Calorimetric Study of the Adsorption of Short-Chain Nonionic

to study the nature of the adsorption process for short- ... surfactant, n-octyl β-D-monoglucoside, and a short-chain ..... (42) Frank, H. S.; Evans,...
0 downloads 0 Views 100KB Size
8842

Langmuir 2000, 16, 8842-8849

Calorimetric Study of the Adsorption of Short-Chain Nonionic Surfactants on Silica Glass and Graphite: Dimethyldecylamine Oxide and Octyl Monoglucoside† Z. Kira´ly*,‡ and G. H. Findenegg§ Department of Colloid Chemistry, University of Szeged, Aradi Vt. 1, H-6720 Szeged, Hungary, and Iwan-N.-Stranski-Institut fu¨ r Physikalische und Theoretische Chemie, Technische Universita¨ t Berlin, Strasse des 17. Juni 112, D-10623 Berlin, Germany Received January 20, 2000. In Final Form: April 10, 2000 The material and enthalpy balances of adsorption of the nonionic surfactants N,N-dimethyldecylamineN-oxide (C10DAO) and n-octyl β-D-monoglucoside (C8G1) from dilute aqueous solutions onto hydrophilic silica glass and hydrophobic graphite (graphitized carbon black) were determined at 298.15 K up to the critical micelle concentration. An automated flow sorption/microcalorimeter system was used for simultaneous measurements of the adsorption isotherm and the enthalpy isotherm of displacement. The formation of the adsorption layer is discussed in terms of the differential molar enthalpy data of adsorption as a function of surface coverage, and the results are related to the aggregated structure of nonionic surfactants at silica/solution and graphite/solution interfaces studied by atomic force microscopy. On silica, a low-density adsorption region (exothermic) is followed by a high-density adsorption region to produce globular surface aggregates of both C8G1 and C10DAO. On graphite, the formation of a flat, ordered monolayer (exothermic) is followed by the formation of C8G1 surface hemicylinders or, probably, a flat, less ordered bilayer of C10DAO. In either case, the adsorption in the high-density adsorbate region is endothermic, like micelle formation in aqueous bulk solution, as is to be expected on the basis of current models of surface aggregation of nonionic surfactants on hydrophilic and hydrophobic surfaces.

Introduction Measurements of adsorption isotherms and calorimetric determinations of the enthalpies of adsorption of alkyl polyoxyethylenes,1-10 alkylbenzene polyoxyethylenes,11-23 and alkylsulfinylalkanols24 from aqueous solutions onto * Corresponding author. E-mail: [email protected]. † Part of the Special Issue “Colliod Science Matured, Four Colloid Scientists Turn 60 at the Millennium”. ‡ University of Szeged. § Technische Universita ¨ t Berlin. (1) Gellan, A.; Rochester, C. H. J. Chem. Soc., Faraday Trans. 1 1985, 81, 3109. (2) Gellan, A.; Rochester, C. H. J. Chem. Soc., Faraday Trans. 1 1986, 82, 953. (3) Seidel, J.; Wittrock, C.; Kohler, H. H. Langmuir 1996, 12, 5557. (4) Kira´ly, Z.; Bo¨rner, R. H. K.; Findenegg, G. H. Langmuir 1997, 13, 3308. (5) Hey, M. J.; MacTaggart, J. W.; Rochester, C. H. J. Chem. Soc., Faraday Trans. 1 1984, 80, 699. (6) Gellan, A.; Rochester, C. H. J. Chem. Soc., Faraday Trans. 1 1985, 81, 1503. (7) Corkill, J. M.; Goodman, J. F.; Tate, J. R. Trans. Faraday Soc. 1966, 62, 979. (8) Findenegg, G. H.; Pasucha, B.; Strunk, H. Colloids Surf. 1989, 37, 223. (9) Klimenko, N. A.; Polyakov, V. E.; Permilovskaya, A. A. Colloid J. USSR 1979, 41, 913. (10) Klimenko, N. A. Colloid J. USSR 1980, 42, 466. (11) Denoyel, R.; Rouquerol, F.; Rouquerol, J. In Proceedings of the 2nd Engineering Foundation Conference on Fundamentals of Adsorption (Santa Barbara, CA, May 4-9, 1986), Liapis, A. I., Ed.; American Institute of Chemical Engineers: New York, 1987; p 199. (12) Partyka, S.; Lindheimer, M.; Zaini, S.; Keh, E.; Brun, B. Langmuir 1986, 2, 101. (13) Partyka, S.; Lindheimer, M.; Faucompre, B. Colloids Surf. 1993, 76, 267. (14) Denoyel, R.; Rouquerol, F.; Rouquerol, J. In Adsorption from Solution; Ottewill, R. H., Rochester, C. H., Eds.; Academic: London, 1983; p 225. (15) Noll, L. A. Calorim. Anal. Therm. 1985, 16, 12. (16) Lindheimer, M.; Keh, E.; Zaini, S.; Partyka, S. J. Colloid Interface Sci. 1990, 138, 83. (17) Giordano, F.; Denoyel, R.; Rouquerol, J. Colloids Surf. 1993, 71, 292.

silica1-4,11-23 and graphitized carbon black2,5-10,24 exemplify the significant differences between the adsorption behavior of nonionic surfactants on hydrophilic and hydrophobic solid surfaces. It has been established that in both cases the adsorption proceeds in two distinct steps. On hydrophilic silica,1-4,11-23 the first step is a low-affinity adsorption involving an exothermic adsorption of monomers, but leading only to a low surface coverage. The second stage, which occurs at close to the critical micelle concentration (cmc), is a highly cooperative adsorption process in which the measured enthalpy changes are similar in magnitude to the corresponding enthalpies of micelle formation in the bulk solution. This finding has been taken as an indication that adsorption in this second step involves a surface aggregation at the solid/liquid interface, driven by hydrophobic interactions similar to surfactant aggregation in the bulk solution.3,4,11-14,16-18,22,23 Depending on the systems studied, classical bilayer or globular associates have been proposed as the structure of the surface aggregates. In the case of graphite,2,5-10,24 on the other hand, the first step is a high-affinity adsorption, which is again exothermic but effectively irreversible, leading to a surface coverage of about one close-packed layer of molecules, oriented with their long axes parallel to the surface. The second stage of adsorption is athermal or endothermic, but there is some debate concerning what actually happens in this region. A gradual transition from (18) Giordano-Palmino, F.; Denoyel, R.; Rouquerol, J. J. Colloid Interface Sci. 1994, 165, 82. (19) Noll, L. A. Colloids Surf. 1987, 26, 43. (20) Seidel, J. Thermochim. Acta 1993, 229, 257. (21) Noll, L. A.; Gall, B. L. Colloids Surf. 1991, 54, 41. (22) Denoyel, R.; Rouquerol, J. J. Colloid Interface Sci. 1991, 143, 555. (23) Thomas, F.; Bottero, J. Y.; Partyka, S.; Cot, D. Thermochim. Acta 1987, 122, 197. (24) Corkill, J. M.; Goodman, J. F.; Tate, J. R. Trans. Faraday Soc. 1967, 63, 2264.

10.1021/la000065f CCC: $19.00 © 2000 American Chemical Society Published on Web 06/01/2000

Adsorption of Short-Chain Nonionic Surfactants

a horizontal monolayer to a vertical monolayer,2,5-7,24 the formation of three-dimensional associates, like halfmicelles,9,10 or the formation of subsequent layers on top of the preceding layer (parallel bilayer or multilayer)8 have been proposed to occur. Calorimetry provides primary information on the energetics and thermodynamics of adsorption, but a characterization of the structure of the adsorption layer at a molecular level requires complementary structuresensitive methods. Atomic force microscopy (AFM) has proved to be particularly well suited for this purpose.25-29 AFM studies on the adsorption of a series of alkyl polyoxyethylenes (CmEn) on hydrophilic silica indicated that the monomers adsorbed in the first step may serve as nuclei for globular surface micelles,26 while the classical bilayer structure occurs less frequently.25,26 The AFM images of CmEn26,27 and N-alkylmaltonamides28 adsorbed on the basal planes of graphite revealed strong surfacedirected ordering of the surfactant molecules in a headto-head, tail-to-tail arrangement. In most cases, this closepacked, horizontal monolayer templated further adsorption to form half-cylindrical adsorbed micelles (surface hemicylinders), though planar bilayers or unstructured uniform layers were sometimes identified.26-28 The results of our calorimetric study on the adsorption of dodecyltrimethylammonium bromide on graphitized carbon black30 were consistent with those of AFM studies on the same and related systems,29 suggesting that for ionic surfactants the hemicylinder is the preferred configuration on graphite at concentrations near the cmc. It has become clear that many of the aggregate geometries which exist in bulk solution may also develop at interfaces. However, the presence of the solid surface invokes geometric constraints on, and energetic interactions with, adsorbing surfactant molecules, thereby modifying their self-assembly behavior. It may be inferred from the results of AFM studies on graphite that surfaceinduced alignment of the first layer occurs only above a critical alkyl chain length within a homologous series. Below the critical chain length, the arrangement of the surfactant molecules is laterally homogeneous, driven by a minimization of the water-graphite interfacial area. However, the critical alkyl chain length depends on the nature of the polar headgroup. The present work sets out to study the nature of the adsorption process for shortchain nonionic surfactants, i.e., chain lengths close to the critical alkyl chain length of ca. 8-10 carbon atoms. Our calorimetric technique is most suitable for such shortchain surfactants for which the region below the cmc is readily accessible. We report on the adsorption isotherms and enthalpies of adsorption of a short-chain sugar surfactant, n-octyl β-D-monoglucoside, and a short-chain amine oxide surfactant, N,N-dimethyldecylamine-Noxide, from aqueous solutions onto silica and graphite. The objective is a comparison of the adsorption properties of these representative nonionic surfactants on model hydrophilic and hydrophobic surfaces. The adsorption and calorimetric data will be related to the structure of the adsorption layer as revealed by AFM images for closely related systems. (25) Rutland, M. W.; Senden, T. J. Langmuir 1993, 9, 412. (26) Grant, L. M.; Tiberg, F.; Ducker, W. A. J. Phys. Chem. B 1998, 102, 4288. (27) Patrick, H. N.; Warr, G. G.; Manne, S.; Aksay, I. A. Langmuir 1997, 13, 4349. (28) Holland, N. B.; Ruegsegger, M.; Marchant, R. E. Langmuir 1998, 14, 2790. (29) Manne, S. Prog. Colloid Polym. Sci. 1997, 103, 226. (30) Kira´ly, Z.; Findenegg, G. H. J. Phys. Chem. B 1998, 102, 1203.

Langmuir, Vol. 16, No. 23, 2000 8843

Experimental Section Materials. Controlled-pore glass 10-240 (designated CPG), a wide-pore (24 nm) silica, and Vulcan 3G (V3G), a graphitized carbon black, were taken from the same lots and pretreated in a similar way as in our previous studies.4,30 The specific surface areas, as determined by the standard BET method with N2 at 77 K, were 88 and 68 m2‚g-1 for CPG and V3G, respectively. Anomeric pure n-octyl β-D-monoglucoside (designated C8G1) was used as supplied by CalBiochem. A stock solution was prepared and diluted volumetrically to the desired concentrations with Milli-Q water. A stock solution of N,N-dimethyldecylamine-Noxide (C10DAO), 100 mM in water, was purchased from Fluka. The natural pH of the diluted solutions was in the range 7.28.5, depending on the concentration. Methods. Thermometric titrations were effected by using the titration unit of a multichannel thermal activity monitor (TAM) isothermal heat-flow microcalorimeter (ThermoMetric LKB 2277, Lund, Sweden) in a twin arrangenment.4 The titration vessel was equipped with a stirring facility and a motor-operated syringe. Typically, 2 mL of water was titrated with 0.5-1 mL of micellar solution in 10-20 µL aliquots and the associated enthalpies of dilution were recorded. The concentration of the feed solution (c1 . cmc) was chosen in such a way that, with increasing surfactant concentration in the titration cell, the cmc was reached and gradually exceeded during the experiment. The cmc and the enthalpy of micelle formation of the surfactant, ∆mich, were determined from the enthalpogram. Isotherms of the integral enthalpy of displacement at the solid/ solution interfaces were measured by the (step-by-step) cumulative method by using the flow unit of the multichannel TAM calorimeter.30 The sorption vessel was loaded with 0.1-0.2 g of adsorbent, and a constant liquid flow rate of ca. 6 mL‚h-1 was applied by using a HPLC micropump (Knauer, Berlin, Germany). The flow rate was measured with the use of a liquid microflowmeter (PhaseSep, Clwyd, England). Initially, pure water was percolated through the column until thermal equilibrium was reached. The liquid flow was then switched to that of a dilute surfactant solution in continuous operation. As the new solution entered the column, surface-bound water molecules were displaced by adsorbing surfactant molecules until the new equilibrium state was established. The flow replacement experiment was successively repeated up to the desired concentration (slightly beyond the cmc) by using small concentration increments, and the associated heat effects (calorimetric peaks) were recorded for each step. Typically, the cmc was reached in 12-14 consecutive concentration steps. The exit port of the calorimeter was connected to a differential refractometer (Knauer), and the concentration waves (break-through curves) of the solutions passing through the sorption vessel were continuously recorded. From the concentration profiles, the retention volumes and, hence, the amounts successively adsorbed were determined on the principle of flow frontal analysis solid/liquid chromatography.31,32 The step-by-step data were gathered and summed over the concentration range of interest to construct the cumulative enthalpy isotherm of displacement ∆21H vs c1 and the cumulative adsorption isotherm Γ1 vs c1. In the present step-by-step dynamic experiments, the adsorption increments ∆Γ1 and the associated enthalpy changes ∆(∆21H) were measured simultaneously for small concentration steps ∆c1, allowing a direct determination of the (pseudo) differential enthalpies of displacement ∆21h1 ) ∆(∆21H)/∆Γ1. All measurements were made at 298.15 ( (2 × 10-4) K and at the natural pH of the solutions at the respective equilibrium concentrations. Detailed descriptions of the automated titration4 and automated flow sorption experiments30 have been reported previously.

Results Titration Microcalorimetry. The cmc and the enthalpy of micelle formation ∆mich for C10DAO were determined from the thermometric titration curve as (31) Wang, H. L.; Duda, J. L.; Radke, C. J. J. Colloid Interface Sci. 1978, 66, 153. (32) Koch, C. S.; Ko¨ster, F.; Findenegg, G. H. J. Chromatogr. 1987, 406, 257.

8844

Langmuir, Vol. 16, No. 23, 2000

Figure 1. Determination of the cmc and the enthalpy of micelle formation of C10DAO in water at 298.15 K: circles, thermometric titration; solid line, sigmoidal fit; dashed line, first derivative of the sigmoidal fit.

displayed in Figure 1. In the initial flat region up to about 15 mM, micelles are broken up exothermically to give free monomers; in this region, the measured enthalpy stems from dilution of the micelles, demicellization, and dilution of the resultant monomers. In the region around the cmc, only partial demicellization takes place and the measured enthalpy changes decrease; accordingly, the enthalpy function increases sigmoidally in this region and finally levels off at about 25 mM. This new baseline is determined by the dilution enthalpy of the micelles, this process being nearly athermal; in those regions where the function runs parallel to the concentration axis, the surfactant solution behaves ideally. Physically, -∆mich consists of the heat measured in the cmc region divided by the amount of surfactant molecules belonging to destroyed micelles in this region. The titration curve was fitted with a sigmoidal expression, adopted from a previous study of Kresheck33 on the micellar behavior of decyldimethylphosphine oxide C10DPO, whose solution properties are in many respects similar to those of C10DAO. The maximum of the first derivative of the fitted curve was taken to be the cmc, and the enthalpy difference between the post- and premicellar regimes, interpolated to the cmc, was taken to represent ∆mich (Figure 1). This evaluation procedure yielded 19.7 mM and 9.6 kJ‚mol-1 for the cmc and ∆mich, respectively. Desnoyers et al. reported cmc values in the range from 14 to 21 mM and ∆mich values of ca. 13 kJ‚mol-1.34,35 However, the liquid flow calorimetric method they used provided less accurate results for C10DAO than for the lower homologue members of the alkyldimethylamine oxide series studied. The present results can further be compared with cmc ) 20 mM and ∆mich ) 11.3 kJ‚mol-1 obtained from heat of solution measurements by Benjamin.36 If his calorimetric data are converted to a reference state around the cmc (as in the present study), rather than his chosen reference state of infinite dilution (which involves some uncertainties originating from experimental difficulties), ∆mich ) 9.8 kJ‚mol-1 is obtained, in excellent agreement with the present result. (33) Kresheck, G. C. J. Colloid Interface Sci. 1997, 187, 542. (34) Desnoyers, J. E.; Roberts, D.; DeLisi, R.; Perron, G. In Solution Behaviour of Surfactants; Mittal, K. L., Fendler, E. J., Eds.; Plenum: New York, 1982; Vol. 1; p 343. (35) Desnoyers, J. E.; Caron, G.; DeLisi, R.; Roberts, D.; Roux, A.; Perron, G. J. Phys. Chem. 1983, 87, 1397. (36) Benjamin, L. J. Phys. Chem. 1964, 68, 3575.

Kira´ ly and Findenegg

Figure 2. Adsorption isotherm and integral enthalpy isotherm of displacement for the system C10DAO (1)-water (2)/silica glass CPG at 298.15 K.

For C8G1, we obtained cmc ) 27.1 mM and ∆mich ) 7.5 kJ‚mol-1. The results were discussed in detail previously in comparison with available literature data.4 Flow Sorption Measurements. Figure 2 presents the isotherm of C10DAO adsorption from aqueous solution onto CPG and the integral enthalpy isotherm for the displacement of water by C10DAO at the CPG/water interface. In adsorption-desorption experiments, the sigmoidal isotherms were found to be reversible. We may extrapolate the initial flat and the subsequent steep regions of the isotherm to define the critical surface aggregation concentration (abbreviated as csac) at the point of intersection of the two straight lines. The adsorption is weak and exothermic up to the csac of ca. 5.5 mM, which is appreciably below the cmc (viz., csac/cmc ) 0.28). The reason for this unusually low value is not clear, but it may be related to the presence of some cationic surfactant species (R-N+-(CH3)2-OH) at natural pH, adsorbing on negative surface sites of the silica glass at that pH. The extent of adsorption increases dramatically after the csac, and the associated displacement process becomes endothermic over a wide range of concentration. The surface saturation concentration of Γ1,max ) 5.36 µmol‚m-2 is reached at about 20 mM, i.e., in the close vicinity of the cmc. Figure 3 shows the adsorption isotherm and the calorimetric enthalpy isotherm of displacement for C10DAO on V3G. The shapes of these isotherms differ substantially from those observed on CPG. On the hydrophobic surface of V3G, strong and exothermic adsorption occurs even from highly dilute solutions in water. This effectively irreversible region of the isotherm is not resolved in the present study. Above a concentration of ca. 1 mM the adsorption continues to rise endothermically up to the plateau of Γ1,max ) 3.54 µmol‚m-2, reached at about the cmc. It should be emphasized that the adsorption isotherm is not of Langmuirian type. Figures 4 and 5 depict the results for the adsorption of C8G1 on CPG and V3G, respectively. Although the mechanism and the energetics of the adsorption of C8G1 resemble those of C10DAO, there are some marked differences in adsorption behavior between the two surfactants. The csac of C8G1 on CPG (ca. 20 mM) occurs at a significantly higher relative concentration; csac/cmc is ca. 0.72, as compared to ca. 0.28 for C10DAO. Accordingly, the transition region from the low-affinity adsorption domain to the high-affinity adsorption domain is narrower

Adsorption of Short-Chain Nonionic Surfactants

Langmuir, Vol. 16, No. 23, 2000 8845

Figure 3. Adsorption isotherm and integral enthalpy isotherm of displacement for the system C10DAO (1)-water (2)/ graphitized carbon black V3G at 298.15 K.

Figure 5. Adsorption isotherm and integral enthalpy isotherm of displacement for the system C8G1 (1)-water (2)/graphitized carbon black V3G at 298.15 K.

Figure 4. Adsorption isotherm and integral enthalpy isotherm of displacement for the system C8G1 (1)-water (2)/silica glass CPG at 298.15 K.

where ∆21H is the integral (or cumulative) enthalpy of displacement and Γ1s is the amount of solute per unit area actually present in the adsorption layer which displaced an amount rΓ1s of solvent from the solid/solution interface. For dilute solutions and preferential adsorption of the solute, the surface excess concentration Γ1 may safely be set equal to the real amount adsorbed, Γ1s.40 A major advantage of the experimental method used in the present work is that, at any concentration, the differential molar enthalpy data ∆21h1 are measured directly. Accordingly, the values of ∆21h1 are free from systematic errors which might arise from a combination of separately measured adsorption and calorimetric enthalpy isotherms. C10DAO on CPG and V3G. ∆21h1 data for the displacement of water by C10DAO on CPG and V3G are plotted against relative surface coverage Γ1/Γ1,max and against relative concentration c1/cmc in Figure 6. The characteristic enthalpy data are given in Table 1. On CPG, the adsorption of monomeric surfactants is connected with the enthalpy of displacement at infinite dilution, ∆21h1I(c1f0) ) ∆21h1∞ ≈ - 10 kJ‚mol-1 (Table 1), and becomes less and less exothermic as the surface concentration increases toward the csac (region I). This may be explained in terms of the presence of surface heterogeneities on the hydrophilic surface of silica. Further, the low csac/cmc of this system may be suggestive of “polar” interactions between ionized surfactant species with negatively charged surface sites. In the cooperative adsorption region (region II, csac < c1 e cmc), ∆21h1II is positive and apparently independent of surface concentration up to Γ1,max, to give 9.2 kJ‚mol-1 on average. This value is close to ∆mich ) 9.6 kJ‚mol-1, indicating that hydrophobic bonding is the thermodynamic driving force of surfactant aggregation at the silica/aqueous solution interface. The fact that ∆21h1II is constant and nearly equal to ∆mich over the entire region II indicates that interactions with the surface play no significant role here; this is quite different from the adsorption mechanism in region I. Two possible explanations emerge for the constancy of ∆21h1II: the molecules in the surface aggregates are not in contact

for C8G1 than for C10DAO. Most markedly, the plateau value of the adsorption isotherm, Γ1,max ) 0.45 µmol‚m-2 for C8G1 on CPG, is 1 order of magnitude smaller than the corresponding value for C10DAO. By way of contrast, on the hydrophobic surface of V3G, Γ1,max ) 4.63 µmol‚m-2 for C8G1, which is slightly higher than the Γ1,max of C10DAO on the same substrate. Some characteristic parameters of the four systems studied in this work are listed in Table 1. Discussion Combined material and enthalpy balances of adsorption from solution are best described in terms of the differential molar enthalpy of displacement, ∆21h1, as a function of the surface concentration Γ1s. In general, ∆21h1 is defined as the difference between the partial molar enthalpies of component 1 (surfactant) and component 2 (water) in the adsorption layer (denoted by superscript s) and the equilibrium bulk solution (denoted by superscript l):37-39

∆21h1 )

[

]

∂(∆21H) ∂Γ1s

T,p,as

) (h1s - h1l) - r(h2s - h2l) (1)

(37) Liphard, M.; Glanz, P.; Pilarski, G.; Findenegg, G. H. Prog. Colloid Polym. Sci. 1980, 67, 131. (38) Kira´ly, Z.; De´ka´ny, I. Klumpp, E.; Lewandowski, H.; Narres, H. D.; Schwuger, M. J. Langmuir 1996, 12, 423. (39) Denoyel, R.; Rouquerol, F.; Rouquerol, J. J. Colloid Interface Sci. 1990, 136, 375. (40) Kira´ly, Z.; De´ka´ny, I. Colloid Polym. Sci. 1988, 266, 663.

8846

Langmuir, Vol. 16, No. 23, 2000

Kira´ ly and Findenegg

Table 1. Adsorption and Microcalorimetric Data for the Adsorption of Nonionic Surfactants C10DAO and C8G1 from Aqueous Solutions onto Silica Glass CPG and Graphitized Carbon Black V3G at 298.15 Ka surfactant/ adsorbent

cmc (mmol‚dm-3)

∆mich (kJ‚mol-1)

csac/cmc

Γ1,max ≡ Γ1II (µmol‚m-2)

∆21Hmax (mJ‚m-2)

∆21h1∞ or ∆21h1I (kJ‚mol-1)

∆21h1II (kJ‚mol-1)

Γ1II/Γ1I

C10DAO/CPG C10DAO/V3G C8G1/CPG C8G1/V3G

19.7 ( 0.2 19.7 ( 0.2 27.1 ( 0.2 27.1 ( 0.2

9.6 ( 0.2 9.6 ( 0.2 7.5 ( 0.1 7.5 ( 0.1

0.28 (∼0.018) 0.72 (∼0.009)

5.36 3.54 0.45 4.63

48.5 -33.8 2.28 -50.7