Article pubs.acs.org/Langmuir
Amphiphile Behavior in Mixed Solvent Media I: Self-Aggregation and Ion Association of Sodium Dodecylsulfate in 1,4-Dioxane−Water and Methanol−Water Media A. Pan,† B. Naskar,† G. K. S. Prameela,‡ B. V. N. Phani Kumar,‡ A. B. Mandal,*,‡ S. C. Bhattacharya,† and S. P. Moulik*,† †
Centre for Surface Science, Department of Chemistry, Jadavpur University, Kolkata 700032, India Chemical Physics Laboratory and Chemical Laboratory, Central Leather Research Institute, Chennai-600020, India
‡
S Supporting Information *
ABSTRACT: Mixed aquo-organic solvents are used in chemical, industrial, and pharmaceutical processes along with amphiphilic materials. Their fundamental studies with reference to bulk and interfacial phenomena are thus considered to be important, but such detailed studies are limited. In this work, the interfacial adsorption of sodium dodecylsulfate (SDS, C12H25SO4−Na+) in dioxane−water (Dn−W) and methanol−water (Ml−W) media in extensive mixing ratios along with its bulk behavior have been investigated. The solvent-composition-dependent properties have been identified, and their quantifications have been attempted. The SDS micellization has been assessed in terms of different solvent parameters, and the possible formation of an ion pair and triple ion of the colloidal electrolyte, C12H25SO4−Na+ in the Dn−W medium has been correlated and quantified. In the Ml−W medium at a high volume percent of Ml, the SDS amphiphile formed special associated species instead of ion association. The formation of self-assembly and the energetics of SDS in the mixed solvent media have been determined and assessed using conductometry, calorimetry, tensiometry, viscometry, NMR, and DLS methods. The detailed study undertaken herein with respect to the behavior of SDS in the mixed aquo-organic solvent media (Dn−W and Ml−W) is a new kind of endeavor.
1. INTRODUCTION Solvent properties such as the polarity, hydro- and lipophilicity, hydrogen-bonded structure, and fluidity play significant roles in amphiphile self-organization.1 Such properties can be controlled/varied in mixed solvents with suitable combinations and compositions, making the self-association process favorable or unfavorable.2,3 This is an issue that requires appropriate planning for exploration. Such investigations are required for the optimum utilization of amphiphiles (lipids and surfactants) in industrial, medicinal, and pharmaceutical fields in relation to emulsification,4,5 stabilization,6,7 extraction,8 cosmetic formulation,9 drug encapsulation,10,11 and synthesis of nanomaterials.12,13 Cosolvents such as straight-chain alcohols (with varied carbon atoms),14,15 ethylene glycol, and glycerol generally act in dual capacities. In low proportions, the critical micelle concentration (cmc) of the surfactant decreases; in higher proportions, the cmc increases by way of increased hydrophobicity of the medium and breakdown of the water structure by cosolvent molecules.16,17 In a majority of reports, different types of cosolvents (alkanols, glycols, dimethyl sulfoxide, 1,4dioxane, acetonitrile, and amides) were used with a minimum concentration of 10 vol %.18−28 The effect of their low proportions in water on the surfactant micellization remains © 2012 American Chemical Society
under-reported. In this respect, fewer studies on the lower alkanols (methanol in particular)14a,d have been reported. There are fewest reports on 1,4-dioxane.19a Desnoyers et al.29 examined different types of alkanols (lower to higher) from the measurements of the transfer of their volumes and heat capacities from water to surfactant solutions as well as that of surfactants from water to the aqueous alkanol solutions. The alkanols were found to interact with the surfactant monomers, micelles, and water and shift the cmc. Possible premicellar association of surfactants by methanol and ethanol was suggested. In recent studies, conventional surfactants such as sodium dodecylsulfate (SDS), alkyl n-trimethyl ammonium bromides with varied chain lengths (AnTAB), sodium bis-2-ethyl hexyl sulfosuccinate (AOT), p-tert-octyl-phenoxy-polyoxyethylene (9.5) ether (TritonX-100), alkyl propoxy ethoxylate family (CiPOnEOj), and gemini surfactants have been used with attempts at empirical data correlation in terms of solvent parameters, viz, permittivity, viscosity, the Hildebrand parameter, ET(30), the Gordon parameter, the Hansen solubility Received: August 13, 2012 Revised: September 3, 2012 Published: September 5, 2012 13830
dx.doi.org/10.1021/la303281d | Langmuir 2012, 28, 13830−13843
Langmuir
Article
to 6 mL of a mixed (water−organic) solvent using a Hamilton microsyringe. The results were graphically processed by plotting the equivalent conductance (Λ) against [SDS]1/2 in accordance with the Onsager rationale to evaluate the cmc from the inflection point.31 The micellar ionization degree and hence the extent of counterion binding of the micelle (β) were determined by the slope ratio method (ratio between the post cmc the pre-cmc linear courses) amply described and discussed in the literature32 (Figure SM1). A nonlinear fitting method has been used by Carpena et al.32f to treat the specific conductance (κ) data to estimate the cmc. In this method, the accuracy of the data should be an important factor as for the Onsager relation (using Λ instead of κ) showing the error-involved fluctuations in the Λ values. Such sets of results we discarded in our data acquisition. 2.2.2. Isothermal Titration Calorimetry (ITC). An Omega ITC microcalorimeter (Microcal, USA) was used for thermometric measurements. A concentrated solution of SDS (∼15 times the cmc taken in a 325 μL microsyringe) was injected for a duration of 30s into 1.325 mL of solvent in the calorimetric cell at equal time intervals (210 s) in multiple steps (32−50 additions) under constant stirring (350 rpm).33,34 All measurements were made under thermostatted conditions maintained by a Nesleb RTE 100 circulating water bath. The heat released or absorbed at each step of dilution of surfactant solution in either water or a mixed solvent was recorded, and the enthalpy change per mole of injectant was calculated with the ITC Microcal Origin 2.9 software. The reproducibility was checked from repeat experiments. The procedure for the evolution of the cmc and the enthalpy of micellization (ΔH0m) can be found elsewhere35 (Figure SM 2). 2.2.3. Tensiometry. Tensiometric measurements were made with a calibrated du Noüy tensiometer (Krüss, Germany) by the ring detachment technique. A volume of 6 mL of a mixed solvent solution was taken in a double-wall jacketed container placed in a thermostatted water bath (accuracy, ±0.1 K) at the requisite temperature of 303 K into which a stock SDS solution of the desired concentration (∼10−15 times the cmc) was added stepwise with a Hamilton microsyringe as required (allowing ∼10 min for mixing and thermal equilibration). The detailed procedure of surface tension measurements has been reported.33,36,37 Determined surface tension (γ) values were accurate to within ±0.1 mN m−1 (Figure SM 3). 2.2.4. Viscometry. A Ubbelohde viscometer of 103.6 s average flow time for 10 mL of water at 303 K was used in the study. It was placed in a thermostatted water bath of accuracy ±0.1 K. The mixed-solvent solutions (10 mL) without and with SDS were taken in the viscometer, and the flow times were measured after thermal equilibration.38 In the latter, the change in [SDS] was made by the stepwise addition of a concentrated solution of the amphiphile with a microsyringe. Errors in measurements were within ±0.5%. Each set of measurements was duplicated, and their flow times were found to be close. The mean values were recorded and used. 2.2.5. Fluorimetry. The steady-state fluorescence measurements were taken in a Perkin-Elmer LS 55 (USA) fluorescence spectrophotometer. The aggregation numbers of SDS micelles in the corresponding solvent mixtures above the critical micelle concentration (cmc) were determined at 373 nm from the static fluorescence quenching (SFQ) of pyrene using CPC as the quencher. The concentration of the probe pyrene in the micellar medium was 0.13 μM. Excitation occurred at 332 nm, and the emission spectra were recorded in the range of 350−500 nm. The slit widths of excitation and emission were 5.0 and 2.5 nm, respectively, and the scan speed was 250 nm/min. The following equation was used for the determination of the micellar aggregation number
parameter, and the Snyder parameter. In these endeavors, solvents such as alkanols (2−7), isopropanol, ethylene glycol, 2methoxyethanol, glycerol, dimethyl sulfoxide, acetonitrile, amides (formamide, dimethylformamide, N-methyl formamide, N-methyl acetamide, and N-methyl propionamide), tetrahydrofuran, and 1,4-dioxane were used in combination with water.2b,18−28 The investigations also evaluated the cmc's (and hence the standard Gibbs free energy of micellization, ΔGm0 ) of amphiphiles, counterion binding (β for ionic micelles), micelle shape, structure, and aggregation number (Nagg), and excess amphiphile concentration at the interface (Γ). With the variation of solvent composition, the micelle-forming and nomicelle-forming zones were indentified.24,30 These results have generated information on the composition-dependent solvent behavior with respect to water toward the amphiphile selfassociation. In this work, we have made a detailed study of SDS using two nonaqueous solvents, 1,4-dioxane (Dn) and methanol (Ml) mixed with water in a large number of proportions between 0 and 100 vol %. It is anticipated that higher percentages of nonaqueous (organic) water-soluble solvents can make a dramatic difference in the amphiphile behavior. The cmc, counterion binding, energetics of the process, nature of the interaction between SDS and the solvent, ion association of Na+and DS−, micellar size, and aggregation number have been evaluated using conductometry, tensiometry, calorimetry, viscometry, DLS, and NMR methods. The results have been analyzed in terms of the energetics of the process, kinds of species formed in the media, and solvent parameters, viz, permittivity (ε), Reichardt’s parameter (ET(30)), Gordon parameter (G), and viscosity (η0). A detailed study of this kind with very low to very high volume percentages of nonaqueous solvents such as Dn and Ml was hardly considered in the past. The energetics study by the method of calorimetry (ITC) in mixed-solvent media was studied only in a limited manner.16 In a number of reported studies,2b,18−28 thermodynamic parameters were calculated, and attempts were made to correlate these with solvent composition and different solvent parameters. Because the concentration of the surfactant at high levels of nonaqueous solvent addition also makes the cmc high, its conversion to activity with respect to both concentration and solvent effects are required, which we have attempted. Nonaqueous solvents Dn and Ml are chosen for the purpose of looking into the effects of a very low polar aprotic Dn and an appreciably polar protic Ml on the solution behavior of SDS in mixed solvent media. In a low proportion they moderately decreased the cmc of SDS whereas in a high proportion the cmc was greatly increased.
2. EXPERIMENTAL SECTION 2.1. Materials. Sodium dodecylsulfate was 99% pure and an ARgrade product of SRL (India). AR-grade pyrene and cetyl pyridinium chloride (CPC) of Sigma were used for a steady-state fluorescence quenching study. Cosolvents 1,4-dioxane (Dn) and methanol (Ml) were GR- and AR-grade products of Merck (India), respectively. Doubly distilled conductivity water (specific conductance, κ = 2 to 3 μS cm−1 at 303 K) was used in the study. 2.2. Instruments and Methods. 2.2.1. Conductometry. Conductometry measurements were made with a (Eutech, Singapore) conductometer using a dip-type cell with cell constant of 1 cm−1. All measurements were made in a double-walled glass container at 303 K maintained by a Hahntech DC-2006 circulating bath with an accuracy of ±0.1 K. A concentrated surfactant solution was progressively added
⎛I ⎞ Nagg[CPC] ln⎜⎜ 0 ⎟⎟ = ⎝ IQ ⎠ [SDS] − cmc
(1)
where I0 and IQ are the fluorescence intensities of pyrene in the absence and presence of quencher CPC, respectively, [SDS] is the total concentration of SDS in solution, which was kept constant, and 13831
dx.doi.org/10.1021/la303281d | Langmuir 2012, 28, 13830−13843
Langmuir
Article
Figure 1. Conductometry, calorimetry, and tensiometry plots for SDS in Dn−W and Ml−W media at different solution compositions at 303 K. (A) Λ vs [SDS]1/2 profiles with 0, 10, 20, and 30 Dn vol %. (Inset) 60 (■) and 75 (O) Dn vol %. (B) Λ vs [SDS]1/2 profiles with 0, 10, 20, and 30 Ml vol %. (Inset) 80 (■) and 100 (○) Ml vol %. (C) ΔH vs [SDS] profiles with 0, 10, 40, and 50 Dn vol %. (Inset) 60 (Δ), 70 (■), 75 (☆), 80 (○), 85 (▽), and 88 (●) Dn vol %. (D) ΔH vs [SDS] profiles with 0, 10, 40, 50, 60, and 70 Ml vol %. (Inset) 100 (○) Ml vol %. (E) Tensiometry: γ vs log[SDS] profiles with 40, 50, 60, 70, 80, and 88 Dn vol %. (F) Tensiometry: γ vs log[SDS] profiles with 40, 50, 60, 70, 85, and 100 Ml vol %. where I0 is the peak intensity in the absence of gradient pulses and parameter k = (γnΩg)2(Δ − Ω/3) where γn is the gyromagnetic ratio and D is the translational self-diffusion coefficient. For all samples, a single-exponential decay was observed as deduced from plots of intensity versus g2 and the corresponding estimated error in D was about 3%. All data processing was done with the aid of Jeol-DeltaNMR software package. The self-diffusion coefficients of all SDS protons were used in the data analysis. Errors in the measured diffusion coefficients were within ±4%. 13 C chemical shift measurements were also made on separate samples containing SDS (97 mM)/D2O (A), SDS (97 mM)/75 vol % Dn/D2O (B), SDS (97 mM)/80 vol % Dn/D2O (C), SDS (97 mM)/ 85 vol % Dn/D2O (D), and SDS (97 mM)/88 vol % Dn/D2O (E) as well as SDS (97 mM)/70 vol % Ml/D2O (F), SDS (97 mM)/85 vol % Ml/D2O (G), and SDS (97 mM)/100 vol % CD3OD (H), which were prepared to probe the preferential solubilization site of Dn/Ml in SDS. For 13C chemical shift studies, SDS in D2O was used as an external reference in a coaxial tube. The resonance position of the methyl carbons of SDS was arbitrarily set to zero, and other resonances are expressed relative to this. 23 Na spin−lattice relaxation rate (R1 = 1/T1) measurements were made for SDS/100 vol % CD3OD (system VIII) as a function of [SDS] using inversion−recovery pulse sequence. The 23Na magnetization recovery profile was observed to be single-exponential. The R1 data were analyzed using a nonlinear three-parameter fit with an
[CPC] is the concentration of quencher in the system that was varied. The slope of the linear plots between ln(I0/IQ) and [CPC] yielded the aggregation number (Nagg) of the micelle (Figure SM 4). Errors in Nagg values were within ±5%. 2.2.6. NMR. NMR (nuclear magnetic resonance) experiments were performed (at 298 K) in an ECA-500 JEOL spectrometer operating at 500 MHz. The SDS proton self-diffusion coefficient (D) measurements were made on SDS/D2O, SDS/Dn/D2O, and SDS/Ml/D2O mixtures: SDS/D2O (I), SDS/75 vol % Dn/D2O (II), SDS/80 vol % Dn/D2O (III), SDS/85 vol % Dn/D2O (IV), and SDS/88 vol % Dn/ D2O (V) as well as SDS/70 vol % Ml/D2O (VI), SDS/85 vol % Ml/ D2O (VII), and SDS/100 vol % CD3OD (VIII), where the SDS concentration was varied but the Dn/Ml concentration was fixed. All diffusion measurements were performed with the bipolar pulse pair longitudinal encode−decode (BPPLED) pulse sequence.39 For the diffusion measurement, the experimental variable gradient duration (Ω) and diffusion time (Δ) were fixed at 6 and 100 ms, respectively, while the amplitude of the gradient (g) varied from 20 to 280 mT/m in 15 steps. For the diffusion data, in view of a reasonable signal-tonoise ratio, a line broadening of 5 Hz was applied to each freeinduction decay (FID) and Fourier transformed. The diffusion coefficients were obtained by the nonlinear fitting of the experimental data to the Stejaskal−Tanner equation39 I = I0e−kD
(2) 13832
dx.doi.org/10.1021/la303281d | Langmuir 2012, 28, 13830−13843
Langmuir
Article
Table 1. cmc and β for SDS Determined in Mixed-Solvent Media by Different Methods at Varied Concentrations of Dioxane (Dn) and Methanol (Ml) in Water at 303 Ka,b cmcave
β
solvent composition vol %(XDn, XMl)
cond
cal
st
cond
cal
st
Dn−W (Ml−W)
Dn−W (Ml−W)
0(0, 0) 1(0.002, 0.004) 2(0.004, 0.009) 3(0.006, 0.013) 4(0.008, 0.018) 6(0.013, 0.027) 8(0.018, 0.036) 10(0.023,0.046) 20(0.050, 0.097) 30(0.083, 0.156) 40(0.123, 0.223) 50(0.173, 0.301)
8.5 8.2 8.0 7.9 7.9 8.2 8.7 9.4 15 30 48 70
8.6 8.0 7.8 7.5 7.6 7.7 8.5 9.9 15 30 50 71
8.7
8.5 8.4 8.4
8.6 8.3 7.6
8.7
8.2 8.5 8.9 9.7 11 15 25 43
7.2 7.5 8.0 9.4 11 13 21 45
8.3
8.6 8.1 (8.4) 7.9 (8.0) 7.7 (--) 7.8 (7.9) 8.0 (8.0) 8.6 (8.5) 9.7 (9.4) 15 (11) 30 (14) 49 (24) 71 (46)
cmc (Dn−W)
cmc (Ml−W)
7.6
9.4 15 30 50 71
9.0 11 14 26 49
0.60 0.55 0.52 0.42 0.44 0.38 0.35 0.32 0.22 0.19 0.18 0.14
(0.58) (0.54) (---) (0.48) (0.50) (0.50) (0.49) (0.38) (0.24) (0.15) (0.10)
cmc is expressed as a millimolar concentration and the standard deviations (SD) for the cmc are ±3, ±2, and ±4% for conductometric (cond), microcalorimetric (cal), and tensiometric (ST) methods, respectively. bParameter β was obtained from the relation β = 1 − (S2/S1),40a where S1 and S2 are the premicellar and postmicellar slopes of the linear conductance vs concentration plots. (See SM1.) a
estimated error of 2%. All data processing was done with the aid of the Jeol-Delta-NMR software package. 2.2.7. DLS. Dynamic light scattering measurements were made using a Nano ZS Zetasizer (Malvern, U.K.) at a 90° scattering angle with a He−Ne laser (λ = 632.8 nm) at 303 K. All of the experimental solutions were filtered two to three times through membrane filters (with a pore size of 0.25 μm) to remove the dust particles. The mean values in the reproducible experimental results are reported. Errors in the measured hydrodynamic radius were within ±3%. 3. Results and Discussion. 3.1. Interfacial and Micelle Parameters of SDS in Aqueous Solutions with Dn and Ml of up to 50 vol %. Conductometry, calorimetry, and tensiometry measurements have witnessed the micelle formation of SDS of up to 50 vol % for both Dn and Ml in water. The representative results are shown in Figure 1A−F. The 50 vol % of the nonaqueous solvents corresponds to mole percentages of 17.3 and 30.1 for Dn and Ml, respectively (equivalent to mole fractions XDn = 0.173 and XMl = 0.301). The presence of the two nonaqueous solvents (one with low permittivity (εDn = 2.21) and the other with fair permittivity (εMl = 32.26)) produced noteworthy effects. The dependence of the cmc values on solvent compositions are presented in Table 1 (average values shown in column 4). It was found that in all of the different methods used to determine cmc, the transitions became less and less sharp with increasing presence of nonaqueous solvents (Dn and Ml). Water is more cohesive and structured by hydrogen bonding among its molecules. Less-cohesive Dn and Ml broke the water structure and formed hydrogen bonds with water molecules but failed to form a dense structure such as water. The mixed solvents had both polar and nonpolar zones, and the latter increased with increasing proportion of the solvent. The strength of interaction and hence the manifestation of a physical property such as the self-assembly of SDS became weaker and less sharp. The balance of forces (cohesive, solvophobic, entropic, and electrostatic) changed with the change in solvent composition; the resultants of these obviously manifested the experimental patterns with broad transition regions. It was observed that the cmc initially decreased up to 3 vol % Dn and 4 vol % Ml and thereafter increased as reported.41As stated above, Desnoyers et al.29 from the transfer functions of volumes and heat capacity found that methanol and ethanol do not self-aggregate in water like higher alkanols but modify the water structure and can interact with hydrophobic solutes such as surfactants, and the interaction can influence the premicellar or preaggregation states. Dn and Ml modify the water structure and lower the dielectric constant: their hydrophilic−hydrophobic interaction with SDS (according to Desnoyers et al.’s rationale) at lower concentration (under the premicellar condition) leads to the formation of micelles with slightly decreased cmc values. At higher concentration, the
cosolvent medium becomes more hydrophobic with prevalent contact among the alkyl chains with the loss of hydrophobic hydration. The medium becomes less structured to restrict the self-aggregation of SDS with increased cmc. The alkanols interact with the surfactant monomers, micelles, and water; the cmc, size, and shape of micelles are affected. Hetu et al.29b investigated these aspects with a chemical thermodynamic model using transfers of volumes and heat capacities of the alkanols to surfactant solutions. Depending on the micelle charge density, a fraction of the counterions are bound to the interface in the Stern layer. This is termed as β, which is expected to be guided by the prevailing solution environment. The β values are presented in Table 1 (evaluation in Figure SM 1). Although the cmc passed through a minimum, the β values declined throughout. The initial 60% Na+ counterion binding to the micelle decreased to 14% in 50 vol % Dn and 10% in 50 vol % Ml. Ml was more effective than Dn because XMl = 1.73XDn (at equal vol %, Ml contributed more molecules to the mixture than Dn). The continuous decrease in β with increasing nonpolarity of the medium resulted from the continuous increase in the counterion dissociation from the micelles reported in the recent literature2b,14,19,20,26,27 in mixed media of water with nonaqueous solvents of low polarity. In the presence of nonaqueous solvents, a change in the micellar geometry and/or aggregation number (Nagg) was anticipated. We have determined the aggregation number (Nagg) by the SSFQ method42,43 at varied vol % for both Dn and Ml; the results are shown in Table 5. It was found that Nagg values in the mixed-solvent media decreased with increasing proportions of both Dn and Ml, and the changes were greater with Dn than with Ml. Although the decrease in β with decreasing Nagg was reasonable, as an overall consideration the quantitative effects of the two types of cosolvents were different. A lower aggregation number produced a lower micelle surface charge density with decreased counterion association (i.e., decreased β14). Nagg in Ml−W was greater than that in Dn−W and hence βMl−W > βDn−W. Along with other parameters, we have calculated the energy parameters of micelle formation of SDS and its interfacial adsorption in varied solvent compositions. Because the concentrations of SDS used in this study were mostly not small, hence the calculations were made using activity instead of concentration. All calculations were conducted up to the maximum point of micellization (i.e., up to 50 vol % of both Dn and Ml, the respective maximum cmc's were 71 and 49 mM). The [SDS] values were