in aqueous solution?

0 downloads 0 Views 1MB Size Report
Mar 28, 2012 - a Dept. of Chemistry, University of Coimbra, 3004-535 Coimbra,. Portugal. E-mail: ..... 4 shows the concentration of free calcium(II) in solu- tions, i.e. ..... 41 G. C. Whitaker, Kirk–Othmer Encyclopedia of Chemical Technology,.

PCCP

Dynamic Article Links

Cite this: Phys. Chem. Chem. Phys., 2012, 14, 7517–7527

PAPER

www.rsc.org/pccp

What drives the precipitation of long-chain calcium carboxylates (soaps) in aqueous solution?w Rui F. P. Pereira,a Artur J. M. Valente,*a Mariana Fernandesab and Hugh D. Burrows*a Received 30th December 2011, Accepted 28th March 2012 DOI: 10.1039/c2cp24152h The interaction of sodium octanoate, decanoate or dodecanoate with calcium(II) in aqueous solutions has been studied using turbidity, conductivity and potentiometric measurements. These show a marked alkyl chain length dependence on the behaviour. At the calcium concentration used (1.0 mM), there is little interaction with the octanoate, the decanoate shows initially formation of a 1 : 1 complex, followed by precipitation, while the dodecanoate precipitates at low surfactant concentrations. The solid calcium carboxylates were prepared, and show lamellar, bilayer structures with planes of calcium(II) ions coordinated to carboxylate groups through bidentate chelate linkages. Thermogravimetry and elemental analyses indicate the presence of coordinated water with the calcium decanoate but not with longer chain carboxylates. The results of both the solution and solid state studies suggest that precipitation of long-chain carboxylates depends on a balance between hydration effects and hydrophobic (largely van der Waals’) interactions. Electrostatic effects make little energetic contribution but play the important structural role of ordering the carboxylate anions before precipitation. Differences are observed in the interactions between calcium(II) and long chain alkylcarboxylates and alkylsulfates, and are interpreted in terms of stronger binding to the metal of the carboxylate group. A general mechanism is proposed for calcium carboxylate precipitation from aqueous solution based on this and previous studies.

Introduction The interaction of cations with anionic surfactants in aqueous solutions is of both theoretical1–5 and practical6,7 importance, and can have dramatic effects on the mixed solution phase behaviour. The factors responsible for the formation of aggregates by amphiphilic molecules in aqueous solutions are well established,6,8,9 and there is a vast literature of experimental data which supports theoretical predictions. However, understanding of the factors that drive phase separation and precipitation from these solutions is still incomplete. Various methods have been used to study the interaction of cations with anionic surfactants, including surface tension,10–12 electrical conductivity,13–15 potentiometry,16,17 nuclear magnetic resonance 18,19 and electron spin resonance spectroscopy,20 self-diffusion measurements,21 a

Dept. of Chemistry, University of Coimbra, 3004-535 Coimbra, Portugal. E-mail: [email protected], [email protected], [email protected]; Fax: +35 1239827703; Tel: +35 1239854482 b Dept. of Chemistry and CQ-VR, University of Tra´s-os-Montes e Alto Douro, 5001-801 Vila Real, Portugal. E-mail: [email protected]; Fax: +35 1259350480; Tel: +35 1259350273 w Electronic supplementary information (ESI) available: Dependence of pH on sodium carboxylate concentration, representative X-ray diffractogram and FTIR spectrum of calcium decanoate. See DOI: 10.1039/c2cp24152h

This journal is

c

the Owner Societies 2012

ultrafiltration,22 ellipsometry,12 and time-resolved fluorescence quenching.23 Although these systems have been widely studied and a number of reviews have been reported,24–28 there is not much information linking the formation and structure of the resulting aggregates with that of the corresponding precipitates, which complicates the development of models for these systems. Over the last 100 years, a large number of scientific articles have been published on the interaction of divalent ions, such as calcium(II), with surfactants. One of the issues that has received special attention is related to the precipitation observed in these systems.29–35 Anionic surfactants tend to precipitate with cations such as Ca2+ and Mg2+ and other positively charged molecules such as cationic surfactants and polymers. Precipitation of surfactants can be desirable in some applications such as metal ion or surfactant recovery, surfactantbased separation processes, and enhanced oil recovery, but can be also detrimental in many applications including detergency, due to the loss of surfactant activity.36 The precipitated long chain carboxylates of divalent metal ions (metal soaps) are an important group of compounds, which find applications as emulsifiers, paint driers, grease thickeners, dispersant agents, etc.37–40 They have a long history. There is evidence of the use of calcium soaps for lubricating axles of chariots in ancient Egypt nearly 3500 years ago.41 Phys. Chem. Chem. Phys., 2012, 14, 7517–7527

7517

However, they are also materials of current interest and are used in solvent extraction procedures,42 as well as showing considerable potential in materials applications, such as in metal–organic mesogen systems.43,44 In this work we focus our study on the interaction of calcium cations with carboxylate anions of different even alkyl chain lengths (octanoate, decanoate and dodecanoate) in aqueous solutions, at 25 1C. We have followed a methodology similar to that used in our recent study of interactions between Al(III) and Cr(III) and carboxylate-based surfactants.45 Calcium ions are of strong interest due to their industrial, environmental and biological importance, and the carboxylates of these ions play an important role in a number of different areas, such as biomaterials,46–48 food industry,49 paper recycling industry50,51 and detergency.52,53 The precipitates formed from precipitation of calcium or magnesium salts of long chain carboxylic acids are the well-known ‘‘soap scums’’ formed around washbasins and bathtubs in areas with hard water,54 and the development of good formulations to avoid these is of considerable importance. Precipitation and redissolution of long chain calcium carboxylates also have important implications on the ingestion of dietary calcium.55 Our aim in this study is to obtain a detailed insight into the mechanism of interaction of these ions in solution using conductimetric, potentiometric and spectroscopic techniques. In addition, we have isolated and characterized the solid phase formed from calcium(II) with decanoic, dodecanoic and tetradecanoic acid. The effects of metal ion hydration on the interaction process and of hydrophobic interactions promoted by the use of different surfactant alkyl chain lengths are analysed to understand the driving force for precipitation. Experiments have also been carried out with sodium dodecyl sulfate in the presence of calcium(II) and results compared with those obtained with sodium dodecanoate. These surfactants have the same twelve carbon atom chain length, but SDS has the sulfate (R–OSO3ÿ) and sodium dodecanoate the carboxylate headgroup (R–COOÿ). These are salts of strong and weak acids, respectively. We feel that these results will contribute to a deeper understanding of the solution interactions with these systems which will be valuable in both trying to control precipitation and in the development of new applications, such as metal–organic frameworks.

Materials and methods Reagents and sample preparation Calcium nitrate tetrahydrate was purchased from Riedel-de-Hae¨n and calcium acetate from Carlos Erba. Sodium octanoate (99%), sodium decanoate (98%), sodium dodecanoate (99–100%), dodecanoic (99–100%) and tetradecanoic acid (99–100%) were purchased from Sigma. Decanoic acid (99%) was from BDH. For the sake of simplicity the surfactants will be designated as C7COONa, C9COONa and C11COONa. All the experiments with sodium carboxylates have been carried out at concentrations below their critical micelle concentration: 340 mM, 94 mM and 24 mM, respectively.56 Sodium dodecyl sulfate ( Z 98%) was purchased from Sigma. Dimethylformamide was purchased 7518

Phys. Chem. Chem. Phys., 2012, 14, 7517–7527

from Riedel-de-Hae¨n and xylene from Jose´ Vaz Pereira. These reagents were used as received, and all solutions were prepared using Millipore-Q water. No control was made on the pH, which was the natural value for each solution. Synthesis of calcium carboxylates Calcium decanoate, dodecanoate and tetradecanoate were prepared by metathesis in alcoholic solution.57 The corresponding acid was neutralised with potassium hydroxide in boiling ethanol. A solution containing the stoichiometric quantity of calcium nitrate or acetate, dissolved in the minimum volume of water, was added dropwise to the stirred hot ethanolic solution of the potassium salt of the acid. The mixture was allowed to cool slowly to room temperature with constant stirring. The precipitate was washed with distilled water and ethanol and dried in a vacuum oven at 30–40 1C. Various solvents were tested for recrystallization of the solids. The best results were obtained with xylene and dimethylformamide. Solids were first recrystallized from xylene, allowed to dry and a second recrystallization carried out from dimethylformamide. Calcium soaps do not show simple melting transitions, but instead form a variety of solid and mesophases over the temperature range from room temperature up to the decomposition temperature (around 400 1C).58 The lowest temperature phase transitions for calcium decanoate, dodecanoate and tetradecanoate were obtained by differential scanning calorimetry (DSC), giving values for the onset of the transition of 137.5, 112 and 120 1C respectively. For the latter two compounds, these are close to literature values.58 Elemental analyses were carried out for hydrogen and carbon and results are presented in Table 1. Solids were also characterized by FTIR spectroscopy, X-ray powder diffraction and thermogravimetry, as will be discussed later. Conductance measurements Electrical conductance measurements were carried out with a Wayne-Kerr model 4265 automatic LCR meter at 1 kHz, through the recording of solution electrical resistances, measured by a conductivity cell with a constant of 0.1178 cmÿ1, uncertainty 0.02%.59 The cell constant was determined from electrical resistance measurements with KCl (reagent grade, recrystallized, and dried) using the procedure and data of Barthel et al.60 Measurements were taken at 25.00 1C (0.02 1C) in a Thermo Scientific Phoenix II B5 thermostat bath. Solutions were always prepared immediately before the experiments. In a typical experiment, 20 mL of calcium(II) nitrate solution (1.0 mM) was placed in the conductivity cell; then aliquots of the surfactant solution were added at 4 minute intervals by a Gilson Pipetman micropipette. Table 1 Results for carbon and hydrogen elemental analysis for calcium carboxylates Calcium decanoate

Calcium Calcium dodecanoate tetradecanoate

C/% H/% C/%

H/% C/%

H/%

Found 58.76 9.75 64.36 10.27 66.98 Expected for Ca(O2CR)2 62.52 10.01 65.70 10.57 67.97 Expected for 59.99 10.06 63.12 10.59 65.58 Ca(O2CR)2H2O

10.75 10.99 11.00

This journal is

c

the Owner Societies 2012

The specific conductance of the solution was measured after each addition and corresponds to the average of three ionic conductances (uncertainty less than 0.2%), determined using homemade software. The specific electrical conductance of the solutions, k, is calculated from the experimental specific conductance, kexp, and corrected for the specific conductance of water, k0: k = kexp ÿ k0. Potentiometric measurements Potentiometric measurements were carried out with a pH Radiometer PHM 240. The ion concentration present in solution was measured using a calcium-selective polymer membrane electrode and an Ag/AgCl reference electrode both from Metrohm. Before each set of measurements, calibration was carried out using calcium-containing standard solutions. pH measurements were made with a pH conjugated electrode (Ingold U457-K7); the pH was measured on fresh solutions, and the electrode was calibrated immediately before each experimental set of solutions using IUPAC-recommended pH 4 and 7 buffers. In a typical experiment, using a Gilson Pipetman micropipette, aliquots of the surfactant solution were added to 20 mL of divalent metal solution. The electrode potential was recorded after signal stabilisation, and all measurements were carried out at 25.00 1C (0.02 1C). Turbidity measurements Turbidity measurements were made by measuring the optical transmittance at 550 nm with a Shimadzu UV-2450 spectrophotometer. In a typical experiment, different amounts of surfactant were added to a 1.0 mM calcium nitrate solution with continuous stirring for about five minutes before each measurement. Infrared spectroscopy Infrared spectra of calcium carboxylates were run as KBr pellets on Perkin Elmer 1760 FTIR and Philips 9800 FTIR spectrometers over the wavenumber range 600–4000 cmÿ1 using 32 scans with a resolution of 4 cmÿ1.

Results and discussion Turbidity measurements The presence of calcium metal ions (Ca2+) in aqueous solutions of long chain sodium alkylcarboxylates (CnCOONa) normally leads to precipitation, due to formation of the calcium(II) carboxylates (metal soaps). As has been discussed elsewhere,39,61 the stoichiometry of these systems depends on both the nature of the metal ion and the precipitation conditions (pH, temperature, etc.). The focus of this study is on understanding the interaction mechanism between calcium ions and sodium carboxylates, and turbidity (optical transmittance) measurements were made to see how carboxylate chain length affects precipitation with calcium ions. Fig. 1 shows the effect of adding sodium carboxylates (CnH2n+1COONa, which for simplicity we will designate as CnCOONa) on the transmittance of aqueous 1.0 mM Ca2+ solutions. When C7COONa is added to 1.0 mM calcium aqueous solutions no precipitation or change in turbidity is observed, with the transmittance staying constant at 100% over the whole molar ratio range. Calcium soaps show a small, but measurable, solubility in water, which increases with decreasing chain length62 and with introduction of unsaturation into the alkyl chain,63 and at this calcium(II) concentration, the interaction between these two salts does not lead to the formation of insoluble products. Upon increasing the surfactant chain length to ten, no turbidity changes are observed until a molar ratio, r, around one. Above r > 1, the transmittance starts to decrease, reaches minimum values for r around 3, and then remains constant with increasing sodium decanoate concentration. Visual observations confirm that interaction of calcium ions with decanoate anions only leads to the formation of insoluble complexes for r > 1. It seems that there is a critical minimum concentration of surfactant that is needed to initiate the interaction. Upon addition of the twelve carbon carboxylate, C11COONa, to the aqueous calcium solution, the transmittance immediately starts to decrease, reaching minimum values when r = 2, and then remains constant for higher molar ratios.

X-ray diffraction X-ray powder diffraction studies were performed using a Philips difractometer equipped with a PW-1730 generator and a PW-1840 goniometer, using monochromatic Cu-Ka (l = 1.54 A˚) radiation. Estimated errors are normally 0.02 A˚ but are slightly higher in the small angle region. Thermal analysis Thermogravimetric analysis (TG) was carried out over the temperature range from room temperature to 800 1C using a Polymer Laboratories PL STA-1500 TGA thermobalance. For DSC studies a Perkin Elmer DSC-7 calorimeter was used, with a sensitivity 0.001 1C and 0.001 mW. Finely divided powder samples were sealed in perforated 25 mL aluminium pans and run at heating rates of 2.5 1C minÿ1. Thermograms were recorded in a nitrogen atmosphere using empty aluminium pans as reference. This journal is

c

the Owner Societies 2012

Fig. 1 Effect of the addition of sodium octanoate (&, n = 7), sodium decanoate (J, n = 9) and sodium dodecanoate (D, n = 11) on aqueous 1.0 mM Ca2+ solution transmittance, at 25 1C.

Phys. Chem. Chem. Phys., 2012, 14, 7517–7527

7519

These turbidity measurements clearly demonstrate the importance of surfactant hydrophobicity on the Ca2+:CnCOOÿ interaction. No precipitation is observed with the shortest chain system, and in the other two cases, the interaction is strongest for the system where n = 11. It is worth noting that dissociation constant (K) measurements on calcium(II):acetate complexation only show a very low degree of binding,64,65 and although K increases slightly upon increasing alkyl chain length due to short range inductive effects, electrostatic interactions between Ca2+ and carboxylate can only play a very minor energetic role in the initial binding. Conductivity and pH measurements From the turbidity analysis, no significant interaction occurs between calcium(II) and sodium octanoate, while the behaviour with calcium(II) and sodium decanoate or dodecanoate is significantly different. Electrical conductivity provides a valuable tool to assess the effect of ionic 15,66 or non-ionic surfactants67 on the structure of ionic solutions, since it allows one to follow modifications of complex electrolyte solutions, resulting from changes in the size and shape of moving particles and/or effective particle changes.68 Electrical conductance measurements were carried out to evaluate the overall effect of addition of sodium alkyl carboxylates to 1.0 mM aqueous solutions of calcium(II) nitrate (Fig. 2). In Fig. 2 it is possible to distinguish three different types of behaviour of k as a function of the surfactant concentration, depending on the alkyl chain length. With sodium octanoate there is a progressive increase in electrical conductivity of the resulting solution with no discontinuities, which indicates the absence of any significant calcium(II)–carboxylate interaction within the sensitivity of the technique. These results are in complete agreement with the turbidity data. With sodium decanoate, the electrical conductivity behaviour is very similar to what has previously been seen in solutions of sodium dodecyl sulfate and trivalent ions.69 Addition of decanoate to a solution containing Ca2+ leads to an initial increase of solution electrical conductivity since there is little or no interaction

Fig. 2 Effect of Ca2+ (1.0 mM) on the specific conductivity of aqueous solutions of sodium octanoate (&, n = 7), sodium decanoate (J, n = 9) and sodium dodecanoate (D, n = 11), at 25 1C.

7520

Phys. Chem. Chem. Phys., 2012, 14, 7517–7527

between the species in this concentration region; however, for a molar ratio (r) [C9COONa]/[Ca2+] of 1.11  0.02, the specific conductance tends to stabilize, this can be interpreted either as a critical aggregation concentration or ion pair formation. Above this point, the Ca2+ species interact with the decanoate, until the stoichiometric molar ratio of 2.3  0.4. For higher molar ratios, the specific conductance increases with increasing concentration of carboxylate. This behaviour can readily be understood, since after all the Ca2+ has been ‘‘consumed’’ by the carboxylate, with formation of the precipitate seen in turbidity measurements, excess surfactant produces an increase in the solution conductivity, similar to what occurs in the absence of the divalent ions. With sodium dodecanoate, interaction with Ca2+ occurs from the first addition of carboxylate to the divalent ion solution, and leads to a constant conductivity value until a molar ratio of 2.15  0.02 is reached, above which the conductivity increases. The conductimetric data are in agreement with turbidity results, and suggest precipitation of the 1 : 2 calcium dodecanoate species, and after this is complete there is an increase in conductivity due to excess sodium carboxylate. From these results we can conclude that the calcium:carboxylate interaction is only weakly affected by ionic interactions, and that a precursor to this involves the hydrophobic chain of the amphiphilic ion. The increase in alkyl chain length increases the Ca2+:CnCOOÿ interaction. Two distinct interpretations are possible. One comes from the structuring effect that the increase in hydrophobic chain length has on solvent water molecules, that is, the metal will be more available for the surfactant since the water–water binding energy increases.70 A second explanation comes from thermodynamic analysis of the factors stabilizing the solid phase in metal soaps. Detailed theoretical and experimental studies on long chain lead(II) carboxylates indicate that the dominant contribution to the melting enthalpy comes from van der Waals’ interactions between the alkyl chains, which increase with chain length.71 It is likely that similar factors are involved in the case of solubility, and that interchain van der Waals’ interactions play a major role in bringing the carboxylate chains together as a precursor to precipitation. In the analysis of these systems, we need to consider the possible influence of hydrolysis on the interaction between the carboxylate and divalent metal ions.72 Fig. 3 shows the variation of pH in a titration of a 1.0 mM solution of calcium ion with sodium carboxylates. It is interesting to note that in the case where octanoate and decanoate are added to Ca2+ solutions there is a pH increase, which is similar to what occurs in the absence of divalent ions (Fig. SI1, ESIw), and that the hydrolysis of alkanoates increases with increasing alkyl chain length. This behaviour was expected for the system with octanoate from the observation that turbidity and conductimetric measurements show there is no significant interaction between the salts. With sodium decanoate, the biggest change occurs at carboxylate concentrations below the region for formation of a 1 : 1 species, and although the pH continues to increase upon adding carboxylate, this is probably being mediated by the Ca2+:decanoate precipitation. Marked differences in the effect on pH are observed between the effects of addition of decanoate and dodecanoate. This journal is

c

the Owner Societies 2012

Fig. 3 Effect of the addition of sodium alkyl carboxylates on the pH of 1.0 mM aqueous Ca(NO3)2 solutions, at 25 1C; (&) sodium octanoate, (J) sodium decanoate and (D) sodium dodecanoate.

The addition of C11COONa to solutions of Ca2+ initially leads to a pH decrease until r = 0.5. Similar behavior has previously been observed on the interaction of aluminium(III) with sodium dodecylsulfate in aqueous solution,69 and is accompanied by changes in the line width of the 27Al NMR spectrum, suggesting changes in the coordination sphere. The pH decrease cannot be justified in terms of hydrolysis of Ca(II), since its first hydrolysis constant, pK, is equal to 12.6.72 Instead, we can feel that the pH behavior comes from the equilibria involving dimers and oligomers of dodecanoate following interaction with metal ions, leading to the consequent partitioning of hydronium ions.31,45,69 Turbidity measurements show that this region (r o 0.5) corresponds to the onset of precipitation, and studies on alkali metal carboxylates reveal that precipitation is likely to involve formation of neutral soap, acid soap and alkanoic acid crystallites.31 Calcium laurate can also form an acid soap complex with lauric acid,55 and similar complex equilibria may explain the observed pH changes up to the electroneutrality ratio r = 2. Above this, the pH increase is caused by the presence of free sodium laurate.

Fig. 4 Effect of addition of sodium decanoate (J) and sodium dodecanoate (D) on the concentration of free Ca2+ (in equilibrium), at 25 1C. Solid lines are obtained by fitting eqn (7) and (8) to the experimental data.

Assuming that the complexation occurs in a two-step mechanism, two equilibrium equations can be written: K1

þ Ca2þ þ Cn COOÿ ÿ! ÿ CaðCn COOÞ K2

CaðCn COOÞþ þ Cn COOÿ ÿ! ÿ CaðCn COOÞ2

This journal is

c

the Owner Societies 2012

½CaðCn COOÞþ Š ½Ca2þ Š½Cn COOÿ Š

ð3Þ

½CaðCn COOÞ2 Š ½CaðCn COOÞþ Š½Cn COOÿ Š

ð4Þ

K1 ¼

K2 ¼

Conservation of mass gives [Ca2+]free = [Ca2+]total ÿ [Ca(CnCOO)+] ÿ [Ca(CnCOO)2] (5) [CnCOOÿ]free = [CnCOOÿ]total ÿ [Ca(CnCOO)+] ÿ 2  [Ca(CnCOO)2]

All the above results indicate that in the systems Ca :C9COONa and Ca2+:C11COONa, the addition of these carboxylates to a solution containing calcium ions leads to precipitation, with the consequent removal of the metal ions from aqueous solution. The fraction of calcium ions remaining in solution can be quantified by using a calcium(II) selective electrode. Fig. 4 shows the concentration of free calcium(II) in solutions, i.e. Ca2+ that did not precipitate in the presence of the surfactant. As with the turbidity data, when C9COONa is added to the calcium salt solution, the concentration of calcium only decreases significantly for r > 1.0, whereas with C11COONa the decrease is abrupt and is seen on the first addition of surfactant. With the decanoate, it seems that there is initially formation of a 1 : 1 complex (or ion pair) whereas with the dodecanoate, the stoichiometric 1 : 2 (calcium(II):carboxylate) solid rapidly precipitates.

ð2Þ

The stabilities of the complexes Ca(CnCOO)+, Ca(CnCOO)2 can be described in terms of the association constants, K1 and K2:

Concentration of free Ca2+ 2+

ð1Þ

(6)

Combination of eqn (3)–(6) gives, after some algebraic manipulations, the following expressions: 2K2K1[Ca2+][CnCOOÿ]2 + (1 + K1[Ca2+])[CnCOOÿ] ÿ [CnCOOÿ]total = 0

(7)

and ½Ca2þ Š ¼

½Ca2þ Štotal 1 þ K1 ½Cn COOÿ Š þ K2 K1 ½Cn COOÿ Š2

ð8Þ

The [CnCOOÿ]free can be estimated through an analytical solution of the real solution of a second-degree equation, by using the quadratic formula. Association constants, K1 and K2, have been obtained from a least-square fit of eqn (7) and (8) to the selective calcium(II) electrode experimental data (Fig. 4). The calculated association constants correspond to Phys. Chem. Chem. Phys., 2012, 14, 7517–7527

7521

those where the sum of squares of the residuals between calculated and observed values of free calcium concentration is minimized. For the system Ca2+:C9COONa, K1 and K2 are equal to 3.2 Mÿ1 and 2.5  105 Mÿ1, respectively, with a residual of 5.5  10ÿ8 M2. The fitting was only done for r > 1, in order to focus on the effective interaction zone based on the above presented data. The global association constant for this system is 8.1  105 Mÿ2. It is worth noting that reported values for the association constant for the 1 : 1 calcium(II) acetate complex fall in the range 3.0–6.3 Mÿ1,65 similar to the K1 value seen here. Whilst the good agreement is possibly fortuitous, it does suggest that we are looking at similar complexation equilibria of carboxylate by the metal ion. For the system Ca2+:C11COONa, K1 and K2 are equal to 4.8  103 Mÿ1 and 8.5  103 Mÿ1, respectively, with a residual of 7.0  10ÿ10 M2. The global association constant for this system is 4.1  107 Mÿ2. This analysis greatly oversimplifies the system since it ignores the effect of precipitation on the two equilibria. However, it is at least qualitatively useful, since the largest association constant is found for the system Ca2+:C11COONa (two orders of magnitude higher) which is in agreement with a stronger interaction in this case. It is also interesting to note that K1 is lower than K2 for both systems but, it is significantly lower in the Ca2+:C9COONa system, where the complex Ca(C9COO)+ does not initially precipitate. With the system Ca2+:C11COONa, K1 and K2 are very similar, indicating that as soon as one carboxylate is bound, it helps complexation of the second one, possibly through some cooperative interactions resulting from the subsequent precipitation. Studies of precipitation of calcium(II) carboxylates show that under these conditions this is a homogeneously nucleated process,29,73 and it is likely that dimerization and oligomerization of the neutral Ca(O2CR)2 species to form the precrystalline nucleus becomes energetically more favourable for the longer chain homologues. The above results suggest a dominant role of hydrophobic effects, particularly interchain van der Waals’ interactions, in the aggregation and precipitation. Although electrostatic interactions between the oppositely charged calcium(II) and carboxylate chains seem to play a secondary role, they act as chaperones and are essential for bringing the long chain carboxylates together. Studies on calcium(II) binding to long chain carboxylate monolayers at the air–water interface using vibrational sum frequency generation spectroscopy provide more information on this binding, and show that both bridging and bidentate chelating configurations are possible.74 They also show that binding has effects on the interfacial water structure.75 This is supported by molecular dynamics simulations on related systems.76 Ionic hydration is likely to be one of the factors favouring dissolution of the shorter chain length calcium(II) carboxylates. Increasing evidence is suggesting that water loss may be a driving force in binding metal ions to surfactants or polyelectrolytes,77 such that precipitation or dissolution of the long chain calcium carboxylates in water would seem to be driven by a balance between hydrophobic and hydration interactions. It is known that under certain conditions, calcium soaps precipitate as their hydrates.78 7522

Phys. Chem. Chem. Phys., 2012, 14, 7517–7527

We have therefore synthesized calcium soaps of different chain lengths to test if they contain bound water molecules. Isolation and analysis of solid calcium(II) carboxylates The synthesis of calcium decanoate, dodecanoate and tetradecanoate was carried out as described in the experimental section. Longer chain homologues were chosen for these experiments than for the solution measurements because of ease of precipitation and recrystallization. DSC studies suggest that the carboxylates prepared with calcium acetate are slightly purer than those prepared with calcium nitrate. The calcium soaps were recrystallized from xylene and then from dimethylformamide (DMF). This is expected to remove any adsorbed water. Infrared spectra showed the absence of free acid. However, elemental analysis for carbon (Table 1) showed rather lower values from those expected for the anhydrous carboxylates, and with the decanoate are closer to those expected for the monohydrate. The calcium(II) carboxylates were also studied by thermogravimetry (TG). Earlier work on long chain calcium(II) carboxylates79,80 have shown that these initially start to break down above 350 1C with weight losses corresponding to the formation of ketones, leaving calcium carbonate as a solid residue. The calcium carbonate decomposes at higher temperatures with loss of carbon dioxide leaving calcium oxide. These two steps were observed in the thermograms of our recrystallised samples of the three calcium soaps around 500 and 700 1C, respectively (Fig. 5). However, with the decanoate there was an earlier sharp weight loss around 100 1C, which we attribute to loss of a coordinated water molecule. In contrast, with the dodecanoate and tetradecanoate, there were no well defined water loss steps in this region. Very similar behaviour has previously been observed with long chain cerium(III) carboxylates, where the octanoate and decanoate show a sharp weight loss in the 80–110 1C region, while the longer chain homologues only show a gradual weight loss in this region which is much smaller than that expected for the monohydrate.81 We believe that with the lower chain length compounds, the metal ions are present with some coordinated water, and that this must reflect some balance between hydration effects

Fig. 5 TG curves of: calcium decanoate (black), calcium dodecanoate (red) and calcium tetradecanoate (blue).

This journal is

c

the Owner Societies 2012

(favoured by short chain soaps) and hydrophobic interactions (which are favoured with the long chain members). Samples were also analysed by X-ray powder diffraction. In agreement with previous studies,58,80,82,83 the X-ray diffractograms of the solid soaps show a series of sharp peaks for the low angle region, up to at least the fifth order reflection with the corresponding Bragg d spacing values in the ratio 1 : 1/2 : 1/3. . .1/n. A typical diffractogram is shown in the ESIw (Fig. SI2). As with the earlier studies on X-ray diffraction of calcium carboxylates, these reflections are attributed to the long spacings from calcium ion containing planes separated by the alkylcarboxylates, with the hydrocarbon chains in an all-trans conformation orthogonal to the basal planes. The d values correspond to the successive (00l) reflections through these planes of metal atoms in the lamellar bilayer structure and are proportional to twice the length of the carboxylate molecule. Average bilayer spacings were determined from the diffractograms for the three calcium soaps and are presented in Table 2. There are also a number of reflections in the 20–251 region, which are associated with shorter distances, probably corresponding to interchain separations. For metal carboxylate having two chains in the fully extended conformation, the predicted lamellar spacings (dmax) are given by:81 dmax = 2dC–H + 2(n ÿ 1)dC–C sin 551 + 2dC–O + 2rCa2+ (9) Taking dC–H = 1.09 A˚, dC–C = 1.54 A˚, dC–O = 1.36 A˚, rCa2+ = 1.00 A˚,84 values of the long spacings were calculated for the three calcium carboxylates and are compared with the experimental values in Table 2. In general, there is a good agreement, confirming the lamellar bilayer arrangement. Infrared spectra of long chain carboxylates can provide information on metal ion binding, chain conformation and chain packing.39 IR spectra were run to obtain data for the three solid calcium carboxylates as KBr pellets. A typical spectrum is shown in ESIw (Fig. SI3), and the main bands are presented in Table 3, with vibrational assignments based on results on related systems.81 Data are not shown in the table for the broad absorptions seen in the 3000–3500 cmÿ1 region, where the water OH stretching Table 2 Comparison between the experimental X-ray diffraction data for the long spacing and values calculated for lamellar spacing

Experimental dmax/A˚ Calculated dmax/A˚

CaC10

CaC12

CaC14

30.79 29.61

35.56 34.65

40.38 39.70

Table 3 Frequency (in cmÿ1) of the main bands observed in the infrared spectra of calcium(II) carboxylates and their vibrational assignment CaC10

CaC12

CaC14

Vibrational assignmenta

2919 2849 1539 1472 1435 1379 721 671

2921 2851 1541 1464 1439 1352 719 698

2921 2851 1541 1462 1437 1377 719 698

uas (CH2) us (CH2) uas (COO) d (CH2)def us (COO) ds (CH3) r (CH2) d (COO)

a

u, stretching; d, bending; r, rocking; as, antisymmetric; s, symmetric.

This journal is

c

the Owner Societies 2012

vibrations are observed, because some of this may be due to water in the host material. The dodecanoate and tetradecanoate show single broad absorptions, centred around 3350 cmÿ1 due to adsorbed water (either from the sample or the KBr pellet). In contrast, calcium decanoate shows both this absorption and a band of medium intensity around 3550 cmÿ1. As with previous results for cerium(III) carboxylates,81 we believe that this is due to the presence of a coordinated water molecule. The data in Table 3 for the methylene group vibrations are consistent with the model from X-ray powder diffraction of the carboxylate chains being present in an extended all-trans conformation. The most important result concerns the antisymmetric and symmetric carboxylate stretching vibrations. The frequencies and relative positions of these two bands are related to the type of metal carboxylate coordination.85,86 The two most likely binding modes for calcium with the carboxylate anion are chelating bidentate and bridging bidentate. From the observed difference between the asymmetric and symmetric stretching frequencies for the three calcium soaps (Du = 103 cmÿ1) and comparison with literature data (Duchel E100 cmÿ1, Dubr E 150–170 cmÿ1), we believe that in the solid phase of these calcium soaps the carboxylate anion binds the metal ion through a chelating bidentate structure. Studies of calcium binding to palmitate monolayers using vibrational sum-frequency generation spectroscopy provide evidence for both bridging and bidentate chelating configuration, depending on the concentration.74 It seems probable that on precipitation of calcium soaps, they may initially associate through bridging structures before precipitating in the chelated form. Surfactant acidity effect on calcium(II):surfactant interactions It is well established that the nature, in particular the acidity, of headgroups of anionic surfactants can play an important role in their interactions with divalent metals. Since sodium carboxylates are salts of weak acids,87 the behaviour will be compared for one chain length (a 12 carbon chain) with those obtained with a surfactant derived from a strong acid, to shed light on the importance of the surfactant acid–base character on calcium(II):surfactant interactions. In this context, it is worth remembering that the salt of the weaker acid is likely to act as a stronger Lewis base upon coordination. The comparison will be based on conductimetric, potentiometric and turbidity measurements upon addition of sodium dodecanoate and sodium dodecyl sulfate to aqueous calcium nitrate (1.0 mM) solutions. The addition of sodium dodecyl sulfate to the calcium ion solution leads to drastic changes in the electrical conductivity, in particular in the low surfactant concentration (pre-micellar) region (Fig. 6). The conductimetric profile is very similar to that observed with aqueous solutions of trivalent ions and SDS.14,69 By subtracting the effect of the SDS concentration on the electrical conductivity in the absence of metal ions it is possible to define linear behavior, with two different slopes at the beginning and the end of the concentration range studied. The transition point, attributed to the critical micelle concentration (cmc), indicates the concentration at which a pseudophase transition takes place between unimers and micelles.15 However, in the presence of Ca2+, the addition of SDS Phys. Chem. Chem. Phys., 2012, 14, 7517–7527

7523

Fig. 6 Effect of SDS on the specific conductance of aqueous calcium nitrate (1.0 mM) solution, at 25 1C. Solid lines show transition points corresponding to critical aggregation concentration (cac), maximum interaction concentration (mic) and apparent critical micelle concentration (cmcap).

leads to several modifications in the solution structure, since two further transition points can be detected (lines of Fig. 6). The first corresponds to the SDS concentration at which a strong interaction between Ca2+ and dodecyl sulfate occurs; after this point, insoluble complexes are formed in solution and there is a slight decrease followed by constant values of the electrical conductivity upon addition of SDS up to considerably higher surfactant concentrations. Such solution electrical behaviour can be justified by the formation of large charged species and/or by the charge collapse of ionic species. Both explanations are in agreement with the formation of aggregates involving Ca2+ and dodecyl sulfate (DSÿ) and, consequently, the concentration at which aggregation takes place is called the critical aggregation concentration (cac) and has a value of 1.27  0.02 mM. Further evidence to support the onset of interaction between Ca2+ and DSÿ comes from the observation of a second transition point; at SDS concentrations above this, the shape of the plot of k versus [SDS] is similar to that found in pure aqueous solutions, suggesting that no more Ca2+ are available to induce the formation of Ca2+/DSÿ aggregates and, consequently, SDS will be in excess, and will behave in a similar way to that occurring in the absence of divalent salt. The SDS concentration at which all Ca2+ is consumed can be defined by the maximum interaction concentration (mic), and has a value of 2.71  0.09 mM. Although we are only interested in the Ca2+:surfactant interactions in the premicellar zone, the third transition point in Fig. 6 corresponds to the onset of SDS micellization. Data presented for this region will not be compared with sodium dodecanoate, since in that case we could not study such high surfactant concentrations due to irreversible formation of precipitate. At first glance, the micellization of SDS in the presence of Ca2+ appears to occur at SDS concentrations above the normal surfactant cmc (8.3 mM),69 which cannot be thermodynamically justified. However, because of the initial complexation of dodecyl sulfate ions with Ca2+, the critical micelle concentration of the SDS, in the presence of trivalent ions, cmc0 , should be calculated by: cmc 0 = cmcap ÿ mic = 6.6  0.1 mM. 7524

Phys. Chem. Chem. Phys., 2012, 14, 7517–7527

This is close to the value 5.75 mM previously reported for the cmc of SDS in the presence of calcium(II).29 This is similar to the behaviour seen with SDS in the presence of gelatine,88 where complexation lowers the cmc of SDS. The fact that the conductimetric behaviour observed for the systems Ca2+:C11COOÿ and Ca2+:DSÿ is completely different supports the existence of different mechanisms of calcium(II):surfactant interactions. With C11COONa, interaction occurs immediately after the addition of surfactant, whereas with SDS this only occurs after the cac, showing that the binding capacity of dodecanoate to calcium ions is greater than that of SDS. If the anions are binding directly to the Ca2+ cation, the degree of complexation will depend on the Lewis basicity of the anion. DFT calculations on sodium and lithium complexes show that the carboxylate group acts as a stronger Lewis base than sulfate,89 which is consistent both with its stronger binding of dodecanoate to calcium(II) and its weaker proton acidity, compared with dodecylsulfate. Further information on difference between systems with sodium dodecanoate and SDS comes from turbidity measurements (Fig. 7). The presence of a critical aggregation concentration, as discussed above, is also evident in optical studies on the Ca2+:SDS system, where the formation of precipitate only occurs for concentrations of SDS greater than 1.28 mM. Upon increasing the SDS concentration still further, a drastic decrease in the transmittance of the mixed solution is observed, due to formation of precipitate. The transmittance continues to decrease until the maximum interaction concentration is reached. The behavior up to this region is very similar for the dodecanoate and SDS. However, the addition of further SDS around and above the cmc leads to redissolution of the precipitate.90 For concentrations of SDS above 18.4 mM, the mixed Ca2+:SDS solutions are completely free of precipitate and the transmittance returns to 100%, suggesting SDS micelles help solubilize the aggregates of Ca2+/DSÿ , probably through hydrophobic interactions. In contrast with the behavior of calcium dodecylsulfate, no dissolution of the precipitate formed with Ca2+:C11COONa, or the other carboxylates, was observed upon adding excess surfactant,

Fig. 7 Effect of the addition of sodium dodecyl sulfate on the turbidity of a 1 mM Ca2+ aqueous solution, at 25 1C. Dashed line represents the cmc of SDS.

This journal is

c

the Owner Societies 2012

Conclusions

Fig. 8 Effect of the addition of sodium dodecyl sulfate on the pH of 1.0 mM Ca(NO3)2 solutions, at 25 1C.

indicating either stronger interactions within the precipitate or weaker interactions involving the alkyl chains of the carboxylates. This important difference between systems with SDS and dodecanoate provides further evidence that interactions involving Ca2+ are stronger with carboxylate than with sulfate, while the greater redissolution capacity of SDS may be linked to its stronger acidity. Potentiometric studies were also made on the system, and in Fig. 8 we can see that addition of SDS to calcium(II) (1.0 mM) solution leads to a continuous increase of pH of the mixed solution, which is only slightly affected when the SDS critical aggregation concentration is reached. In contrast, with pH measurements for the dodecanoate systems, addition of surfactant initially led to a decrease of solution pH up to the zone where Ca2+:C11COONa interaction occurs. These results again support different mechanisms for interaction of calcium ions with dodecanoate and SDS. Unfortunately, although it would be valuable to study potentiometrically how calcium concentration varies upon addition of SDS to compare with the dodecanoate system, this was not possible because SDS interferes with calcium concentration measurements using the calcium ion selective electrode.

Both solution measurements and studies on isolated solids have been used to obtain an insight into what drives the precipitation of long chain calcium carboxylates from aqueous solution. Studies using turbidity, electrical conductivity, pH and calcium selective electrodes in aqueous solutions show a marked dependence upon the carboxylate chain length. At the calcium ion concentration used (1.0 mM), no precipitate was formed upon addition of sodium octanoate, while with the decanoate and dodecanoate, precipitation became more pronounced upon increasing chain length. With the isolated calcium(II) carboxylates, these precipitated as lamellar solids with the carboxylate bound to calcium through a bidentate chelating structure and, in addition, with the shortest chain homologue there was evidence for coordinated water. The results strongly suggest that whether the calcium soap precipitates or stays in solution involves a careful balance between hydrophobic factors (probably mainly involving interchain van der Waals’ interactions) and hydration effects. Increasing alkyl chain length has the effect of squeezing out the water from the solid structure. In water, calcium(II) is present with six water molecules in its primary hydration sphere and a number of more weakly bound molecules at greater distances.91 Molecular dynamics simulations show that hydration effects are also important with long-chain alkylcarboxylate headgroups.76 From studies on the interaction of calcium(II) with sodium decanoate in water, complexation occurs in at least two distinct stages, initial ion association to form a 1 : 1 complex, in which the carboxylate probably displaces two water molecules, followed by binding of the second carboxylate, which is a precursor to precipitation. The electrostatic interactions in water are weak, as seen by the small dissociation constant, but undoubtedly have an important role in acting as a template to bring the long chain carboxylates together. Upon formation of the long-chain Ca(O2CR)2 species, these can then form dimers, trimers and other oligomers, driven by hydrophobic interactions. The oligomers are likely to act as nuclei for formation of crystalline solids. A detailed study of the precipitation of calcium dodecanoate has been presented by Clarke et al.,29 and a schematic idea of the precipitation is given in Fig. 9.

Fig. 9 Schematic representation of complexation between calcium(II) and dodecanoate.

This journal is

c

the Owner Societies 2012

Phys. Chem. Chem. Phys., 2012, 14, 7517–7527

7525

Changing the surfactant head group from carboxylate to the weaker Lewis base sulfate decreases the degree of coordination between the surfactant and calcium(II) and favours redissolution. At least two effects are responsible for these differences. In addition to differences associated with the binding, molecular dynamics simulations support the importance of effects on the hydration shell of both the surfactant and cation.76,92

Acknowledgements R.F.P.P. thanks FCT for a PhD grant (SFRH/BD/38696/2007). We are grateful to Dr M. E. Serra for the elemental analysis, Professor M. T. Vieira for the TG analysis and Professor A. M. de Matos Beja for the X-ray diffraction experiments.

References 1 D. Stigter, J. Phys. Chem., 1975, 79, 1008–1014. 2 D. Stigter, J. Phys. Chem., 1975, 79, 1015–1022. 3 G. Gunnarsson, B. Jonsson and H. Wennerstrom, J. Phys. Chem., 1980, 84, 3114–3121. 4 D. F. Evans, D. J. Mitchell and B. W. Ninham, J. Phys. Chem., 1984, 88, 6344–6348. 5 D. F. Evans and B. W. Ninham, J. Phys. Chem., 1986, 90, 226–234. 6 K. Holmberg, B. Jo¨nsson, B. Kronberg and B. Lindman, Surfactants and Polymers in Aqueous Solution, Wiley, Chichester, U.K., 2003. 7 C. A. Bunton, F. Nome, F. H. Quina and L. S. Romsted, Acc. Chem. Res., 1991, 24, 357–364. 8 J. N. Israelachvili, D. J. Mitchell and B. W. Ninham, J. Chem. Soc., Faraday Trans. 2, 1976, 72, 1525–1568. 9 J. N. Israelachvili, Intermolecular and Surface Forces, Academic Press, London, 1985. 10 S. Miyamoto, Bull. Chem. Soc. Jpn., 1960, 33, 375–379. 11 M. Hato and K. Shinoda, J. Phys. Chem., 1973, 77, 378–381. 12 R. Teppner, K. Haage, D. Wantke and H. Motschmann, J. Phys. Chem. B, 2000, 104, 11489–11496. 13 C. A. Bunton, K. Ohmenzetter and L. Sepulveda, J. Phys. Chem., 1977, 81, 2000–2004. 14 R. F. P. Pereira, A. J. M. Valente and H. D. Burrows, J. Mol. Liq., 2010, 156, 109–114. 15 A. J. M. Valente, H. D. Burrows, R. F. Pereira, A. C. F. Ribeiro, J. Pereira and V. M. M. Lobo, Langmuir, 2006, 22, 5625–5629. 16 S. G. Cutler, P. Meares and D. G. Hall, J. Chem. Soc., Faraday Trans. 1, 1978, 74, 1758–1767. 17 J. F. Rathman and J. F. Scamehorn, J. Phys. Chem., 1984, 88, 5807–5816. 18 I. D. Robb, J. Colloid Interface Sci., 1971, 37, 521–527. 19 J. B. Rosenholm and B. Lindman, J. Colloid Interface Sci., 1976, 57, 362–378. 20 P. Stilbs and B. Lindman, J. Colloid Interface Sci., 1974, 46, 177–179. 21 J. Clifford and B. A. Pethica, Trans. Faraday Soc., 1964, 60, 216–224. 22 A. Hafiane, I. Issid and D. Lemordant, J. Colloid Interface Sci., 1991, 142, 167–178. 23 L. T. Okano, E. O. Alonso, O. L. Waissbluth and F. H. Quina, Photochem. Photobiol., 1996, 63, 746–749. 24 N. Pilpel, Chem. Rev., 1963, 63, 221–234. 25 P. A. Winsor, Chem. Rev., 1968, 68, 1–40. 26 H. D. Burrows, The Structure, Dynamics and Equilibrium Properties of Colloidal Systems, Kluwer, Dordrecht, 1990. 27 J. Lyklema, Adv. Colloid Interface Sci., 1968, 2, 67–114. 28 A. S. Kertes and H. Gutman, Surface and Colloid Science, Plenum, New York, 1975. 29 D. E. Clarke, R. S. Lee and I. D. Robb, Faraday Discuss., 1976, 61, 165–174. 30 C. H. Rodriguez, L. H. Lowery, J. F. Scamehorn and J. H. Harwell, J. Surfactants Deterg., 2001, 4, 1–14. 31 P. A. Kralchevsky, K. D. Danov, C. I. Pishmanova, S. D. Kralchevska, N. C. Christov, K. P. Ananthapadmanabhan and A. Lips, Langmuir, 2007, 23, 3538–3553.

7526

Phys. Chem. Chem. Phys., 2012, 14, 7517–7527

32 S. Soontravanich and J. F. Scamehorn, J. Surfactants Deterg., 2010, 13, 13–18. 33 S. Soontravanich and J. F. Scamehorn, J. Surfactants Deterg., 2010, 13, 1–11. 34 J. Lucassen, J. Phys. Chem., 1966, 70, 1824–1830. 35 K. L. Stellner and J. F. Scamehorn, Langmuir, 1989, 5, 70–77. 36 S. Soontravanich, S. Walsh, J. F. Scamehorn, J. H. Harwell and D. A. Sabatini, J. Surfactants Deterg., 2009, 12, 145–154. 37 F. J. Buono and M. L. Feldman, Kirk–Othmer Encyclopedia of Chemical Technology, Wiley, New York, 1979. 38 G. D. Parfitt and J. Peacock, Surface and Colloid Science, Plenum, New York, 1978. 39 M. S. Akanni, E. K. Okoh, H. D. Burrows and H. A. Ellis, Thermochim. Acta, 1992, 208, 1–41. 40 R. G. Bossert, J. Chem. Educ., 1950, 27, 5. 41 G. C. Whitaker, Kirk–Othmer Encyclopedia of Chemical Technology, Interscience, New York, 1965. 42 H. Yamada and M. Tanaka, Adv. Inorg. Chem., 1985, 29, 143–168. 43 K. Binnemans and C. Gorller-Walrand, Chem. Rev., 2002, 102, 2303–2345. 44 B. Donnio, Curr. Opin. Colloid Interface Sci., 2002, 7, 371–394. 45 R. F. P. Pereira, M. J. Tapia, A. J. M. Valente and H. D. Burrows, Langmuir, 2012, 28, 168–177. 46 C. Elvira and J. SanRoman, J. Mater. Sci. Mater. Med., 1997, 8, 743–746. 47 K. Sakamoto, S. Yamaguchi, M. Kaneno, I. Fujihara, K. Satoh and Y. Tsunawaki, Thin Solid Films, 2008, 517, 1354–1357. 48 M. Hayashi, T. Sakai, Y. Hasegawa, T. Nishikawahara, H. Tomioka, A. Iida, N. Shimizu, M. Tomita and S. Awazu, J. Controlled Release, 1999, 62, 141–148. 49 A. Valor, E. Reguera, E. Torres-Garcia, S. Mendoza and F. Sanchez-Sinencio, Thermochim. Acta, 2002, 389, 133–139. 50 D. A. Acosta, J. F. Schmehorn and S. D. Christian, J. Pulp Pap. Sci., 2003, 29, 35–41. 51 Y. Zhao, Y. Deng and J. Y. Zhu, Prog. Pap. Recycl., 2004, 14, 5. 52 H. Zhang, C. A. Miller, P. R. Garrett and K. H. Raney, J. Colloid Interface Sci., 2003, 263, 633–644. 53 A. Zapf, R. Beck, G. Platz and H. Hoffmann, Adv. Colloid Interface Sci., 2003, 100, 349–380. 54 S. Soontravanich, H. Lopez, J. Scamehorn, D. Sabatini and D. Scheuing, J. Surfactants Deterg., 2010, 13, 367–372. 55 R. Jandacek, Lipids, 1991, 26, 250–253. 56 P. Mukerjee and K. J. Mysels, Critical Micelle Concentrations of Aqueous Surfactants Systems, NBS, Washington, DC, 1971. 57 S. N. Misra, T. N. Misra and R. C. Mehrotra, J. Inorg. Nucl. Chem., 1963, 25, 195–199. 58 P. A. Spegt and A. E. Skoulios, Acta Crystallogr., 1964, 17, 198–207. 59 A. C. F. Ribeiro, A. J. M. Valente, V. M. M. Lobo, E. F. G. Azevedo, A. M. Amado, A. M. A. da Costa, M. L. Ramos and H. D. Burrows, J. Mol. Struct., 2004, 703, 93–101. 60 J. Barthel, F. Feuerlein, R. Neueder and R. Wachter, J. Solution Chem., 1980, 9, 209–219. 61 S. L. Young and E. Matijevic, J. Colloid Interface Sci., 1977, 61, 287–301. 62 J. T. Yoke, J. Phys. Chem., 1958, 62, 753–755. 63 D. Beneventi, B. Carre´ and A. Gandini, J. Colloid Interface Sci., 2001, 237, 142–144. 64 C. A. Colman-Porter and C. B. Monk, J. Chem. Soc., 1952, 4363–4368. 65 J. W. Bunting and K. M. Thong, Can. J. Chem., 1970, 48, 1654–1656. 66 R. F. P. Pereira, M. J. Tapia, A. J. M. Valente, R. C. Evans, H. D. Burrows and R. A. Carvalho, J. Colloid Interface Sci., 2011, 354, 670–676. 67 H. D. Burrows, M. J. Tapia, S. M. Fonseca, A. J. M. Valente, V. M. M. Lobo, L. L. G. Justino, S. Qiu, S. Pradhan, U. Scherf, N. Chattopadhyay, M. Knaapila and V. M. Garamus, ACS Appl. Mater. Interfaces, 2009, 1, 864–874. 68 P. Turq, J. Barthel and M. Chemla, Transport, Relaxation and Kinetic Processes in Electrolyte Solutions, Springer-Verlag, Berlin, 1992. 69 R. F. P. Pereira, A. J. M. Valente, H. D. Burrows, M. L. Ramos, A. C. F. Ribeiro and V. M. M. Lobo, Acta Chim. Slov., 2009, 56, 45–52.

This journal is

c

the Owner Societies 2012

70 G. Nemethy and H. A. Scheraga, J. Chem. Phys., 1962, 36, 3382–3400. 71 C. G. Bazuin, D. Guillon, A. Skoulios, M. A. Amorim da Costa, H. D. Burrows, C. F. G. C. Geraldes, J. J. C. Teixeira-Dias, E. Blackmore and G. J. T. Tiddy, Liq. Cryst., 1988, 3, 1655–1670. 72 C. F. Baes and R. E. Mesmer, The Hydrolysis of Cations, Wiley, New York, 1976. 73 T. Havre, Colloid Polym. Sci., 2004, 282, 270–279. 74 C. Y. Tang, Z. Huang and H. C. Allen, J. Phys. Chem. B, 2010, 114, 17068–17076. 75 C. Y. Tang, Z. Huang and H. C. Allen, J. Phys. Chem. B, 2011, 115, 34–40. 76 H. Yan, X.-L. Guo, S.-L. Yuan and C.-B. Liu, Langmuir, 2011, 27, 5762–5771. 77 H. D. Burrows, T. O. Chimamkpam, T. Encarnac¸a˜o, S. M. Fonseca, R. F. P. Pereira, M. L. Ramos and A. J. M. Valente, J. Surf. Sci. Technol., 2010, 26, 197–212. 78 R. D. Vold, J. D. Grandline 2nd and M. J. Vold, J. Colloid Sci., 1948, 3, 339–361. 79 R. A. Hites and K. Biemann, J. Am. Chem. Soc., 1972, 94, 5772–5777. 80 K. N. Mehrotra and S. K. Upadhyaya, Recl. Trav. Chim. Pays-Bas, 1987, 106, 625–627.

This journal is

c

the Owner Societies 2012

81 E. F. Marques, H. D. Burrows and M. D. Miguel, J. Chem. Soc., Faraday Trans., 1998, 94, 1729–1736. 82 R. D. Vold and G. S. Hattiangdi, Ind. Eng. Chem., 1949, 41, 2311–2320. 83 A. Valor, S. Kycia, E. Torres-Garcıa, E. Reguera, C. Va´zquez-Ramos and F. Sa´nchez-Sinencio, J. Solid State Chem., 2003, 172, 471–479. 84 D. R. Lide, Handbook of Chemistry and Physics, CRC Press, 1995–1996. 85 K. Nakamoto, Infrared Spectra of Inorganic and Coordination Compounds, Wiley, New York, 1963. 86 G. B. Deacon and R. J. Phillips, Coord. Chem. Rev., 1980, 33, 227–250. 87 E. P. Serjeant and B. Dempsey, Ionization Constants of Organic Acids in Aqueous Solution, Pergamon, Oxford, 1979. 88 P. C. Griffiths, J. A. Roe, B. L. Bales, A. R. Pitt and A. M. Howe, Langmuir, 2000, 16, 8248–8254. 89 M. Remko, P. T. Van Duijnen and C. W. von der Lieth, THEOCHEM, 2007, 814, 119–125. 90 M. Baviere, B. Bazin and R. Aude, J. Colloid Interface Sci., 1983, 92, 580–583. 91 H. Ohtaki and T. Radnai, Chem. Rev., 1993, 93, 1157–1204. 92 H. Yan, S.-L. Yuan, G.-Y. Xu and C.-B. Liu, Langmuir, 2010, 26, 10448–10459.

Phys. Chem. Chem. Phys., 2012, 14, 7517–7527

7527