pubs.acs.org/Langmuir © 2009 American Chemical Society
Adsorption and Desorption of Cationic Surfactants onto Silica from Toluene Studied by ATR-FTIR Rico F. Tabor,† Julian Eastoe,*,† and Peter Dowding‡ †
School of Chemistry, University of Bristol, Bristol BS8 1TS, U.K., and ‡Infineum UK Ltd., Mill Hill Business Park, Abingdon, Oxfordshire OX13 6BB, U.K. Received June 24, 2009. Revised Manuscript Received August 26, 2009
The adsorption and desorption behavior of cationic dialkyldimethylammonium bromide surfactants (Di-CnDABs where n = 10, 12, 14) at the silica-toluene interface has been studied. Adsorption is a rapid process, consistent with transport control, whereas desorption appears to occur in a two-stage process, with varying proportions of surfactant desorbing in fast and slow modes. These proportions appear to be affected by trace moisture present in the adjacent toluene solvent, possibly owing to competition between surfactant and water molecules for surface sites. Surprisingly, the surfactant tail length (n) has a significant impact on solubility in toluene, and this appears to affect bulk-surface partitioning. The results are compared with previous experiments utilizing nonionic surfactants (Tabor, R. F.; Eastoe, J.; Dowding, P. Langmuir 2009, 25, 9785), and also with work on surfactant-stabilized silica in nonpolar solvents (Tabor, R. F.; Eastoe, J.; Dowding, P. J.; Grillo, I.; Heenan, R. K.; Hollamby, M. Langmuir 2008, 24, 12793). Observations are explained in terms of the balance of interactions between the surfactant, solvent, and surface.
Introduction Surfactant adsorption at solid surfaces from aqueous solutions is a well-understood process, generally and predictably controlled by the structure and morphology of the surfactant and the nature of the surface.1 When water is swapped for a low-dielectric organic solvent, the relationships between the surfactant structure, solvent, and surface change and hence so does the adsorption and desorption behavior. The principles of the hydrophobic effect and charging and aggregation behavior seen in aqueous systems no longer strictly apply when apolar solvents are employed. Despite the importance of surfactant adsorption behavior in apolar solvents for applications such as oil processing, paints and coatings, electonics and lubrication, the understanding of such systems is still underdeveloped when compared to that of aqueous analogues. The aim of this study is to begin to elucidate the nature and kinetics of surfactant adsorption and desorption in such systems and the effects of variables such as surfactant chemical structure, solvent quality, and trace moisture (which is particularly relevant to industrial systems). This information will aid in the design and selection of surfactants for lubricants, particle stabilization, and other systems of industrial significance. It is well understood that when ionic surfactants adsorb to a solid surface from water, the dissociated surfactant ion preferentially adsorbs, hence changing in the local structure of water as a result of adsorption.2 Hydrophobic effects may then favor bilayer or surface aggregate formation.3 However, in oily solvents, surfactant ion dissociation is less energetically favored and hence surfactants are believed to adsorb as ion pairs (although they may then dissociate), and the attractions driving adsorption are polar headgroup-surface interactions (and/or Lewis acid-Lewis base interactions).1 (1) Morrison, I. D.; Ross, S. Colloidal Dispersions: Suspensions, Emulsions, and Foams; Wiley: New York, 2002. (2) Somasundaran, P.; Shrotri, S.; Huang, L. Pure Appl. Chem. 1998, 70, 621– 626. (3) Atkin, R.; Craig, V. S. J.; Wanless, E. J.; Biggs, S. Adv. Colloid Interface Sci. 2003, 103, 219–304.
Langmuir 2010, 26(2), 671–677
Structure-behavior relationships for surfactants are often understood in terms of Traube’s rule: surfactants lower the surface tension to a greater degree as the carbon chain length increases.4 Freundlich extended this idea to suggest that the adsorbed amount at a solid surface from water would also increase for larger carbon numbers.5 Holmes and McKelvey studied the adsorption of fatty acids on silica from toluene to determine if the reverse trend would be seen in organic solvents (lower adsorbed amount for longer-chained acids); their research suggested that this was indeed the case.6 A similar reversal in behavior was found by Paleos for the adsorption of fatty acids onto polymer resins from toluene.7 Fowkes and Mostafa made important advances by studying the adsorption of poly(methyl methacrylate) from low dielectric solvents onto silica.8 They found that adsorption was maximized in solvents lacking Lewis acid or Lewis base characteristics (benzene and carbon tetrachloride) because solvent molecules compete less for surface sites or solvate the surfactant less effectively. For more polar solvents such as chloroform and 1,4-dioxane (which display greater Lewis acid/base properties), competition for surface sites or increased solvation was believed to reduce adsorption. The role of surface and surfactant acidity was studied by Krishnakumar et al. using the surfactants aerosol-OT (AOT) and dodecylamine on alumina and silica, adsorbed from various organic solvents.9 Their results also showed that adsorption decreased as solvent polarity increased and that desorption was possible only if the solvent had a sufficiently high dielectric constant. On the basis of these findings, a model was developed for equilibrium surfactant adsorption from organic solvents in terms of interaction parameters (derived from solubility/cohesion parameters) for solid-solvent, solute-solvent, and solid-solute.10 (4) Traube, J. J. Prakt. Chem. 1886, 34, 292–311. (5) Freundlich, H. Colloid and Capillary Chemistry; Methuen: London, 1926. (6) Holmes, H. N.; McKelvey, J. B. J. Phys. Chem. 1928, 32, 1522–1523. (7) Paleos, J. J. Colloid Interface Sci. 1969, 31, 7–18. (8) Fowkes, F. M.; Mostafa, M. A. Ind. Eng. Chem. Prod. Res. Dev. 1978, 17, 3–7. (9) Krishnakumar, S.; Somasundaran, P. Langmuir 1994, 10, 2786–2789. (10) Somasundaran, P.; Krishnakumar, S.; Mehta, S. C. J. Colloid Interface Sci. 2005, 292, 373–380.
Published on Web 09/18/2009
DOI: 10.1021/la902270e
671
Article
Tabor et al.
This study is an extension of previous work11 that covered the adsorption and desorption of standard, monodisperse CnEm nonionic surfactants onto silica from toluene. For nonionics, it was found that the ethylene oxide (EO) headgroup size significantly affected the equilibrium adsorbed amount and desorption kinetics, with larger headgroups favoring slower desorption. This is consistent with expectations based on increased surface binding as a function of ethylene oxide number. Additionally, trace levels of water in the solvent impacted the adsorbed amount, inhibiting adsorption for smaller headgroups but increasing adsorption for larger headgroups. In this work, the adsorption and desorption behavior of cationic dialkyldimethylammonium bromide surfactants (Di-CnDABs) from toluene onto silica is studied to add to the previous work11 and to begin to elucidate a general relationship between surfactant structure and adsorption/desorption, as well as the effect of solvent quality and trace moisture. The benefit of using these cationic surfactants is that moisture can be more accurately controlled and deuterated analogues allow a range of solvents to be used, providing a larger scope for probing the adsorption environment. The results may aid in understanding the factors affecting the formulation and long-term stability of nonaqueous dispersions, which has relevance to research and industrial applications of interfacial nanochemistry.
Experimental Section Materials. Surfactants from the dialkyldimethylammonium bromide series (Di-CnDABs) with chain lengths of 10, 12, and 14 carbons were used. To avoid confusion between acronyms, they are referred to in the text as Di-C10DAB, Di-C12DAB (or DDAB), and Di-C14DAB, respectively. These surfactants were obtained from Sigma and used as received. Surfactants were dried and stored over refreshed P2O5. Toluene was Fisher reagent grade and was further purified by distillation. For the moisture-controlled experiments, three water-conditioned samples of toluene were produced. Dry toluene was made by first drying the solvent over sodium wire, passing twice through an alumina column, and passing once through a copper catalyst column. Ambient toluene was left open to the air for 2 h before use. Water-saturated toluene was made by partitioning the solvent against excess water. Moisture levels were then determined by Karl Fischer titration (the average of three measurements with less than 5% deviation) as 8 ppm (dry), 102 ppm (ambient), and 1600 ppm (watersaturated). Deionized water was obtained from an Elga Purelab apparatus (resistivity 18.2 MΩ cm). Chain-deuterated DiC12DAB (D-Di-C12DAB) was synthesized and purified as described previously.12 Determination of Surfactant Solubility. Excess solid surfactant was added to a tared vial with a known volume of toluene. Samples were shaken and left to equilibrate in a thermostatted water bath overnight and were then centrifuged (6000 rpm, 10 min) to generate a pellet of undissolved material. The supernatant was carefully decanted, and the pellet was dried and weighed in situ to determine the mass of dissolved surfactant. IR Measurements. All IR spectra were obtained using a Bruker Tensor 27 Fourier-transform spectrometer at a resolution of 4 cm-1. For ATR spectra, a liquid-nitrogen-cooled MCT detector was used; for transmission measurements, a roomtemperature DTGS detector was used. Other acqusition parameters were the following: an IR aperture of 3.5 mm, a scanner velocity of 10 kHz, double-sided forward-backward acquisition, and a 14 218-point interferogram. The Fourier transform used was the Blackman-Harris three-term with a zero-filling factor (11) Tabor, R. F.; Eastoe, J.; Dowding, P. J. Langmuir 2009, 25, 9785–9791. (12) Eastoe, J.; Hetherington, K. J.; Dalton, J.; Sharpe, D.; Lu, J. R.; Steytler, D.; Heenan, R. K. J. Colloid Interface Sci. 1997, 190, 449.
672 DOI: 10.1021/la902270e
Figure 1. Representative early-time ATR-FTIR spectra for the adsorption of Di-C14DAB onto silica from ambient toluene. of 2 using Mertz phase correction with a phase resolution of 32. For ATR measurements, the accessory used was made by Pike Technologies and contained a silicon trapezoidal prismatic crystal cut at 45° (80 mm long 4 mm thick). This was held in a flow cell (volume 0.1 mL) by a Teflon O-ring. The incident angle of the IR beam was 45°, giving 10 solution-sensing reflections at the top face of the crystal. The flow cell was connected to a reservoir of solvent or surfactant solution via a microperistaltic pump (Williamson Pumps, U.K.) using Viton tubing (i.d. = 1 mm). The flow rate of the pump was set to 2.6 mL/min. For the moisture-controlled experiments, the reservoir was sealed with a rubber septum. The crystal was cleaned between experiments by removal from the flow cell, followed by wiping with cotton buds soaked in ethanol and distilled water, followed by drying in a nitrogen stream. The crystal was then remounted, and a spectrum was taken to ensure that no hydrocarbon signal was present in the 3000-2800 cm-1 region. For all ATR kinetic experiments, a fixed bulk surfactant concentration of 1.3 10-6 mol cm-3 was used. A correction factor of 6% accounting for the overlapping CH3 band of toluene was applied to H-surfactant spectra, as previously described.11 Reproducibility tests have previously shown that the adsorption of D-Di-C12DAB and H-Di-C12DAB from toluene is effectively identical to within an error of 2%.11 To perform an adsorption measurement, the clean crystal was mounted in the cell and pure toluene was flowed through for 20 min. After this time, a background spectrum was taken. Surfactant solution was then added to the solvent reservoir to achieve the desired concentration; 4-scan spectra were measured every 4 s for the first 80 s of adsorption, after which 10-scan spectra were collected every 30 s. Representative ATR spectra for adsorption are shown in Figure 1. Data acquisition was stopped when no further spectral changes were observed. For desorption measurements, the reservoir was replaced with pure toluene and spectra measured at the same time intervals. To measure the adsorption isotherm, the same experimental setup was used, but each sample was allowed to equilibrate for 20 min before spectral acquisition. For the calculation of adsorbed amounts, spectra were integrated in the alkyl stretching region of 2960-2814 cm-1 for the H surfactant and 2270 and 2054 cm-1 for the D surfactant. Transmission measurements were obtained using a liquid transmission cell (Specac, U.K.) furnished with 5-mm-thick calcium fluoride windows and a 0.05 mm Teflon spacer. The precise cell path length was calculated from an interference-fringe counting method13 as 0.051 mm. Calculation of Adsorbed Amounts. As in previous work,11 a two-phase approximation was used in the calculation of adsorbed (13) G€unzler, H.; Gremlich, H.-U. IR Spectroscopy; Wiley: Weinheim, Germany, 2002.
Langmuir 2010, 26(2), 671–677
Tabor et al.
Article Table 1. Fitted Equilibrium Adsorbed Amounts and Desorption Rate Constants for Dialkyldimethylammonium Bromides onto Silica from Ambient Toluene (102 ppm Water) surfactant
Γmax Γmax Γmax A B TOT -10
(10
-2
mol cm )
kdes A -10
(10
kdes B mol cm
-2 -1
s )
solubilitya mol L-1
Di-C10DAB 0.65 0.95 1.60 0.010 0.0002 Di-C12DAB 0.70 1.10 1.80 0.010 0.0005 0.690 ( 0.02 Di-C14DAB 0.90 1.10 2.00 0.016 0.0005 0.023 ( 0.002 a Maximum solubility in ambient toluene at 25 °C. The maximum solubility of Di-C10DAB was too high to be easily measured (>1 mol L-1).
Figure 2. Adsorption isotherms for Di-C12DAB and C12E10 from ambient toluene (102 ppm water) onto silica. The dotted line is drawn as a guide to the eye. The solid line is a fit to the initial points for Di-C12DAB using the Langmuir isotherm.
amounts. The Sperline equation gives the adsorbed amount (surface excess, Γ) of surfactant at the crystal interface:14 Γ ¼
ðA=NÞ -ECb de 1000Eð2de =dp Þ
ð1Þ
In eq 1, A is the integrated absorbance of the alkyl band in the ATR spectrum, N is the number of solution-sensing reflections, and ɛ is the molar absorption coefficient of the alkyl band (as determined from transmission measurements). Desorption Kinetics. A simple model for the two-rate desorption process observed was introduced in previous work; the justification for employing this model as well as its strengths and weaknesses have been previously covered.11 It assumes that surfactant may be adsorbed in one of two different modes, A or B, and that there is both a characteristic maximum adsorbed amount for each mode of adsorption, Γmax and Γmax A B , and a characteristic rate constant for desorption for each mode, kdes A and kdes B . The total equilibrium maximum adsorbed amount is max Γmax þ Γmax TOT = ΓA B . The change in the adsorbed amount over time for each mode can be represented as Γmax A
dθA ¼ -kdes A θA dt
ð2Þ
Γmax B
dθB ¼ -kdes B θB dt
ð3Þ
Results and Discussion Adsorption Isotherm and Adsorption Kinetics. The equilibrium adsorption isotherm for Di-C12DAB onto silica from toluene (Figure 2) shows strong affinity behavior with high levels of adsorption at low concentration (
