)OUR.'lAL OF COU.OID AND INTERFACE SCIENCE 191, 202-208 ( 1997)ARTICLE NO. CS974923

Lei Zhang. P. Somasundaran, I and C. Maltesh 2

lAngmuir Center for Colloids & Inteifaces, Columbia Universi~'. 911 Mudd Building, New York. Ne\v York 10027

Received January 31. 1997; accepted AprilS. 1997

Adsorption of a typical sugar-based surfactant, n-dodecyl-{3-o-maltoside (OM), on hydrophilic solids, silica, alumina, titania,and hematite, and a hydrophobic solid, graphite, was studied.Effects of salts and pH on the adsorption on alumina as well as theelectrokinetic potential of the particles after surfactant adsorptionwere studied to determine the adsorption mechanisms. Hydropho-bicity and settling rate were measured to explore the surfactantconformation on the particle surfaces. For hydrophilic solids, OMwas found to adsorb strongly on alumina, titania, and hematitebut weakly on silica. While hydrogen bonding is postulated to bethe major driving force for the adsorption on hydrophilic solids, forhydrophobic solid, the adsorption is mainly due to the hydrophobicinteractions. The different behaviors of surfactant on hydrophilicand hydrophobic solids were attributed to the different interactionsbetween surfactant and solids. Also. the surfactant is estimated toform a bilayer on alumina while on graphite it forms a monolayer.The surface hydrophobicity and stability of the solids are discussedin terms of the adsorbed monolayer I bilayer formation on the parti-cles. Q 1997 A£ademic Press

Key Words: adsorption; zeta potential; settling rate; hydropho-bicity; n -dodecyl-{3-o-maltoside; sugar-based surfactant; alumina;hematite; titania; silica; graphite.

behavior of these surfactants (9-11), infonnation on ad-sorption of alkyl polyglucosides on solid substrates is lim-ited. Smith et al. (12) measured the adsorption isothenns ofthree alkyl polyglucosides on titanium dioxide. The resultsshowed the adsorption of alkyl polyglucoside on a TiO2surface to exceed tha( required to fonD a theoretical mono-layer. Smith et al. postulated that the hydroxyl groups ofthe surfactants are slightly acidic in nature and can hydrogenbond with the basic OH groups on the TiO2 particles. Resultsof the mechanism studies of these reagents have been oflimi(ed use due to the impurities in the surfactants, and thetests were done using commercial samples composed ofcomplex mixtures.

Adsorption behavior of another group of nonioDic surfac-tants, the ethoxylated surfactants (C;Ej), has been exten-sively investigated using a range of techniques (13-18). Itwas found that ethoxylated surfactants can adsorb on silicabut not on alumina (15-17). Hydrogen bonding is postu-lated as the adsorption mechanism in this case (18). It isuseful to compare the adsorption properties of alkyl polyglu-cosides to those of ethoxylated surfactants, particularly inreference to their application in processes such as wetting,dispersion, and separation.

In this study the adsorption behavior of n-dodecyl-,8-o-maltoside (OM) on hydrophilic alumina, silica, hematite,and (itania and hydrophobic graphite was studied. The chem-ical structure and molecular model of n-dodecyl-,8-o-malto-side are shown in Fig. I. Notice the dimensions of the hydro-philic maltose head group and the hydrophobic dodecylchain. To elucidate the mechanism of adsorption, the effectsof salt and pH on the adsorption of OM on alumina werealso studied. Elec(rophoretic measurements were conducted(0 determine the effect of surfactant adsorption on the in-terfacial potential of the particles. Wettability and settlingrate were measured to monitor the relative hydrophobicityof particle surfaces and to explore the confonna(ion of ~eadsorbed species.

INTRODUCTION

Alkyl polyglucosides are a relatively new class of non-ionic surfactants finding increasing applications in many in-dustrial fields using processes such as detergency, emulsifi-cation, dispersion, wetting, and solubilization (I, 2). Theyare used in soaps and cosmetic products since they offergood detergency properties and are very mild to skin (3,4). Also, they have potential biological and pharmaceuticalapplications and are used in biological studies for solubiliz-ing membrane proteins without denaturation (5, 6). In addi-tion to the above properties, these surfactants can be obtainedfrom renewable materials and are easily biodegradable (7,8), and hence their use has considerable potential.

Although there are a number of articles on the solutionEXPERIMENTAL

Surfactants. n -Dodecyl-.8-o-maltoside was obtainedfrom Calbiochem and was used as received. The purity deter-

1 To whom correspondence should be addressed.! Current address: Corporate Research Division, Nalco Chemical Co.

Naperville. IL 00563.

2020021-9797/97 $25.00Copyright e 1997 by Academic PlessAll rights of reproduction in any fonn reserved.

n-OODECYL-.8-D-MALTOSmE ADSORPnON 203

~

aHO

"V~,,"""---""",/

H

b

FIG. 1. (a) Chemical stnIcmre of n-dodecyl-,8-D-maltoside; (b) Molecular model of n-ckl98% withdodecanol and residues of maltose as impurities. Elementalanalysis showed that it contains 55.27% carbon and 9.06%

hydrogen.

Reagents. NaOH purchased from Fisher Scientific Co.was certified as volumetric standard solutions (0.1 N).Na2S04 was from Fisher Scientific Co.; the water used wastriple distilled.

Solids. Alumina AKP-50 obtained from Sumitomo hada mean diameter of 0.2 JLm. The BET specific surface areameasured using nitrogen with a Quantasorb system was 10.8m2 / g and the isoelectric point

204 ZHANG. SO~IASUNDARAN. AND MAL TESH

when these isothenns are examined. In the first stage ofadsorption, the surfactant is proposed to adsorb individuallyand sparsely on the surface and chain-chain interaction isnegligible. A sharp increase in the adsorption density occursin the second stage, and this is proposed to be the associationof the surfactant into hemimicelles due to the chain-chaininteraction ( 19, 20). The adsorption isothenn reaches a pla-teau region at the onset of the third stage. The inflectionpoint between stages II and III corresponds to critical micelleconcentration of the surfactant.

In the plateau region, the adsorption density is about 5.5X 10-6 mol/m2 and the surface area per molecule adsorbedis calculated to be 30 A 2. Compared with the value derivedfrom the surface tension at the solution-air interface, whichis 48.5 A:!, the amount of surfactant adsorbed on the particlesis estimated to be more than that required to form a theoreti-cal monolayer and is in fact close to that required for abilayer. However, this value can be explained equally wellby assuming adsorption of surface micelles since a widerange of values can be obtained depending on the shape ofadsorbed micelles and their distribution on the surface.

Infonnation on changes in relative hydrophobicity of thesolid surface due to surfactant adsorption can shed light onthe confonnation of surfactant on the solid surface and helpto elucidate the mechanism involved. The effect of surfactantadsorption on the wettability of the alumina is illustrated inFig. 3 along with the adsorption isothenn. In the absence ofthe surfactant the alumina exhibits complete hydrophilicity.With an increase in adsorption on alumina. the surface be-comes hydrophobic due to increasing amounts of surfactantadsorbing with their hydrophobic tails oriented toward thebulk solution. The hydrophobicity reaches a maximum atthe beginning of region II and then drops. The drop in hydro-phobicity after the maximum suggests that the onset of the

1 10 100 1000 .10000

a-do4ecyI+O ;de - c-InIICMI. I 0 "\t

FIG. 2. Adsorption isodlCrms 9fn~l-p-o-!naltosidC on hydro-philic solids at 2S"C and ~tI'81 pH: (~)SiOz, (6) AlzO), (0) Fe2~'and (0) TiOz.

shaken for 1 min manually and then allowed to settle for 10min. The bulk of the aqueous phase with hydrophilic solids,as well as the toluene phase with hydrophobic solids, wasthen allowed to flow out of the funnel separately. The twophases containing the solids were evaporated and the weightof alumina was recorded. The relati:vepercentage hydropho-bicity was determined as

Weight of alumina in toluene phase x 100.Weight of alumina in toluene phase +

weight of alumina in aqueous phase

;-~

l81

t-JO

~:... .

:.~

-.K) :£'

:';0

..5

-10

- _1- ~ r --= -""'1 I I r - ._.a-g.~

! ~

{t\;'

Settling rate. The settling rate measurement was donein 40 ml vials under the same conditions as those used foradsorption tests. The dispersion was tumbled manually, thedescent of the upper interface was observed, and the positionand time of the upper interface were recorded. The slope ofthe plot of the height of the interface against time was usedas a measure of settling rate.

All the experiments were carried out at 25°C.

fI t~IO' ,.RFSULTS AND DISCUSSIONS cv100

~

1.10

Adsorption of n-DodeL-yI-{3-o-maltoside on HydrophilicSolids

Adsorption isotherms of n-dodecyl-{3-o-maltoside on alu-mina. silica. titania, and hematite are shown in Fig. 2. Thesurfactant is found to adsorb on alumina, hematite, and tita-nia, while it does not absorb on silica to the same extent,particularly in the high concentration range.

The adsorption isotherms of n-dodecyl-{3-o-maltoside onalumina and hematite are very similar and that on titania issomewhat similar. A three-stage adsorption becomes evident

I 10 100 1000 100004

.~__10 )(

FIG. 3. Adsorption of n-dodecyl-,8-o-maltoside and its effect on thehydrophobicity of alumina panicles as detennined by two phase separation:(0) adsorption density and ( . ) hydrophobicity.

205n -DODECYL-.8-D-MAL TOSIDE ADSORPTION

,wo

Ixl."

~

fl...

,.0'I 10 100 1000 10000

-,I~-__IO"MFIG. 4. Effect of sodium sulfate on the adsorption of n-dodecyl-{3-o.

maltoside on alumina: (. ) 0.25 M Na:SO.. and (0) without salt.

-1 10 100 1000 10000

.-dooIecyI~__IO"J,I

AG. S. Effect of pH on the adsorption of n-dodec~i-/3-D-maltoside onalumina: (8) pH 10.5 and (0) pH 7.4.

by the change in pH in the range tested. Since the isoelectricpoint of AKP-50 alumina is 8.9, the zeta potential of thealumina is very different under the two pH conditions tested.The identical adsorption isothenns obtained under the twopH conditions suggest that the electrostatic interaction is nota dominant factor in determining the adsorption of n -dode-cyl-,8-o-maltoside on alumina. Electrophoretic measure-ments also show that the zeta potentials of alumina andtitania are not altered significantly by the adsorption of OMsurfactant (Fig. 6). These results are in accord with thenonionic nature of n -dodecyl-,8-o-maltoside and eliminatethe possibility of electrostatic interactions between the solidand the surfactant.

The driving force for the adsorption in the present caseis postulated to be hydrogen bonding. In their study of dis-persion of TiO2 by alkyl polyglucoside surfactants, Smith etat. proposed that the hydroxyl groups on the surfactant were

,.I

10 tOO 1000

.-,I~_-.IO"~

100)}

chain-chain interaction is causing some hydrophilic groupsto orient toward the aqueous phase. The hydrophobicity ofthe alumina drops further as the adsorption reaches the pla-teau region. The minimum hydrophobicity at the plateauregion is possibly caused by the bilayer adsorption. sincethat can render the alumina surface hydrophilic. Further vari-ation in hydrophobicity in the plateau region is possiblyrelated to the effect of high residual DM on the two-phaseseparation that is used here to monitor hydrophobicity.

Note that the bilayer confonnation of n-dodecyl-{3-D-mal-toside on alumina may not be a regular homogeneous bilayer.Due to the bulky nature of the maltose head group. closepacking of the adsorbed layer will require the hydrocarbonchains of opposing surfactants to interpenetrate each otherso that there is actually one layer of hydrocarbon chainswith two layers of head groups on each side. The cross-sectional area of the head group is 48.5 A.2, while that ofthe paraffin chain is about 20 A 2. Hence the arrangementproposed above can be considered a strong possibility.

The effect of Na2S04 salt on the adsorption of n-dodecyl-{3-D-maltoside on alumina is shown in Fig. 4. The adsorptionisothenn is found to shift to the left in regions I and II anddownward in region III in the presence of salt. Similar resultshave been reported for the adsorption of ethoxylated surfac-tant on silica in the presence of Na2S04 ( 18). The effect ofsalt is attributed to the salting out of the hydrocarbon chainof the surfactant by the salt. The inflection point betweenregions II and III drops from 1.8 x 10-5 to about 9 x 10-6moltL. At the same concentration, Na2S0.. has been reportedto reduce tl-le cmc of n-dodecyl-{3-D-maltoside from 1.8 x10-5 to 9.4 X 10-6 moltL (21). The comparable shiftingof the inflection point on the adsorption isotherm suggeststhat the shifting isotheml.due to Na2S04 can be attributedprimarily to changes in the solution rather than to those onthe solid surface.

The effect of pH on the adsorption of DM on alumina isshown in Fig. 5. The maltoside adsorption is not affected

FIG. 6. Zeta potentials of alumina (.) and titania (.) on n -dodecyl.8-D-maltoside adsorption at neutral pH.

p

206 ZHANG. SOMASUNDARAN. AND MAL TESH

1 10 100 1000 10000

.~ide IaiduoI ~1IU:JQOtI. 10" M

FIG. 7. Adsorption isodtenns of n-dodecyl-8-o-maltoside on hy-drophobic solid graphite (. ) compared with hydrophilic solid alumina (0 )at 2SoC and neutral pH.

slightly acidic in nature and hydrogen bonded with the basicOH groups on the surface of the TiO~ particles, althoughthere was no evidence presented for this mechanism in theirwork ( 12). Hydrogen bonding has been proposed to be thedriving force for the adsorption of ethoxylated surfactantsas well (15, 18). However, polyethylene oxide has beenreported to show strong adsorption on silica, but not onalumina (15-17). Similar results have also been reportedfor the adsorption of polyethylene oxide polymers on thesesolids (22). It is interesting that n -dodecyl-{:J-o-maltosideshows a behavior opposite to that of the polyethylene oxidesurfactant. It adsorbs on alumina and hematite but much lesson silica. This preferential adsorption should have potentialapplications in many processes such as dispersion and flota-tion.

proposed to be due to the reconstruction of the adsorbedlayer with the chains standing vertically and packing closely.

In the plateau region, the adsorption density is 3 X 10-6mol 1m 2 and the surface area per molecule in this region is

estimated to be 55 A 2, which is close to the value, 48.5 A 2,derived from the surface tension at the solution-air interfaceand half of that for alumina. The amount of surfactant ad-sorbed on the graphite is close to a monolayer.

The conformation of the adsorbed layer on graphite isillustrated in the form of the effect of surfactant adsorptionon the wettability and stability (see Fig. 8). Both tests giverelative hydrophobicity of the solid. In wettability tests, thegraphite surface exhibits complete hydrophobicity in the ab-sence of the surfactant. With adsorption on graphite, thesurface becomes increasingly hydrophilic due to the surfac-tant orientation with the hydrophobic heads attached to thesurface and the hydrophilic parts dangling into solution. Theeffect of the surfactant on the stability of graphite particlesas measured by the settling rate is shown in Fig. 9. TheAdsorption of n-dodecyl-f3-D-maltoside on Hydrophobic

Solids

The adsorption isotherm of n-dodecyl-I3-D-maltoside onthe hydrophobic graphite is shown in Fig. 7. The adsorptionisothenn on graphite is quite different from that on alumina.The adsorption density increases very quickly at a very lowDM concentration and reaches a plateau far below the cmcof the surfactant. Aggregation of the surfactant in the bulksolution itself does not appear to affect the adsorption pro-cess. This result is in accord with the literature on the adsorp-tion of nonionic surfactants on hydrophobic surfaces (23).Due to the hydrophobic nature of the solid, surfactants ad-sorb on the solid with hydrophobic groups attaching to thegraphite surface by hydrophobic interactions (24). Thesharp increase in the isotherm at low OM concentration sug-gests very strong interaction with much of the surfactantpartitioning to the solid-solution interface. Above the in-flection point, the slow increase in adsorption density is

, 1 t ~ . [:m.1 10 100 1000 ' I«X»

~ miduol-.IO"..

FIG. 9. Adsorption isotherm of n-dodecyl-/3-o-maltoside on graphiteand its effect on the stability of graphite suspension: (D) adsorption iso-therm and (8) settling rate.

n -DODECYL-.8-o-MAL TOSIDE ADSORPTION 201

Hydrophilic solid Hydrophobic solid

A

B

c

0 hydrophilic ~d group

~ : hydrophobic tailFIG. 10. Proposed orientation model for the adsorption of n-dodecyl-fJ-D-maltoside on hydrophilic and hydrophobic surfaces. showing d1e conformation

of surfactant at the surfaces. A. B. and C indicate the successive stages of adsorption.

graphite suspension is unstable without the surfactant. AsOM adsorbs, the settling rate is found to decrease sharply.Complete dispersion of graphite occurs at concentrationscorresponding to the onset of plateau on the adsorptioncurves. The stabilization of hydrophobic suspensions by non-ionicsurfactants has been observed for many other systems(25,26).

The above observations suggest that the adsorption of n-dodecyl-{3-o-maltoside on graphite is monomolecular withthe nonpolar groups attached to the solid surface and thehead groups oriented toward the bulk solution.

SUMMARY

Adsorption of n-dodecyl-,B-o-maltoside on various hydro-philic ,and hydrophobic solids has been studied. The adsorp-tion of surfactant on hydrophilic and hydrophobic solidshows different behavior due to the markedly different inter-actions between the surfactant and the solid surface. In thecase of hydrophilic solids, the adsorption is proposed to be

due to the hydrogen bonding between the solid surface andthe hydrophilic groups of the surfactant and hydrophobicinteraction between the surfactant chains at high surfactantconcentration. while the hydrophobic solid exhibits exclu-sively hydrophobic interactions between solid and surfactantas well as chain-chain interaction between the surfactants.

An orientation scheme for the adsorption of n-dodecyl-,8-o-maltoside on hydrophilic and hydrophobic solids undervarious conditions is proposed in Fig. 10. The adsorption ofsurfactant on hydrophilic solids such as alumina is dividedinto three regions. At lower concentrations. the surfactantadsorbs on the solid surface individually due to hydrogenbonding. At still higher concentrations interactions betweensurfactant chains take place, leading to a steep rise in adsorp;tion. Above the cmc of the surfactant the adsorption reachesa plateau. The confonnation of the saturated adsorbed layeris close to that of a bilayer with hydrophobic chains interpen-etrating each other.

The adsorption on hydrophilic solids shows the surfactantto absorb on alumina. hematite. and titania but much less

208 ZHANG. SOMASUNDARAN. AND MAL TESH

on silica. This behavior on hydrophilic solids is opposite tothat of alkyl polyethylene oxide surfactant, and this uniquebehavior of the sugar-based nonionic surfactant has practicalimplications. The effect of salt on the adsorption is attributedto the salting out of the surfactant hydrophobic chain. Theadsorption is not affected by the pH changes in the solution,and the zeta potentials of the solids are not altered by thesurfactant adsorption. It is hence proposed that specific inter-actions such as hydrogen bonding between the surfactantand basic solids are responsible for the adsorption of n-dodecy 1-,8- D- maltos ide.

The adsorption on hydrophobic solid graphite is dividedinto two parts. At low concentrations, the surfactant adsorbson the solid surface individually, with the hydrophobic chainattaching to the solid surface. At high concentrations thesurfactant tends to aggregate into a close-packed monolayeron the surface, with hydrophobic parts attaching to the sur-face.

On both types of solids the final adsorption state is theone in which the surface becomes hydrophilic. This is sup-ported by the results obtained for the wettability and stabilitychanges of the solids upon surfactant adsorption. These prop-erties enable n-dodecyl-.B-D-maltoside to be utilized in vari-ous wetting, detergency, dispersion, and flotation processes.

ACKNOWLEDGMENTS

The authors acknowledge financial suppon of the National Science Foun-dation crS-962278 I and technical discussions with Dr. B. Pethica and Dr.Alben Chan (ARCO Oil and Gas Co.).

REFERENCES

1. Hughes, F. A.. and Lew, B. A., J. Am. Oil Chern. Soc. 47, 162 (l970).2. Balzer, D., Tenside Suif Det. 28, 419 (l991).3. Putnik, C. F.. and Borys. N. F., Soap Cosmet. Chern. Spec. 61(6),34

(l986).

4. Salka, B., Cosmet. Tolletries 108, 89 (1993).5. Helenius. A., McCaslin. D. R., Fries, E., and Tanford. C., Methods

Enzymol. 56, 734 (1979).6. Stubbs, G. W., Smith, H. G., Jr., and Litman, B. I., Biochirn. Bioph}-s.

Acta 425, 46 (1976).7. Matsumura, S., lmai, K., Yoshikawa, S., Kawada, K., and Uchibori,

T., J. Am. Oil Chern. Soc. 67,996 (1990).8. BjorkJing, F., Godtfredsen, S. E., and Kirlc. 0., J. Chern. Soc.. Chern.

Commun. 934 (1989).9. Shinoda, K., Yamaguchi. T., and Hori. R., BulL Chern. Soc. Jpn. 34,

237 (1961).10. Drummond, C. I., Warr, G. G., Greiser, F., Ninham, B. W., and Evans,

D. F., J. Phys. Chern. 89,2103 (1985).11. Warr, G. G., Drummond, C. I., Greiser. F., Ninham. B. W., and Evans.

D. F., J.. Phys. Chern. 90,4581 (1986).12. Smith, G.A., Zulli, A. L., Grieser, M. D., and Counts, M. C., Colloids

Su1j: 88,67 (1994).13. Partyka, S., Zaini, S.. Lindheimer, M.. and Brun, B., Colloids Su1j: 12,

255 (1984).14. Tronel-Peyroz, D.. Schuhmann, D., Raous, H., and Bertrand, C., J.

Colloid Interface Sci. 97,541 (1984).15. Somasundaran, P., Snell. E. D., and XU. Q., J. Colloid Interface Sci.

144,165 (1991).16. Fu. E., ., Adsorption of Anionic-Nonionic Surfactant Mixtures on Ox-

ide Minerals:' Ph.D. thesis, Columbia University, New York, 1987.17. Lawrence, S. A., Pilc, I. A., Readman, I. R., and Sennon, P. A., J.

Chern. Soc., Chern. Commun. 1035 (1987).18. Doren. A., Vargas, D., and Goldfrab, I.,lnst. Min. Metall. Trans.. Sect.

C 84,34 (1975).19. Kunjappu, I. T., and Somasundaran. P., J. Colloid Interface xi. 175,

520 (1995).20. Somasundaran, P., and Krishnakumar. S., Colloids SUfj: 93, 79 ( 1994).21. Zhang, L., Somasundaran. P., and Maltesh, C., LiJngmuir 12, 2371

(1996).22. Koksal, E., Ramachandran, R., Somasundaran, P., and Maltesh, C.,

Powder TechnoL 62,253 (1990).23. Corkill, I. M., Goodman, I. F., and Tate, I. R., Trans. Faraday Soc.

62, 979 (1966).24. Ben-nairn, A., .'Hydrophobic Interactions:' Plenum. New York., 1980.25. Glazman, Y., and Blashchuk, z., J. Colloid Interface Sci. 62, 158

(1977).26. Mathai, K. G., and Ottewill, R. H.. Trans. Faraday Soc. 62, 750

(1966).