solution is injected through silicone rubber tubinginto a fluid cell (Digital Instruments), sealed by an O-ring. Allmeasurements are performed at 22 ( 1 °C. The AFM data reflectthe equilibrium structure of the surfactant surface aggregatesbecause no changes are observed between the first image or forcecurve collected, approximately 5 min after the solution injection,and subsequent data captured in a 3-h period. An E scannerwith a maximum scan area of 16 × 16 µm2 is used. Silicon nitrideintegral tips (NP type) are used with factory-specified nominaltip radii of curvature between 20 and 40 nm. All AFM imagesare deflection images unless specified. The integral and pro-portional gains are kept at a minimum of approximately 0.1. Theforce curves shown here are obtained from the same AFM tip sothat the force magnitude can be compared. The spring constantof the cantilever is calibrated using the deflection method againsta reference cantilever (Park Scientific Instruments) of knownspring constant (0.157 N/m).23 The measured spring constant,0.17 ( 0.05 N/m, is used in all the force curves. The forcecalibration plot is converted to a force and surface separationplot by defining the point of zero force and the point of zeroseparation.24 The tip to substrate velocity is maintained at 0.1µm/s.

Results and Discussion

The surface aggregate structures of the three surfac-tants adsorbed on four substrates are studied by simul-taneously imaging surface topography and measuring thesurface force (F) versus the surface separation distance(D) profile, or F(D). All force curves are taken in ContactMode unless specified. The force curves are quantified bythe force onset distance (at an arbitrarily defined valueof 0.2 nN), the maximum value of the force barrier, andthe jump-in distance. The values in three surfactantsolutions on four different substrates are listed in Table2. The CACs for BE6, BE8, and BE12 are 9.6 × 10-5, 1.09× 10-4, and 2.7 × 10-4 mol/L, respectively.25 All surfactantconcentrations are maintained at 2 × CAC. At thisconcentration, BE8 and BE12 form clear solutions whileBE6 forms a slightly cloudy solution.

(18) Dong, J.; Mao, G.; Hill, R. M. In Mesoscale Phenomena in FluidSystems; ACS Symposium Series 861; American Chemical Society,Washington, DC, 2003; Chapter 1, pp 2-16.

(19) Wang, A.; Jiang, L.; Mao, G.; Liu, Y. J. Colloid Interface Sci.2001, 242, 337.

(20) Wang, A.; Jiang, L.; Mao, G.; Liu, Y. J. Colloid Interface Sci.2002, 256, 331.

(21) Kern, W. J. Electrochem. Soc. 1990, 137, 1887.(22) Wu, B.; Mao, G.; Ng, K. Y. S. Colloids Surf., A 1999, 162, 203.

(23) Tortonese, M.; Kirk, M. SPIE 1997, 3009, 53.(24) Prater, C. B.; Maivald, P. G.; Kjoller, K. J.; Heaton, M. G. Probing

Nanoscale Forces with the Atomic Force Microscope; Application NoteNo. 8; Digital Instruments: Santa Barbara, CA, 1995; see also referencestherein.

(25) Svitova, T.; Hill, R. M.; Smirnova, Yu.; Stuermer, A.; Yakubov,G. Langmuir 1998, 14, 5023.

Table 1. Contact Angles (deg) of Pure Water and BEnSolutions at 2 × CAC on Various Substrates

substrate waterBE6 at

2 × CACBE8 at

2 × CACBE12 at

2 × CAC

mica 0 0 0 0silica 0 0 0 0graphite 75 27a 14a

OTS 108 10 19 73a Data taken from ref 32.

Table 2. Summary of Data Extracted from theForce-Distance Measurement Curves

substrate data type BE6 BE8 BE12

mica force onset at 0.2 nN (nm)barrier height (nN)jump-in point (nm) 5.0 4.3 4.8

silica force onset at 0.2 nN (nm) 6.0 13.0 12.0barrier height (nN) 1.8 1.0 1.3jump-in point (nm) 4.3, 13.0 3.9, 7.0 3.9

graphite force onset at 0.2 nN (nm) 6.4 7.5 8.5barrier height (nN) 2.8 3.2 6.6jump-in point (nm) 5.1 4.9 3.9

OTS force onset at 0.2 nN (nm) 8.1 10.8 11.0barrier height (nN) 8.6 10.0 20.6jump-in point (nm) 4.4 3.0 3.4

2696 Langmuir, Vol. 20, No. 7, 2004 Dong et al.

On Mica. The AFM images are featureless on mica forall three trisiloxane surfactants. Figure 1 is the F(D) curvein BE6 solution on mica. A purely attractive force isobserved at 5 nm, and it decreases in magnitude to -0.4nN at contact. The F(D) curves measured in BE8 and BE12

solutions on mica are similar to Figure 1 with slightvariations in the jump-in point. The jump-in point, asmarked in Figure 1, is defined as the start of the unstableregion in the force measurement when the cantilever tipsnaps suddenly onto the substrate. The jump-in occursfrequently during the cantilever-based force measure-ments due to the force gradient exceeding the springconstant. There exists a subtle difference between Figure1 and the F(D) curve obtained in pure water. F(D)measured in pure water displays a weak long-rangerepulsion before the attraction sets in.17 The absence ofthe electrostatic repulsion in the surfactant solutionsuggests some modification to the surface charge due tosurfactant adsorption, perhaps in the form of adsorbedisolated molecules or molecular patches. However, suchsporadic adsorption does not present the necessary stericrepulsion for the Soft-Contact imaging. We can concludethat on mica, the BEn surfactants behave similarly to theCmEn surfactants.26,27

On Hydrophilic Silica. Figures 2 and 3 summarizethe AFM imaging and force measurement results of thetrisiloxane surfactants. All the images presented here areobtained at a minimum image force. However, the readershould bear in mind that the imaged features are stillaffected by the contact force. The films consist of particles

Figure 1. The surface force as a function of surface separationdistance during the approach of the AFM tip to the mica surfaceobtained in 2 × CAC BE6 aqueous solution.

Figure 2. The AFM images on oxidized silicon wafers in BE6, BE8, and BE12 solutions: (a) BE6 (low magnification), (b) BE6 (highmagnification), (c) BE8, and (d) BE12.

Nanoscale Aggregates of Trisiloxane Surfactants Langmuir, Vol. 20, No. 7, 2004 2697

whose center-to-center distance varies with EO number,40-50 nm for BE6 (Figure 2a,b), 12-20 nm for BE8 (Figure2c), and 6-9 nm for BE12 (Figure 2d). All the force curvesshow the steric repulsion due to the dehydration of theEO headgroups, compression, and removal of the surfac-tant molecules from the contact zone. The maximum forceis measured to be 1.8 nN at 4.3 nm for BE6 (Figure 3a),1.0 nN at 3.9 nm for BE8 (Figure 3c), and 1.3 nN at 3.9nm for BE12 (Figure 3d). The force maximum is followedby the jump-in process, which corresponds to the removalof surfactant from the contact zone. After the jump-inprocess, the repulsion rises rapidly. The unperturbed filmthickness is estimated (from the onset force distance takenat F ) 0.2 nN) to be 6.0 nm for BE6, 13.0 nm for BE8, and12.0 nm for BE12. The fully stretched chain length iscalculated to be 2.8 nm for BE6, 3.5 nm for BE8, and 4.8nm for BE12. The bilayer thickness is 3.4 nm in BE6vesicles28 and 4.0 nm in the BE8 lamellar phase,10 in whichthe polyoxyethylene chain is believed to be in a coilstructure. BE12 forms micelles in solution with a diameterof 7.0 nm.29 The higher film thickness compared to therespective bulk structure is present in a number of AFMforce measurement studies of the nonionic surfactantsystem.16,19,20,27,30 It is therefore concludedthat theabsolutethickness value taken from the AFM force curves shouldbe treatedwithcaution dueto theuncertainty in surfactantadsorption on the tip and interpretation of the forcecurves.31 In a later part, we compare the F(D) curvesobtained by a unmodified tip and by a tip that is coatedwith a thiol monolayer in order to determine whether theBEn surfactants adsorb on the AFM tip.

In addition to the primary jump-in point, a second jump-in point at a longer separation is observed at 13.0 nm forBE6 (Figure 3b) and at 7.0 nm for BE8 (Figure 3c). In thecase of BE6, we find that when the tip is on top of a bump,two jump-in points are obtained, and when the tip isbetween the bumps, only one jump-in is recorded in the

F(D) curves. Therefore the bumps represent an additionaldiscrete bilayer with a thickness of 7.0 nm. In the case ofBE8, the coverage of the second bilayer must be higherbecause most of the force curves measured in the BE8solution contained two jump-in points. Both BE6 and BE8are known to form vesicles in solution. It is possible thatafter the formation of the anchoring bilayer by the fastmolecular diffusion, the adsorption and rupture of theirvesicles contribute to the second bilayer patches.

The adsorption of BEn on silica agrees largely with whatis known about CmEn surfactants: (1) hydrogen bondingis necessary for adsorption; (2) both surface micelle andbilayer structures are possible; and (3) the aggregate sizedepends on the molecular packing constraints. Forexample, surfactants with large headgroups such as BE12favor the formation of small, discrete micellar structure.A second discrete bilayer in the case of BE6 and BE8 isconsistent with the presence of dispersed bilayer struc-tures, such as vesicles, in the bulk phase.

On Graphite. The images captured in all threesurfactant solutions showed smooth surfaces. We do notobserve the stripelike structure as reported elsewhere.32

However, the strip pattern for BE8 is only observed afterwaiting for many hours in the previous study, while allour measurements are completed within 2-3 h aftersolution is injected. The stripe pattern is induced by one-dimensional epitaxy between all-trans hydrocarbon chainsand the graphite lattice.33 It is difficult to envision sucha match in the case of trisiloxane surfactants. The F(D)curves in the three surfactant solutions on HOPG areshown in Figure 4. All three curves exhibit repulsive forcesconsistent with the steric repulsion followed by the jumpinto contact. The unperturbed film thickness for BE6, BE8,and BE12 is determined at F ) 0.2 nN to be 6.4, 7.5, and8.5 nm, respectively. The steric barrier height and jump-in point are determined to be 2.8 nN and 5.1 nm for BE6,3.2 nN and 4.9 nm for BE8, and 6.6 nN and 3.9 nm for

Figure 3. The F(D) curves on oxidized silicon wafers in BE6, BE8, and BE12 solutions: (a) BE6 when force curves were takenbetween bumps, (b) BE6 when force curves were taken on top of the bumps, (c) BE8, and (d) BE12.

2698 Langmuir, Vol. 20, No. 7, 2004 Dong et al.

BE12. Both the steric barrier thickness and the heightincrease with increasing EO number. The increase in thefilm thickness with chain length is consistent with theend-on molecular adsorption configuration. Despite thethickness being longer than the calculated fully stretchedchain length, an ordinary planar monolayer structureremains the most likely structure. Bilayer formation isenergetically prohibited because of the hydrophobic natureof graphite, while the formation of 3 or more layers canbe ruled out by the presence of only one jump-in in theF(D) curves.

Unlike the CmEn surfactants that self-assemble intohemicylindrical hemimicelles, the BEn surfactants forma planar monolayer structure with vertically orientedmolecules on graphite. The vertical monolayer structureis further supported by the lack of a 0.4 nm steric barrierclose to the substrate as observed in the case of CmEn dueto the flat-lying template layer.30

On Hydrophobic Silica. On the silica that is hydro-phobized by an OTS monolayer, a smooth surface isobtained in all three surfactant solutions as representedby Figure 5a. Figure 5b shows the F(D) curves for BE6,BE8, and BE12 at 2 × CAC on OTS. The steric barrierthickness values measured at 0.2 nN are 8.1, 10.8, and11.0 nm for BE6, BE8, and BE12, respectively. The barrierheight and jump-in distance are 8.6 nN and 4.4 nm forBE6, 10.0 nN and 3.0 nm for BE8, and 20.6 nN and 3.4 nmfor BE12, respectively. The monotonic increase in the stericforce followed by a single jump-in indicates a monolayerstructure formed on the hydrophobic silica by all threesurfactants. Similar to the graphite case, both the stericbarrier thickness and height increase with increasing EOnumber. The magnitude of the steric barrier gives aqualitative measurement of the adsorbed amount since itis proportional to the surface pressure exerted by the

surfactant film.34 The wider and shorter trisiloxanehydrophobe is able to accommodate a larger hydrophilicgroup before the onset of lateral repulsion betweenneighboring headgroups, which tends to reduce packingdensity. The large flexibility of the trisiloxane group mayalso contribute to closer packing. The BEn surfactantsadsorb more strongly on OTS than on graphite withroughly 200% increase in the barrier height and 30%increase in the film thickness. The increase in the filmthickness suggests decreasing tilt in the surfactantorientation with respect to the substrate as a result ofcloser packing on OTS. The adsorption energy is expectedto be larger on the more hydrophobic OTS substratebecause of higher water/substrate and lower surfactant/substrate interfacial energy terms.

To determine whether BEn adsorbs on the silicon nitridetip, we compare the F(D) curves obtained in two configu-rations: methylated tip/surfactant solution/methylatedsubstrate and bare silicon nitride tip/surfactant solution/methylated substrate. A silicon nitride tip is hydropho-

(26) Rutland, M. W. Colloids Surf., A 1994, 83, 121.(27) Grant, L. M.; Ducker, W. A. J. Phys. Chem. B 1997, 101, 5337.(28) Li, X.; Washenberger, R. M.; Scriven, L. E.; Davis, H. T.; Hill,

R. M. Langmuir 1999, 15, 2278.(29) He, M.; Hill, R. M.; Doumaux, H. A.; Bates, F. S.; Davis, H. D.;

Evans, D. F.; Scriven, L. E. In Structure and Flow in SurfactantSolutions; ACS Symposium Series 578; American Chemical Society:Washington, DC, 1994; pp 192-216.

(30) Grant, L.; Tiberg, F.; Ducker, W. A. J. Phys. Chem. B 1998, 102,4288.

(31) Tiberg, F.; Brinck, J.; Grant, L. Curr. Opin. Colloid InterfaceSci. 2000, 4, 411.

(32) Svitova, T.; Hill, R. M.; Radke, C. J. Colloids Surf., A 2001, 183,607.

(33) Rabe, J. P.; Buchholz, F. Science 1991, 253, 424.(34) Eskilsson, K.; Ninham, B. W.; Toberg, F.; Yaminsky, V. V.

Langmuir 1999, 15, 3242.

Figure 4. The F(D) curves on graphite in BE6, BE8, and BE12solutions.

Figure 5. (a) The AFM image with scan size ) 600 × 600 nm2.(b) The F(D) curves on the OTS surface in BE6, BE8, and BE12solutions.

Nanoscale Aggregates of Trisiloxane Surfactants Langmuir, Vol. 20, No. 7, 2004 2699

bized by first sputter coating it with a thin film of gold (5nm chromium + 7 nm gold) and subsequently depositinga monolayer of 1-hexadecanethiol (>99%, Aldrich). Thetip is placed in 10 mM thiol ethanolic solution for 2 h.Figure 6 shows the F(D) curve measured in the BE8solution on the OTS surface with the thiol-coated tip, whichshows two jump-in points, one at 6.1 nm and the other at4.7 nm. The same measurement with the uncoated tipshows only one jump-in point at 3.0 nm. Two jump-inevents indicate that two layers are being pushed out atdifferent force magnitudes. In the case of the bare tip,only one layer is pushed out. However, we cannot rule outthe possibility of a thin yet strongly bound layer on thetip because such a layer cannot be distinguished from thehard substrate. The force values of the two configurationscannot be directly compared because the coated tip is adifferent tip with a different tip geometry.

In summary, the adsorption behavior of the trisiloxanesurfactants at the solid-liquid interface, as summarizedschematically in Figure 7, bears remarkable similarity tothat of the CmEn surfactants on both hydrophobic andhydrophilic surfaces. Thus our results confirm the perviousconclusion13 that superwetting, albeit enigmatic, sharesits main features with other types of surfactants. We donot find anomalous adsorption behavior for the trisilox-anes.

Conclusion

The self-associating aggregate structures of threetrisiloxane surfactants BE6, BE8, and BE12 are studied bySoft-Contact AFM imaging and direct force measurementsat the solid-liquid interface. There are many similaritiesbetween the BEn and CmEn surfactants that can besummarized below.

(1) Surface micelles or bilayers form on hydrophilicsurfaces. The aggregate size is connected to the molecularpacking parameter as in the bulk phase.

(2) The nonionic surfactant only adsorbs on hydrophilicsurfaces with hydrogen bonding sites.

(3) Monolayers form on hydrophobic surfaces, drivenby hydrophobic attraction.

(4) Surfactant adsorption, measured by both the stericbarrier thickness and height, increases with increasingEO numbers and surface hydrophobicity.

(5) The adsorbed amount of the BEn surfactant iscomparable with that of an equivalent CmEn surfactant.According to the literature, BE6 closely resembles C12E4in phase behavior, BE8 is similar to C12E5, and BE12 issimilar to C12E6.10 We obtain similar steric force magnitudenumbers for the pairs. For example, the steric barrierheight for C12E5 on OTS at 2 × CMC is 9.5 nN while forBE8 it is 10.8 nN at 2 × CAC when the same AFM tip isused.

The different structure formed by the BEn surfactantson graphite can be attributed to the lack of a dimensionalmatch between the trisiloxane hydrophobe and thegraphite lattice. The double bilayer structure found onsilica by the BE6 and BE8 surfactants can be linked to thedispersed bilayer structure in the bulk phase.

In conclusion, we find the surface aggregate structuresof the BEn surfactants essentially paralleling those of theCmEn surfactants, with differences reflecting the size,shape, and chemical nature of the trisiloxane hydrophobe.The surface aggregate structures resemble bulk struc-tures. Therefore, despite the enigmatic nature of thesuperwetting phenomenon, we do not find anomalousadsorption behavior for the trisiloxane superwetters. Thisstudy agrees with previous conclusions by Stoebe et al.that the superwetters display bulk and interfacial featuressimilar to those of other surfactants, especially the CmEnsurfactant series. The surface aggregate structures ofsurfactants are not only important for the understandingof various colloidal phenomena but may also offer insightinto the “bottom-up” approach to nanotechnology by theself-assembly of hierarchical structures.

Acknowledgment. We thank the ACS PetroleumResearch Fund (36149-AC5) and the National ScienceFoundation (CTS-0221586) for financial support.

LA036059B

Figure 6. The F(D) curve on the OTS surface in BE8 solutioncaptured with a thiol-modified AFM tip.

Figure 7. Schematic representations of BEn aggregate struc-tures at the solid/liquid interface on different substratesaccording to the AFM investigation. (a) Little adsorption onmica. (b) Continuous bilayer plus an additional discrete bilayerfor BE6 and BE8 and a discrete micellar layer for BE12 on silica.(c) Planar monolayer on graphite (with lower packing densitythan on OTS). (d) Planar monolayer on OTS-modified silica.

2700 Langmuir, Vol. 20, No. 7, 2004 Dong et al.