ith

moniumree

rationng

degree ofofdrity and thes

Journal of Colloid and Interface Science 286 (2005) 755–760www.elsevier.com/locate/jcis

Aggregation behavior of hexadecyltrimethylammonium surfactants wvarious counterions in aqueous solution

Nan Jianga, Peixun Lib, Yilin Wanga,∗, Jinben Wanga, Haike Yana, Robert K. Thomasb

a CAS Key Laboratory of Colloid, Interface and Chemical Thermodynamics, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100080,People’s Republic of China

b Physical and Theoretical Chemistry Laboratory, Oxford University, South Parks Road, Oxford OX1 3QZ, UK

Received 9 November 2004; accepted 21 January 2005

Available online 19 February 2005

Abstract

Both thermodynamic and microenvironmental properties of the micelles for a series of cationic surfactants hexadecyltrimethylam(C16TAX) with different counterions, F−, Cl−, Br−, NO−

3 , and ½SO2−4 , have been studied. Critical micelle concentration (CMC), deg

of micelle ionization(α), and enthalpy of micellization(�Hmic) have been obtained by conductivity measurements and isothermal titmicrocalorimetry. Both the CMC and theα increase in the order SO2−

4 < NO−3 < Br− < Cl− < F−, consistent with a decrease in bindi

of counterion, except for the divalent anion sulfate.�Hmic becomes less negative through the sequence NO−3 < Br− < Cl− < F− < SO2−

4 ,and even becomes positive for the divalent sulfate. The special behavior of sulfate is associated with both its divalency and itsdehydration. Gibbs free energies of micellization(�Gmic) and entropies of micellization(�Smic) have been calculated from the values�Hmic, CMC, andα and can be rationalized in terms of the Hofmeister series. The variations in�Hmic and�Smic have been comparewith those for the corresponding series of gemini surfactants. Electron spin resonance has been used to assess the micropolamicroviscosity of the micelles. The results show that the microenvironment of the spin probe in the C16TAX surfactant micelles dependstrongly on the binding of the counterion. 2005 Elsevier Inc. All rights reserved.

Keywords: Hexadecyltrimethylammonium; Micelles; Counterion effect

tiononre-tiontoand

thethatics

ome-tonand

fect

ag-ith

onre-n

1. Introduction

There are two competing tendencies in the formaof micelles of ionic surfactants. Removal of hydrocarbchains from water favors aggregation and electrostaticpulsions between the ionic head groups oppose aggregaCounterions stabilize ionic surfactant micelles by bindingthe micelles and screening the electrostatic repulsionshence the binding affinity of the counterion influencesprocess of micellization. Extensive studies have shownthe counterion has a strong effect on the thermodynamand aggregation properties of micellization[1–13]. How-

* Corresponding author. Fax: +86-10-82615802.E-mail address: yilinwang@iccas.ac.cn(Y. Wang).

0021-9797/$ – see front matter 2005 Elsevier Inc. All rights reserved.doi:10.1016/j.jcis.2005.01.064

.

ever, the range of counterions studied has either been swhat limited [1–4] or studies have confined themselvesthe effects of added salt[6–9]. In the latter case, counterioeffects are swamped by more general electrolyte effectsthere are relatively few systematic studies of just the efof the counterion[10,12–14].

We have previously systematically investigated thegregation behavior for six cationic gemini surfactants wvarious counterions [C12H25(CH3)2N(CH2)6N(CH3)2C12H25]X2 (C12C6C12X2, X = F−, Cl−, Br−, Ac−, NO−

3 ,and ½SO2−

4 ) in aqueous solution using isothermal titratimicrocalorimetry supplemented by conductivity measuments[15]. The variation in the enthalpy of micellizatio

was found to differ from that expected from the Hofmeisterseries[16] and the bivalent sulfate anion generated prop-erties markedly different from those of the monovalent

nd Int

ionwfac-

el-ain

y-n awastheolu-e byeze-wasmspi

ivedr-

C

redtiv-edbe-

r

singricll ofL

mMonsnda

ions

ra-ays

oftion

re-

andatAllrnale

on-Thentra-s

am-om

SR-ec-

nen-

n-andtheee

nt

756 N. Jiang et al. / Journal of Colloid a

anions, for example, a positive enthalpy of micellizatand a lower critical micelle concentration (CMC). We noextend that study to the series of single-chained surtants based on hexadecyltrimethylammonium salts C16TAX,where X= F−, Cl−, Br−, NO−

3 , and ½SO2−4 . The aim of this

work is to further understand the counterion effect on miclization and to compare the counterion effect on single-chsurfactants with gemini surfactants.

2. Experimental

2.1. Materials

Hexadecyltrimethylammonium bromide (C16TABr) waspurchased from Fluka BioChemika Co. The C16TAX se-ries with F−, Cl−, NO−

3 , and ½SO2−4 were obtained from

C16TABr by means of ionic exchange. Hexadecyltrimethlammonium bromide was dissolved in water, and thelarge excess of the basic ionic exchange resin (Dowex)added to replace bromide by hydroxide. At the end ofexchange, typically complete in less than an hour, the stion was filtered and neutralized as precisely as possiblthe appropriate acid. The resulting solution was then fredried to give the corresponding salt. The dry surfactantrecrystallized, usually more than twice, until no minimuwas observed in surface tension measurements. Theprobe 5-doxylstearic acid (5-DSA) was used as recefrom Aldrich. Triply distilled water was used in all expeiments.

2.2. Conductivity measurements

Electrical conductivity was used to determine the CMand the degree of ionization(α) of the C16TAX micelles.Conductivities of the surfactant solutions were measuas a function of concentration using a Jenway conducity meter (Model 4320). All measurements were performin a double-walled glass container with the temperatureing maintained at 298.15± 0.05 K using a circulating watebath.

2.3. Isothermal titration microcalorimetry

The calorimetric measurements were performed ua 2277-201 TAM microcalorimetric system (ThermometAB, Järfälla, Sweden) with a stainless-steel sample ce1 mL at 298.15 K. The cell was initially loaded with 0.5 mpure water and concentrated surfactant solution of 5was injected into the stirred sample cell in 45–50 portiof 10 µL using a computer-controlled syringe pump (Lu612) from a 500-µL gastight Hamilton syringe throughstainless-steel cannula. The interval between two inject

was sufficiently long for the signal to return to the baseline.Raw data curves were integrated using Digitam 4.1 softwareas described in the instrument manual. The precision of the

erface Science 286 (2005) 755–760

n

calorimeter was checked periodically by electrical calibtion and the evaluation results of calibration were alwin the required range for the TAM system. The accuracymeasurements was verified by measurement of the diluenthalpies of concentrated sucrose solution (0.985 mol L−1),which were in excellent agreement(±1%) with the literaturevalue[17]. All experiments were repeated twice, and theproducibility was within±4%.

2.4. Electron spin resonance (ESR)

The ESR spectra were recorded at 298.15 ± 0.20 Kon a Bruker ESP 300E spectrometer operating at X-b(9–10 GHz) with 100-kHz magnetic field modulation3.18 mW microwave power to avoid power saturation.experiments were run using a quartz tube with an extediameter of∼1.5 mm. The average relative error for throtational correlation time was better than 10%. The ccentrations of all surfactant solutions were above CMC.spin probe concentration was kept at a constant concetion of 5 × 10−5 mol L−1, which is usually considered aa negligible perturbation to the micelle structure. The sples were stirred for 8 h and then stabilized for 3 h at rotemperature before measurements.

From the ESR spectra, the rotational correlation time(τc)

can be calculated using the following equation[18,19],

(1)τc = (6.6× 10−10)W0

[(h0

h−1

)1/2

+(

h0

h+1

)1/2

− 2

],

whereW0 represents the peak-to-peak linewidth of the Emid-field line (in gauss) andh+1, h0, h−1 are the peak-topeak heights of the low-, mid-, and high-field lines, resptively. The hyperfine coupling constant(AN) can be calcu-lated using the following relation[20,21],

(2)AN = (A‖ + 2A⊥)/3,

whereA‖ is the time-averaged electron-nuclear hyperfitensor (parallel) andA⊥ is the time-averaged electronuclear hyperfine tensor (perpendicular).

3. Results and discussion

Fig. 1 shows the variation of electrical conductivityκ ofthe C16TAX surfactant solutions with the surfactant concetrationC at 298.15 K. There are clear breaks in the plotsthe sharpness of the break helps to confirm the purity ofsurfactant sample[22], while the breaks give the CMCs. Thdegree of micelle ionization,α, was taken as the ratio of thdκ/dC values above and below the CMC[23,24]. Table 1presents the resulting CMC andα values of the C16TAX sur-factants, together with earlier reported values. The preseα

values, except for C16TABr, basically agree well with thosereported by Sepúlveda et al.[12], where C16TAF was notstudied.

nd

N. Jiang et al. / Journal of Colloid and Interface Science 286 (2005) 755–760 757

Table 1Critical micelle concentration (CMC), degree of micelle ionization(α) and thermodynamic parameters for the C16TAX surfactants at 298.15 K

Anion CMC (mM) α �Hmic �Gmicc T �Smic Hydration

No.Conductivity Microcalorimetry (kJ mol−1)

SO2−4 0.61± 0.03 0.54± 0.02 0.29± 0.02 3.1± 0.4 −24.8 27.9 8d

(0.26)a –NO−

3 0.89± 0.03 0.89± 0.03 0.31± 0.02 −7.8± 0.5 −29.4 21.6 –(0.30)a (−10.6)b

Br− 0.95± 0.03 0.94± 0.03 0.33± 0.02 −6.9± 0.3 −28.9 22.0 1.5d

(0.22)a (−8.7)b

Cl− 1.15± 0.03 1.16± 0.02 0.35± 0.02 −4.4± 0.3 −27.6 23.2 2d

(0.37)a (−2.3)b

F− 1.62± 0.02 1.54± 0.04 0.48± 0.02 −0.8± 0.3 −24.4 23.6 5e

– (0.5)b

a From Ref.[12].b From Ref.[10].c Calculated from�Gmic = RT (1 + β) ln CMC and�Gmic = RT (1 + β/2) ln CMC − (RTβ ln 2)/2 for surfactants with monovalent counterions a

divalent counterion respectively[25], whereβ is given byβ = 1− α.

eet-

sig-intore-thehetesra-nden-rvednal-

from

-

oci--

-

s,for

hy-thechctantboth

d From Ref.[1].e From Ref.[26].

Fig. 1. Variations of the conductivityκ of the C16TAX aqueous solutionsversus the C16TAX concentrationC at 298.15 K.

The calorimetric titration curves of C16TAX surfactantsat 298.15 K are given inFig. 2, where the data points arthe enthalpies obtained from isothermal titration calorimric curves. The enthalpy curves in pure water are allmoidal in shape and each curve can be subdividedtwo concentration ranges with an intermediate transitiongion associated with micelle formation, corresponding toCMC. When the final dilution concentration is below tCMC, dilution breaks up the micelles and further diluthe monomer solution. When the final dilution concenttion is above the CMC, only free micelles are diluted a�Hobs drops toward zero. In the present study, the differtial enthalpy was used to determine the CMC. The obseenthalpy curves were differentiated with respect to the fisurfactant concentrationC and the extremum of the differential curve was taken to be the CMC[27]. There is anexcellent agreement between the CMC values obtained

microcalorimetry and conductivity. The enthalpy of micel-lization (�Hmic) can be determined from the difference be-tween the observed enthalpies of the two linear segments o

Fig. 2. Variations of the observed enthalpies(�Hobs) versus the final concentration(C) of the C16TAX surfactants at 298.15 K.

the plots at the CMC[27,28]as shown inFig. 3. The Gibbsfree energies of micellization(�Gmic) are calculated fromthe value of the CMC and the degree of counterion assation to micelle(β), following the procedure in the literature[25]. The entropy of micellization(�Smic) can be thenderived from�Hmic and�Gmic. All the thermodynamic parameters obtained are listed inTable 1.

As seen inTable 1, both the CMC and theα values ofthe C16TAX surfactants increase in the sequence SO2−

4 <

NO−3 < Br− < Cl− < F−. For the monovalent counterion

this order correlates exactly with the Hofmeister seriesanions[9], NO−

3 > Br− > Cl− > F− ∼ SO2−4 . The position

of an anion in this series is considered to depend on itsdrated radius[29]. A smaller hydrated radius enhancesability of the anion to bind at the micellar surface, whidecreases the electrostatic repulsion between the surfahead groups and hence favors aggregation. This lowers

f

the CMC and theα, and these two quantities therefore fol-low the Hofmeister series for the monovalent counterions.The divalent sulfate, however, behaves differently. Hexade-

nd Int

the

ndnte-ionatwa-the

elynte-ies.ss

try

bemict

he

lher-ein

-own

hismi-e

toac-

iontatic

ticy

n of

mM

en-o ad ather-py.the

othion-orop-

that

ere

sfor

ngly

he

ofthet.

mi-ex-f the

t arehefac-es of

758 N. Jiang et al. / Journal of Colloid a

Fig. 3. Determination of the CMC and�Hmic from the calorimetric titra-tion curve of C16TABr into water at 298.15 K. The solid line representsdifferential of the calorimetric titration curve.

cyltrimethylammonium sulfate has values of the CMC aα lower than the other surfactants with monovalent courions. Although the large hydrated radius of the sulfatemay hinder its ability to bind to the micelle, it is clear ththe divalency and the greater distance of the hydrationter from the center of the ion are dominant and enablesulfate ion to neutralize the micellar charge very effectivand hence gives a lower CMC and a lower degree of courion dissociation than expected from the Hofmeister serThe role of the valency of the ion in the micellization procewill be discussed further in connection with the calorimeresults below.

Enthalpies of micellization are generally expected toexothermic and dilution should therefore be an endotherprocess, as observed for the C16TAX series with monovalencounterions. However, for C16TA½SO4, dilution is exother-mic and the enthalpy of micellization is endothermic. Tvalues of�Hmic shown inTable 1fairly agree with previ-ously reported values[10], apart from F−, where our smalexothermic value contrasts with the previous small endotmic value. Both�Hmic and�Gmic become more negativin line with the Hofmeister series, except for an inversion�Gmic between F− and SO2−

4 . The entropies of micellization are all positive and tend to become less positive dthe same series. In all cases−�Hmic < T �Smic indicatesthat the micellization process is mainly entropy-driven. Tis especially the case for sulfate, where the entropy ofcellization is the only driving force for micellization, sinc�Hmic is unfavorable.

As the monovalent counterion changes from nitratefluoride down the Hofmeister series, the dehydrationcompanying the binding of counterion with surfactantincreases, leading to an increasing endothermic electroscontribution to�Hmic, an increasing positive electrostafree energy to�Gmic, and an increasing gain in entropfrom the released water molecules. The strong hydratio

the sulfate ion[30] has the tendency to hinder its interac-tion with the head groups of the surfactant. However, thestronger electric field in the sulfate as a divalent ion and the

erface Science 286 (2005) 755–760

Fig. 4. Representative ESR spectra of 5-DSA in water and in 15C16TAX surfactant solutions.

greater distance of the hydration water from the ion cter in the sulfate than in monovalent ions would lead tstronger electrostatic interaction with surfactant ion angreater release of water molecules, resulting in an endomic contribution to the enthalpy and an increase in entroThere would also be a smaller decrease of entropy frombinding of a single ion rather than two monovalent ions. Bfactors must contribute to the higher entropy of micellizatobserved for the sulfate system. In�Gmic, however, the endothermic contribution to the enthalpy from dehydrationpartial dehydration of the ion more than compensates theposite effect of the large entropy release to give a valueis lower than all the other counterions except F−.

The variation in the enthalpy of micellization of thC12C6C12X2 gemini surfactants was found to be mocomplex than that observed here for the C16TAX surfac-tants[15]. In particular�Hmic for the chloride species wafound to be anomalous in that it was less exothermic thanthe other counterions. This was attributed to the less strobound hydration shell of the chloride ion[31] and the strongelectrostatic field in the doubly charged gemini ion. Tsame effect evidently does not occur for the C16TAX series,since there is no inversion of the order of�Hmic betweenchloride and fluoride. The greater extent of dehydrationthe chloride in the geminis must therefore be a result ofhigh electrostatic field in the doubly charged environmen

In order to further understand the difference of thecelle microenvironments for different counterions, ESRperiments have been conducted. All the ESR spectra oprobe 5-DSA in the C16TAX micellar solutions exhibit thethree-line pattern and the broadened high-field lines thaconsistent with partial hindered rotational mobility of tprobe 5-DSA molecules as it is incorporated into the surtant micelles. Representative ESR spectra in various typ

surfactant solutions are presented inFig. 4.

The variations of the rotational correction timeτc and thehyperfine coupling constantAN of 5-DSA with the surfac-

nd In

pec--yer

dicrre-ndcos-obe

byxy-ent

s-

-. Theu-theithith

in aota-, F

fac-theofore

tant

ne-

ofe-

ri-ty atromichon.yoff

ech-se po-

d

as a

pro-pite

ob-

ely,ries,

of

and

r ofys-tersmi-ied

icel-es.

N. Jiang et al. / Journal of Colloid a

Fig. 5. Variations of (a)τc and (b)AN of 5-DSA in C16TAX aqueous so-lutions at different surfactant concentrations above CMC: (Q) C16TANO3;(") C16TABr; (2) C16TACl; (×) C16TA½SO4; (+) C16TAF. The solidlines are guides to the eye.

tant concentration above the CMC obtained from ESR stra are shown inFig. 5. Kevan et al.[32] reported that the 5DSA spin probe is located in the inner part of the Stern la(the polar head group region) of a micelle with its acigroup at the micelle–water interface. The rotational colation time reflects the rotational mobility of the probe acan be used to monitor the changes in the local microvisity among the surfactant systems experienced by the prThe hyperfine coupling constantAN can give informationabout the micropolarity of the microenvironment sensedthe nitroxide probe. Due to a greater electron density at ogen and spin density at nitrogen, a more polar environmproduces a largerAN.

The values ofτc vary in the order related to the Hofmeiter series, that is, C16TAF ≈ C16TA½SO4 < C16TACl <

C16TABr ≈ C16TANO3, which implies that the microviscosity sensed by the probe increases in the same orderincrease in microviscosity is produced by the gradual ntralization of the micellar surface, which is caused bybinding of counterions to the micellar surface. The ion wthe smallest hydrated radius interacts most strongly wthe oppositely charged head groups. This would resultmore tightly head group packing and according slower rtional motion of the probe. As already discussed above−

−

binds less strongly to the micelle, whereas NO3 binds moststrongly. Meanwhile, the effects of changing concentrationwere also measured. The change ofτc with concentration

terface Science 286 (2005) 755–760 759

.

e

above the CMC shows a slight increase for all the surtants. This increase reflects the motional restriction ofprobe within the micelle accompanying the slow growthmicelle amount and micelle aggregation number and mcompact packing of micelles with an increase of surfacconcentration.

The 5-DSA probe in water shows a high hyperficoupling constant (AN = 15.8 G), due to the highly polar environment[33]. In C16TAX micellar solution, lowerAN values were observed, indicating the lower polaritythe micelles. TheAN values become smaller in the squence C16TAF ≈ C16TA½SO4 > C16TACl > C16TABr >

C16TANO3. Clearly, the increased binding of the counteons along this series leads to a decreasing micropolarithe binding sites of the spin probe as would be expected fthe closer packing associated with tighter binding, whwould reduce water penetration into the interfacial regiAs shown inFig. 5b, the values ofAN also decrease slightlwith increasing surfactant concentration before levelinggradually. This can probably be attributed to the same manism as for the change inτc, i.e., micellar growth hinderwater penetration into the Stern region and decreases thlarity. It is worth noting that SO2−

4 has relatively smallτcand largeAN, corresponding to its low microviscosity anhigh micropolarity, suggesting that the C16TA½SO4 micelle,compared with those having monovalent counterions, hless tight packing structure, which may indicate that SO2−

4might be not as effective as monovalent counterions inmoting the aggregation of the surfactant molecules, desits much lower CMC andα.

The ESR results were also supported by the resultstained from dynamic light scattering (DLS)[1] and SANSdata[3]. The hydrodynamic radiusRh values obtained byBiresaw et al.[1] for the C16TABr, C16TACl, and C16TA½SO4 micelles are 32.2, 29.0, and 27.2 Å, respectivchanging in a manner consistent with the Hofmeister seand supporting the ESR results that SO2−

4 is not as effec-tive as monovalent counterions in promoting the growthmicelles. The aggregation numberN from SANS data indi-cates that nitrate and bromide micelles are similar in sizelarger than chloride micelles[3] and this is qualitatively inline with the results given inTable 1.

4. Summary

The detailed information on the aggregation behavioa series of C16TAX surfactants in aqueous solution was stematically studied. The various thermodynamic parameindicate that the counterion, X, has a large influence oncellization. For the series of monovalent counterions studhere, the CMCs and free energies and enthalpies of mlization generally correlate well with the Hofmeister seri

However, for the divalent sulfate the behavior is more com-plex and there are features of the micellization that are prob-ably associated with the divalency of the ion as well as

nd Int

thes-m.SR

f as

re-

alci-nd,

ciety.

s.

ir 8

5611

13

04

871.0.

. 95

ys.

39)

.P.

. Re-

.g-

83)

. 1 83

58,

M.y-988,

M.J.

760 N. Jiang et al. / Journal of Colloid a

changes in hydration. The results show that entropy ismain driving force for micellization in the monovalent sytems and is the only driving force in the divalent systeThe effects of the counterion are further supported by Eresults. The substantial differences inτc andAN among theC16TAX surfactants suggest that the microenvironment ospin probe in the micelles of C16TAX surfactants dependstrongly on the binding of the counterion X−, exactly aswould be expected from the conductivity and calorimetrysults.

Acknowledgments

We are grateful for financial support from the NationNatural Science Foundation of China, the National Sence and Technology Committee, CNPC Innovation Futhe Chinese Academy of Sciences, and the Royal So(20233010, 20473101, 2001AA602014-2, Grants Q810)

References

[1] G. Biresaw, D.C. McKenzie, C.A. Bunton, D.F. Nicoli, J. PhyChem. 89 (1985) 5144.

[2] M.S. Vethamuthu, M. Almgren, E. Mukhtar, P. Bahadur, Langmu(1992) 2396.

[3] S. Berr, R.R.M. Jones, J.S. Johnson Jr., J. Phys. Chem. 96 (1992)[4] L.S. Romsted, J. Phys. Chem. 89 (1985) 5107.[5] L. Gaillon, J. Lelièvre, R. Gaboriaud, J. Colloid Interface Sci. 2

(1999) 287.[6] L.S. Romsted, J. Phys. Chem. 89 (1985) 5113.[7] H.-P. Lin, C.-P. Kao, C.-Y. Mou, S.-B. Liu, J. Phys. Chem. B 1

(2000) 7885.[8] E.W. Anacker, H.M. Ghose, J. Phys. Chem. 67 (1963) 1713.

erface Science 286 (2005) 755–760

.

[9] J.W. Larsen, L.J. Magid, J. Am. Chem. Soc. 96 (1974) 5774.[10] S. Paredes, M. Tribout, L. Sepúlveda, J. Phys. Chem. 88 (1984) 1[11] C. Gamboa, H. Ríos, L. Sepúlveda, J. Phys. Chem. 93 (1989) 554[12] L. Sepúlveda, J. Cortés, J. Phys. Chem. 89 (1985) 5322.[13] M.M. Knock, C.D. Bain, Langmuir 16 (2000) 2857.[14] P.-A. Bergström, J. Lindgren, O. Kristiansson, J. Phys. Chem

(1991) 8575.[15] N. Jiang, P. Li, Y. Wang, J. Wang, H. Yan, R.K. Thomas, J. Ph

Chem. B 108 (2004) 15385.[16] F. Hofmeister, Arch. Exp. Pathol. Pharmakol. 24 (1888) 247.[17] F.T. Gucker, H.B. Pickard, R.W. Planck, J. Am. Chem. Soc. 61 (19

459.[18] B. Cannon, C.F. Polnaszek, K.W. Butler, L.E.G. Eriksson, I.C

Smith, Arch. Biochem. Biophys. 167 (1975) 505.[19] S. Schreier, J.R. Ernandes, I. Cuccovia, H. Chaimovich, J. Magn

son. 30 (1978) 283.[20] W.L. Hubbell, H.M. McConnell, J. Am. Chem. Soc. 93 (1971) 314[21] K. Tajima, Y. Imai, T. Horiuchi, M. Koshinuma, A. Nakamura, Lan

muir 12 (1996) 6651.[22] R. Zana, M. Benrraou, R. Rueff, Langmuir 7 (1991) 1072.[23] R. Zana, J. Colloid Interface Sci. 78 (1980) 330.[24] H.C. Evans, J. Chem. Soc. (1956) 579.[25] R. Zana, Langmuir 12 (1996) 1208.[26] R.B. Dorshow, C.A. Bunton, D.F. Nlcoli, J. Phys. Chem. 87 (19

1409.[27] Z. Király, I. Dekány, J. Colloid Interface Sci. 242 (2001) 214.[28] I. Johnson, G. Olofsson, B. Jönsson, J. Chem. Soc. Faraday Trans

(1987) 3331.[29] E.R. Nightingale Jr., J. Phys. Chem. 63 (1959) 1381.[30] J.A.A. Ketelaar, Chemical Constitution, Elsevier, New York, 19

p. 102.[31] M. Magini, G. Licheri, G. Paschina, G. Piccaluga, G. Pinna, in:

Magini (Ed.), X-Ray Diffraction of Ions in Aqueous Solutions: Hdration and Complex Formation, CRC Press, Boca Raton, FL, 1p. 160.

[32] E. Szajdzinska-Pietek, R. Maldonado, L. Kevan, R.R.M. Jones,Coleman, J. Am. Chem. Soc. 107 (1985) 784.

[33] M.F. Ottaviani, G. Martini, L. Nuti, Magn. Reson. Chem. 25 (1987)897.