Urea-Induced Decrease of Anion Selectivity in Surfactant Aggregates

Micellar catalysis, a useful misnomer. Laurence S Romsted , Clifford A Bunton , Jihu Yao. Current Opinion in Colloid & Interface Science 1997 2, 622-6...
0 downloads 0 Views 143KB Size
+

+

1166

Langmuir 1996, 12, 1166-1171

Urea-Induced Decrease of Anion Selectivity in Surfactant Aggregates Fa´bio Herbst Florenzano, Luı´s Geraldo Cardoso dos Santos, Iolanda Midea Cuccovia, Maria V. Scarpa, Hernan Chaimovich, and Ma´rio Jose´ Politi* Departamento de Bioquı´mica and Laborato´ rio Interdepartamental de Cine´ tica Ra´ pida, Instituto de Quı´mica, Universidade de Sa˜ o Paulo, Caixa Postal 26077, Sa˜ o Paulo, S.P., 05599-970 Brazil Received July 17, 1995. In Final Form: November 9, 1995X Direct and indirect mechanisms have been used to explain the effect of urea on surfactant aggregates. By studying formation properties, anion selectivities, and stabilities of cationic and zwitterionic micelles and of cationic vesicles, we confirm previous evidence in favor of the direct mechanism (Langmuir 1995, 11, 1715). Differential fluorescence suppression of 2,3-N-butylnaphthalimide by Br- and Cl- is used to investigate in a short time scale (nanoseconds) effects of urea on ion binding and exchange in micelles and vesicles. Enhanced degree of dissociation and decreased anion selectivity of the surfactant aggregates are observed. Micellar catalysis of the unimolecular decarboxylation of 3-carboxy-6-nitrobenzoxazol studied to investigate the effect of the additive on a long range (minutes) also shows a reduced anion selectivity. Formation properties of the thermodynamically stable (cmc) and nonstable aggregates (hydrodynamic radii and molecular weight) are measured to obtain parameters for analysis. Addition of urea caused increase in the cmc and led to increase in vesicles molecular weight and hydrodynamic radii. These results support preferential headgroup solvation in the aggregates (direct mechanism) by urea resulting in the loss, or size reduction, of the interfacial region for counterion specificity (size, polarizability, and other non-Coulombic terms) to manifest.

Introduction The interplay of urea diminishing the hydrophobic effect is responsible for the denaturation of macromolecules native conformation.1,2 The molecular details of its action however are still a matter of high controversy.3-8 More recently amphiphile macromolecular aggregates appeared in this scenario and the effect of urea and its derivatives upon properties of surfactant solutions were investigated. Given the usual straightforward nature of simple surfactant molecules, that is, well-defined hydrophobic and hydrophilic regions, it was hoped that they could provide an understanding of the urea effect. Unfortunately data in the literature did not answer the question of how urea acts. The same two extreme interpretations standing for protein solutions remain alive with micellar ones. In other words direct and indirect mechanisms have been argued to rationalize experimental observations as increased critical micellar concentration (cmc) for ionic and nonionic surfactants9-14 and as increased ionic dissociation degree (R) for ionic ones. We have addressed this question using various approaches starting from urea effects on simple proton * Author for correspondence: e-mail, [email protected]; fax, (55)(011)8155579. X Abstract published in Advance ACS Abstracts, February 1, 1996. (1) Franks, F. Water: A Comprehensive Treatise; Plenum Press: New York; 1978. (2) Robinson, D. R.; Jencks, W. P. J. Am. Chem. Soc. 1965, 87, 2462. (3) Baglioni, P.; Ferroni, E.; Kevan, L. J. Phys. Chem. 1990, 94, 4296. (4) Kang, Y. S.; McManus, H. J. D.; Kevan, L. J. Phys. Chem. 1992, 96, 10055. (5) Kuharski, R. A.; Ressky, P. J. J. Am. Chem. Soc. 1984, 106, 5786. (6) Nandi, P. K.; Robinson, D. R. Biochemistry 1984, 23, 6661. (7) Wetlaufer, D. B.; et al. J. Am. Chem. Soc. 1964, 86, 508. (8) Baglioni, P.; et al. J. Phys. Chem. 1990, 94, 8218. (9) Bruning, W.; Holtzer, A. J. Am. Chem. Soc. 1961, 83, 4865. (10) Emerson, M. F.; Holtzer, A. J. Phys. Chem. 1967, 71, 3320. (11) Evans, H. C. J. Chem. Soc. 1956, part 1, 579. (12) Mukerjee, P.; Ray, A. J. Phys. Chem. 1963, 67, 190. (13) Ruiz, C. C.; Sa´nchez, F. G. J. Colloid Interface Sci. 1994, 165, 110. (14) Schick, M. J. J. Phys. Chem. 1964, 68, 3585.

transfer reactions15 to effects on percolative behavior of AOT reversed micelles.16 In a recent study we examined the effect of the additive on the surface-pressure isotherm properties of quaternary ammonium monolayers.17 Observed increase in the limiting area per monomer and loss of ionic selectivity by the monolayer were attributed to binding of urea to the interface, that is, an unequivocal expression of the direct interaction mechanism. In order to check the validity of these observations to other biomimetic aggregates, we pursue further our studies to three selected systems, namely, cationic cetyltrimethylammonium chloride (CTACl) and bromide (CTABr) micelles, zwitterionic hexadecyldimethylammonium propanesulfonate micelles (HPS), and dialkyldimethylammonium chloride (HERQUAT) small unilamellar vesicles (SUV). These macromolecular aggregates present a common quaternary group residing in the interfacial region. Principal differences among them are surface charge density and packing of molecules in the surfactant aggregate. Here we wish to demonstrate by examining the effect of urea on (i) the formation properties of the aggregates, (ii) the counterion fluorescence dynamic suppression of a fluorophore incorporated in the aggregates (nanosecond time basis), and (iii) the reaction rate constant of a reactive counterion species (minute time basis), that monomer packing, ion binding (condensation), and exchange within these aggregates behave as found for ammonium monolayers.17 The evidence is again very strong toward the direct urea interaction mechanism. Materials and Methods Materials. Urea (Merck) was triply recrystallized from hot ethanol (70%). Conductance of a 10 M aqueous urea solution was lower than 10 µS cm-1, warranting no electrolyte contamination and also no urea hydrolysis. Urea solutions were freshly (15) Politi, M. J.; Chaimovich, H. J. Solution Chem. 1989, 18, 1055. (16) Amaral, C. L. C.; Brino, O.; Chaimovich, H.; Politi, M. J. Langmuir 1992, 8, 2417. (17) De Souza, S. M. B.; Chaimovich, H.; Politi, M. J. Langmuir 1995, 11, 1715.

+

+

Effect of Urea on Surfactant Aggregates prepared and discharged after use. For long standing solutions, as those for determination of vesicles molecular weight (MW) as function of time (days), the extent of urea degradation was followed by NH3 titration according to Forman’s method.18 Typically, after 30 days urea decomposed to a maximum of 0.15%. Dialkyldimethylammonium chloride (Herquat, HERGA Brazil) was Soxlet extracted with peroxide free ether19 followed by triple recrystallization from hot acetone/ethanol (85:15 (v:v)). GC chromatographic analysis via Hofman degradation of the hydroxide form of Herquat20 revealed a tail composition of 85% octadecyl and 15% hexadecyl with traces of tetradecyl. Herquat stock solution concentrations were determined either by halide titration21 or by following the absorbance of the complex between Herquat and Orange G.22 Cetyltrimethylammonium halides (CTACl & CTABr) were recrystallized from acetone/methanol (85:15 (v:v)). Stock solutions were titrated for halide content.21 Hexadecyldimethylammonium propanesulfonate (HPS) was synthesized and purified following a known procedure;23 elemental analysis, H-NMR, and melting point were in agreement with literature.24 HPS solutions were also made with a commercially available compound (Eastman Kodak), which was previously recrystallized from hot acetone/methanol (85:15 (v: v)). N-Butyl-2,3-naphthalimide (β-NBN) was synthesized by one of our group and used as received.25 Sodium 3-carboxy-6nitrobenzoxazol (NBOC) was a gift from Professor C. A. Bunton from Santa Barbara University and used as received. Electrolyte solutions were prepared with analytical grade salts (Merck), and stock solutions were standardized by halide titration.21 Water was bidistilled from an all-glass apparatus and was further deionized and ultrafiltrated (0.2 µm) through a Milli-Q system (Millipore). Methods. Effect of urea on the cmc of HPS was determined by a surface tension method using an automated apparatus (du Nou¨y’s tensiometer, Lauda, model TE1C/3). For CTACl the effect of urea on the cmc was monitored by standard specific conductivity assay using a Digimed (model CD-20) conductivimeter. cmc values were calculated form the line intersection obtained from data linear regression well below and above the cmc region26 (Table 1). Values of the aggregate dissociation degree (R) were calculated from the ratio of the slopes between these lines.27 Herquat vesicles were prepared by tip ultrasonication (Braunsonic model 1510). Herquat suspensions were sonicated for 15 min at a nominal power of 100 mW. For fluorometric titrations, SUV preparations, in absence and presence of urea, had 5 mM NaCl. For light scattering assays the suspensions were centrifuged (5000 rpm/30 min, in Servile RCA-B) for removal of residual titanium particles. Vesicles were further ultrafiltrate through 1 µm Nylon membranes (Gelman) and diluted as necessary with an ultrafiltrated salt solution. For these assays suspensions had 1 or 2 mM NaCl. Solutions were kept with rigorously clean capped vials (scintillation vials) to avoid dusty contamination or solvent loss due to evaporation. Static and dynamic light scattering experiments were realized with a homemade system consisting of a 10 mW He-Ne vertically polarized laser (Hughes), goniometer, lens for focusing the beam in the dynamic mode, slits and apertures to assure coherent scattering volumes, a head-on PMT (Thorn-EMI), high voltage power supply (Tectrol, Brazil), discriminator APED-II (ThornEMI), and an autocorrelator (Brookhaven BI-2030) for data accumulation. For static measurements, scattered intensities were taken from the accumulated number of photons in channel A for a constant time base from the correlator (1 second). Toluene (Merck) was used as reference for system calibration.28 Averaged (18) Forman, D. T. Clin. Chem. 1964, 10, 497. (19) Vogel, A. A Textbook of Practical Organic Chemistry Including Qualitative Organic Analysis; Longmans, Green and Co.: London, 1966. (20) Ribaldo, E. J.; et al. J. Colloid Interface Sci. 1984, 97, 115. (21) Schales, O.; Schales, S. S. J. Biol. Chem. 1941, 140, 879. (22) Stelmo, M.; et al. J. Colloid Interface Sci. 1987, 117, 1. (23) Brochsztain; et al. J. Phys. Chem. 1990, 94, 6781. (24) Fendler, E. J.; et al. J. Phys. Chem. 1972, 76, 1460. (25) Barros, T. C.; et al. J. Photochem. Photobiol. 1993, 76, 55. (26) Mukerjee, D.; Mysels, K. J. In Critical Micelle Concentrations of Aqueous Surfactant Systems; NSRDS; 1971. (27) De Lisi, R.; Fiscaro, E.; Milioto, S. J. Solution Chem. 1988, 17, 1015. (28) Bender, T. M.; Lewis, R. J.; Pecora, R. Macromolecules 1986, 19, 244.

Langmuir, Vol. 12, No. 5, 1996 1167 Table 1. Critical Micellar Concentrations (cmc) of HPS and CTAX and Dissociation Degree (r) of CTAX Micelles, as a Function of Urea Concentration (T ) 30 °C)

surfactant

urea concentration (M)

cmc (M)

R

CTABr (in 10 mM TRIS-HBr buffer, pH ) 8.10)

0 0.05 0.10 0.25 0.50 0.75 1.00 2.00 5.00 0 0.5 1 3 5 0 5

2.09 × 10-5 2.28 × 10-5 2.00 × 10-5 2.28 × 10-5 2.28 × 10-5 2.54 × 10-5 2.68 × 10-5 3.73 × 10-5 8.11 × 10-5 1.60 × 10-3 1.64 × 10-3 1.68 × 10-3 2.31 × 10-3 3.03 × 10-3 4.19 × 10-4 1.25 × 10-3

0.33 0.36 0.37 0.42 0.44 0.24 0.28

HPS

CTACl

weight molecular weight (MW) determinations were analyzed by the usual Zimm plot using a software developed by Professor W. Reed from the University of Tulane. Refractive index increments (dn/dc) for SUV were measured using the interferometric method described by Wilson.29 Determined dn/dc values were 0.166 and 0.175 g-1 cm3 in water and in urea (3 or 5 M), respectively. The value measured in the presence of 5 M urea is not precise since the working area in the refratometric cuvette became, for unknown reasons, very reduced and did not allowed proper scanning of the solution. MW determinations within this condition are therefore qualitative and included in results for the purpose of comparisons. Hydrodynamic radii (Rh) were obtained by the cumulants method using the software from Brokhave, Inc. Fluorometric determinations were performed either in a LS-5 (Perkin-Elmer) spectrofluorometer interfaced to an IBM-PC compatible microcomputer or in a LS-1 (PTI, Canada) spectrofluorometer using the ratio mode. Fluorescence emission spectra are noncorrected. Usually a 20 µL aliquot of β-NBN stock solution ([β-NBN] ∼ 3 × 10-3 M) in acetonitrile was added to a 5 mL volumetric flask and the solvent evaporated with a stream of N2. Proper volumes of surfactant and salts were added to the flask and after 2.5 mL transferred to a 1 cm optical path length quartz cuvette. All extra solutions added to the assay sample had β-NBN in the same concentration to avoid dilution effects. UV-vis spectrophotometric data were recorded on a Hitachi U-2000 interfaced, via RS 232C, to an IBM-PC compatible microcomputer, for data storage and handling. Decarboxylation kinetics of NBOC were monitored with a DU-7 spectrophotometer (Beckman). Product formation was followed at 410 nm with the reaction temperature fixed invariably at 303 K. Kinetic parameters were obtained with standard first-order data fitting software. The decarboxylation process in bulk water and with 5 M urea had a lifetime on the order of 26 h (see below). Reaction rate constants (kobs) for these experiments were calculated from their initial velocities once it is know that the formed nitrile slowly hydrolyses to the corresponding amide resulting in an erroneous reaction end determination.30

Results Prior to examination of the effect of urea on the interfacial properties of the surfactant aggregates under study, it is necessary to establish its effect on the formation properties of the aggregate. Effects of the additive on the cmc’s of CTAX (Cl- and Br-) and HPS micelles are summarized in Table 1. The following are observed: an increase in the cmc for both surfactants with addition of urea, for HPS in the presence of 5 M urea a 4-fold increase is observed, whereas for CTAX the increase is 2-fold. The (29) Wilson, T.; Reed, W. F. Am. J. Phys. 1993, 61, 1046. (30) Kemp, D. S.; Cox, D. D.; Paul, K. G. J. Am. Chem. Soc. 1975, 97, 7312.

+

+

1168

Langmuir, Vol. 12, No. 5, 1996

Florenzano et al. Table 2. Association Constant (KS) of β-NBN in HPS and CTACl Surfactant Micelles and HERQUAT Vesicles 10-3 KSa (M-1)

a

aggregate

in water

in 3 M urea

HPS micelles CTACl micelles HERQUAT vesicles

13.3 11.5 2.51

2.73 2.42 0.45

Accurancy is in order of 5-10%.

Figure 1. Stern-Volmer plots of β-NBN (4.14 × 10-6 M) fluorescence suppression by NaCl (O) and NaBr (0) in aqueous solutions and NaCl (b) and NaBr (9) in 3 M urea: T ) 30 °C, λexc ) 330 nm. Chart 1. Structure of N-Butyl-2,3-naphthalimide (β-NBN)

larger effect for HPS points to its higher monomer solubilization by urea. Overall changes in cmc’s presented in Table 1 are in agreement with previous studies.9-14 R’s calculated from the regression lines below and above the cmc region (see methods) show an increase in the parameter with urea concentration. Differences in cmc and in R by changing the counterion, Cl- to Br-, are related to the higher affinity of Br- by the cationic interface.31 The surfactant employed for vesicle preparation (Herquat, see materials section) did not show detectable solubility even with 5 M urea warranting its use for vesicle preparation. The high difference in collisional quenching efficiencies between Br- and Cl- ions on β-NBN (Scheme 1) fluorescence emission is a convenient tool for monitoring interfacial concentrations of these species.4 Accordingly one can correlation emission intensities with the relative amount of these ions at the interface. Stern-Volmer suppression plots for aqueous β-NBN show that Brquenching efficienty is ∼40-fold higher than that for Cl(Figure 1). Addition of urea reduces individual SternVolmer suppression constants (Ksv) compared to those in pure water but their relative efficiencies remain in the same order (KSV’s ) 40 and 28 M-1 for Br- and 1 and 0.26 M-1 for Cl-, in absence and presence of 3 M urea, respectively) (Figure 1). The association of β-NBN with the macromolecular aggregates was accessed by monitoring the blue shift in wavelength emission maxima of β-NBN in going from water or aqueous urea to the less polar micelle or vesicle microenvironment24 (Table 2). Incorporation of β-NBN in HPS micelles (in the presence of 3 M urea) is presented for exemplification in Figure 2. A blue shift from ∼417 nm in 3 M urea aqueous media to ∼397 nm with HPS > 5 mM is observed. Using these limiting wavelength values as representatives of only free probe (water) or fully incorporated (aggregate), the binding isotherm was con(31) Morgan, J. P. D. PhD Thesis University of Sydney, 1995.

Figure 2. λmax of NBN emission spectra as a function of HPS micelles concentration. This curve shows a typical incorporation of NBN in aggregates. Introduction of this naphthalimide in a more apolar medium causes a blue shift in emission spectra. Insert: double reciprocal plot of emission wavelength maximum versus HPS concentration, according to eq 2.

structed from the usual association equilibrium (eq 1)32 Ks

β-NBNf + aggregate T β-NBNb-aggregate

(1)

where subscripts f and b, refer to free an bound probes, respectively. Magnitudes of the association constant (Ks) were obtained by data linearization using the standard double reciprocal equation (eq 2) (Figure 2, insert)

1 1 1 )1+ Fr Ks CD

(2)

where, Fr is the fraction of bound β-NBN and CD is the detergent concentration in micellar phase (CD ) Ct - cmc, and Ct is the amount of added surfactant). Association of β-NBN into aggregates is obviously well favored. In the presence of urea KS’s are slightly lower showing the higher solubility of the probe in the bulk solvent. The high affinity of β-NBN for the surfactant phase (pseudophase in the case of micelles) with or without urea allowed the use of fairly low surfactant concentrations to ensure complete partitioning of the probe into the supramolecular aggregates. A condition followed here. In Figure 3 the effect of anions on the fluorescence emission yields of bound β-NBN are presented. Figure 3a shows the quenching effect of Br- on the emission of β-NBN in HPS micelles (water and urea) having 1 M NaCl. It can be observed that intensity ratios (I0/I, fluorescence intensities in the absence and presence of added Br-, respectively) increase steadily for both aqueous and urea/ aqueous solutions, leveling off for the urea/aqueous media at about 1.2 M Br- and continues to increase when in absence of urea. Figure 3b shows the effect for CTACl (32) Connors, K. A. Binding Constants. The Measurement of Molecular Complex Stability; John Wiley: New York, 1987.

+

Effect of Urea on Surfactant Aggregates

+

Langmuir, Vol. 12, No. 5, 1996 1169

Figure 4. Decarboxylation rate constant of NBOC (kobs) as a function of CTABr concentration: (O) in absence of urea; (b) in presence of 5 M urea. Lines represent best fits obtained with PPIEM formalism (see text).

Figure 3. (a) Stern-Volmer projection for β-NBN (4.14 × 10-6 M) incorporated in 10 mM HPS micelles with 1 M NaCl in absence (O) and presence of 3 M urea (b) as a function of [NaBr]. (b) Fluorescence suppression ratios I0/I of β-NBN (4.14 × 10-6 M) incorporated in 40 mM CTACl micelles in absence (O) and presence (b) of 3 M urea as a function of [NaBr], solid lines are from data simulation according to eq 3 (see text). (c) Same as part b for 7 mM HERQUAT small vesicles in the absence (O) and presence (b) of 3 M urea as a function of [NaBr]. T ) 30 °C, λexc ) 330 nm.

micelles where the same pattern is observed with the plateau region for urea/aqueous system being observed at ca. 70 mM Br-. Figure 3c shows intensity ratios for β-NBN in HERQUAT vesicles as a function of [Br-]. In the absence of urea a stepper increase is observed initially and maxima are observed for both media at ∼10 mM NaBr. The observed decrease in I0/I thereafter is due to vesicle rupture and precipitation with Br- and therefore data in this region are unreliable.

The slow unimolecular decarboxylation of NBOC (kobs ∼ 5 × 10-6 s-1 in H2O at 300 K33) can be employed to investigate the long time averaged ionic atmosphere around charged surfactant aggregates.34 NBOC is a relatively hydrophilic reactive anion whose decarboxylation rate constant is very sensitive to medium polarity, accordingly reaction rate constants increase by up to a hundredfold in going from bulk water to CTABr micelle/ water interface.35 Measurements of kobs in water and in the presence of 5 M urea reproduced nicely previous determinations33 and demonstrating the lack of effect of urea on the reaction in bulk aqueous media. Effect of CTABr on the reaction is presented in Figure 4, for ureafree solution kobs levels off at approximately 20 mM CTABr. In the presence of 5 M urea the plateau is lately attained at ∼40 mM CTABr (notice that within experimental error the same plateau is obtained), clearly more CTABr is necessary for NBOC to fully partition in the micelle palisade. The effect of urea on vesicle stability was monitored by measuring vesicles Rh and MW as a function of time. For these measurements samples were kept in clean capped vials and only the external vial’s surfaces were cleaned before each measurement. Rh’s versus time show that preparations with or without urea are quite stable (Figure 5). In a time range of 40 days, Rh increased by 20% for urea-free preparation and 50% and 70% with 3 and 5 M urea, respectively. Determined Rh’s in the range investigated (6.0 × 10-4 to 3.5 × 10-3 M) were independent of sample dilution, pointing to a regime of nonelectrostatic intervesicle volume exclusion. The effect on MW’s also shows an increase on MW’s with urea concentration (Table 3). In the absence of urea and in the presence of 3 M urea, increases in MW (and in the calculated mean aggregation number) are 54% and 67%, respectively. With 5 M urea the increase is only 42%, these data however suffer from the poor dn/dc determination (see methods section) and are therefore less reliable. Overall its clear that with increasing urea concentration, vesicle growth process is enhanced. (33) Kemp, D. S.; Paul, K. G. J. Am. Chem. Soc. 1975, 97, 7305. (34) Silva, I. N. Master Dissertation Thesis submitted to Instituto de Quı´mica, USP, Brazil, 1995. (35) Bunton, C. A.; Minch, M. J. Tetrahedron Lett. 1970, 44, 3881.

+

+

1170

Langmuir, Vol. 12, No. 5, 1996

Figure 5. Hydrodynamic radii of HERQUAT small vesicles as a function of incubation time (T ) 25 °C): (b) HERQUAT ) 0.814 mM with 2 mM NaCl; (1) HERQUAT ) 1.092 mM in 3 M urea with 2 mM NaCl; (9) HERQUAT ) 0.812 mM in 5 M urea with 1 mM NaCl. Table 3. HERQUAT Vesicles MW and Aggregation Number as a Function of Incubation Time and Urea Concentration

time (days) 0 1 7 13 20 34

molecular weightc × 10-6 g/mol 3M 5M watera ureaa ureab 4.02 4.93 6.50 7.03 8.54 8.66

3.65 4.59 7.37 8.30 10.30 11.25

mean aggregation numberc × 10-3 (water/3 M urea/5 M urea)

6.25 8.73 14.87 12.19 13.15 10.71

6.92/6.28/10.77 8.49/7.90/15.03 11.20/12.70/25.60 12.11/14.29/21.00 14.71/17.74/22.64 14.92/19.36/18.44

a Vesicle preparation with 2 mM NaCl. b Vesicle preparation with 1 mM NaCl. c Accurancy for light scattering determinations is about 5%.

Discussion The interface of macromolecular aggregates is the site of action. Effects upon chemical reactivity and on several others physical-chemical processes36 originated from the peculiarities of the interfacial microenvironment. In this region properties of the bulk media (water) change drastically to a molten of salt in ethanol-like polarity media.36 Knowledge of the various factors which contribute to this dynamic state is of course of high importance for viewpoints going from understanding the behavior of live matter where interfacial phenomena are abundant to pure technological ones like those of catalyst design. Moreover the ability to modulate the state of this compartment is more than that of simply understanding but to act at the microscopic level to manipulate conditions as our wish. In this context the investigation on the effect of “inert” additives like urea can help to establish interfacial modulators. The biomimetic aggregates here investigated are well characterized. Parameters as ion dissociation degree (R) for the ionic surfactants (CTAX and Herquat) and ion condensation for the neutral (HPS)37,38 are well established. For these systems anion selectivity has been shown to follow the usual affinity order Br > Cl > F ) OH- > acetate, which is related to the anion polarizability.31 In (36) Fendler, J. H. Membrane Mimetic Chemistry; John Wiley and Sons: New York, 1982. (37) Sepu´lveda, L.; Corte´s, J. J. Phys. Chem. 1985, 89, 5322. (38) Cuccovia, I. M.; et al. J. Phys. Chem. 1990, 94, 3722. (39) Bunton, C. A.; Mhala, M. M.; Moffatt, J. R. J. Phys. Chem. 1990, 93, 6781.

Florenzano et al.

the case of neutral interfaces as for HPS (dipole moment pointing outward) the first step toward selectivity is the distinct affinity between cations and anions (anions favored) which by its turn generates an asymmetric electrolyte condensation resulting in a locally positively charged interface similar to an actual cationic interface.38 The main part of results here presented is based on the special properties of the fluorometric probe (Chart 1). It was shown before that β-NBN binds favorably in surfactant aggregates and that from the distinct quenching effect by Br- and Cl-, access to their interfacial relative concentrations was feasible.23 Binding of β-NBN into HPS, CTACl, and HERQUAT aggregates in the presence or absence of urea is again well favored (Table 2). Parallelly, the distinct quenching efficiencies in bulk solution by Brand Cl- with or without urea ensures the applicability of our methodology (Figure 1). Addition of urea to the surfactant media results in a very distinct behavior of the interfacial quenching competition between Br- and Cl-. Starting with HPS in the presence of 1 M NaCl, to ensure close to complete Clcondensation in/on the interface, I0/I curve in the lack of urea increases steadily suggesting that Br- keeps exchanging favorably Cl- from the surface as demonstrated previously23 (Figure 3a). In the presence of urea however I0/I reaches a plateau indicating that no matter how large the Br- concentration (within the physical allowed limits) there is no differential competition between them. It is important to notice that in our previous study with HPS micelles23 it was shown that the capability of electrolyte condensation maxima is about 2 M. Thus the observed initial increase reflects simply the entrance of Br- in the dipolar headgroup compartment (Figure 3a). The distribution of electrolytes in ionic micelles have been adequately described by the PPIEM formalism.40 With this model and using established values for the dissociation degree (R) it is straightforward to simulate I0/I curves as shown for CTACl/Br ion exchange in the absence of urea (Figure 3b, O). The exchange between Br- and Cl- and the resulting decrease in the fluorescence emission is attributed to the increase in the micelle coverage (σ) by Br- according to eq 341

σBr/σCl )

(I0Cl/IBr,Cl) - 1 (I0Cl/I0Br) - (I0Cl/IBr,Cl)

(3)

where I0Cl and IBr,Cl are the fluorescence intensities of β-NBN associated with CTACl in the absence and presence of Br-, respectively, and σBr and σCl are the coverage degrees for Br- and Cl-, respectively. σCl and σBr can be determined by the following PPIEM mass balance equations40

σCl ) Clb/CD ) σ0Cl - σBr ) σ0Cl/[1 + (σBr/σCl)] (4) σBr ) Brb/CD ) σ0Cl(σBr/σCl)/[1 + (σBr/σCl)]

(5)

{-A1 + [(A1)2 + 4(1 - KBr/Cl)BradKBr/Cl(1 - R)CD]1/2} )

[2CD(1 - KBr/Cl)]

(6)

where σ0Cl is the Cl- coverage in absence of Br- added (Brad), Clb and Brb are the concentration of the specific counterion bound to the interface, KBr/Cl is the exchange constant between Br- and Cl- (KBr/Cl ) Br-b‚Cl-f/Br-f‚Cl-b), and (40) Quina, F. H.; Chaimovich, H. J. Phys. Chem. 1979, 83, 1844. (41) Abuin, E.; et al. J. Phys. Chem. 1983, 87, 5166.

+

+

Effect of Urea on Surfactant Aggregates

A1 ) RCD + cmc + KBr/Cl Brad + KBr/Cl(1 - R)CD (7) where CD is the concentration of micellized surfactant (CD ) CT - cmc). Letting R ) 0.28 (Table 1)42 and using the experimental value of I0Cl/I0Br ()14.75), best fit is achieved with KBr/Cl ) 4.6 ( 0.6 (Figure 3b), a value in high concordance with previous calculations.41 Using this approach for data in the presence of urea with I0Cl/I0Br ()12.8) and letting R ) 0.35,42 nonlinear regression minimization occurred with KBr/Cl ) 1.1 ( 0.5. pointing to the loss of ion selectivity by addition of urea, in agreement with that observed for quaternary ammonium monolayers.17 It is very interesting to notice that the effect of urea on the ionic interfacial competition for the three systems investigated (Figure 3) has the decreasing order of HERQUAT > CTACl > HPS or, in other words, the effect of urea being more pronounced as R is lower (or equivalent parameter for HPS). Analysis of NBOC decarboxylation data within the same framework produces also a nice fit for both aqueous and urea/water results (Figure 4). For fitting the data in absence of urea, the following parameters were employed: kwater ) 5.0 × 10-6 s-1, kmicelle ) 6.6 × 10-4 s-1 (where kwater and kmicelle are the rate constants in absence of added surfactant and for fully associated probe), R ) 0.24 (Table 1), and cmc ) 4.19 × 10-4 M (Table 1). In the presence of urea, parameters were kwater and kmicelle as above, and cmc ) 1.25 × 10-3 M (Table 1). In bulk water the best fitted exchange constant between NBOC/Br (KNBOC/Br) is equal to 20, confirming the very high affinity of NBOC for the interface. In the presence of urea and using R ) 0.28 (Table 1) best fitted KNBOC/Br is equal to 11. The decrease found for the exchange constant shows from the viewpoint of a slow process (hours) the same behavior as that for a very fast one (fluorescence suppression, nanoseconds), that is loss of ion selectivity by addition of urea. It is important to observe that the high value of KNBOC/Br (with or without urea) shows the existence of a hydrophobic component in the partition of NBOC; however for the sake of the present analysis it is not necessary to further deconvolve these values in order to differentiate between electrostatic and non-Coulombic contributions. The decrease in the exchange constant due to urea undoubtedly confirms that sorption of urea at the interface or equivalently that headgroup solvation by the additive increases the surface charge density (increase in R). The suggested hypothesis of collapse of the Stern layer due to replacement of water by urea proposed to occur with quaternary ammonium monolayers17 could also be operative with the surfactant aggregates presently investigated. The increase in Rh and MW of HERQUAT vesicles with urea (Figure 5 and Table 3) can also be rationalized within the viewpoint of its direct interaction with the aggregate. Once Rh’s in the investigated range are dilution independent, increase in vesicle size parameters derive from growth processes of single particles. Although its mechanism is presently unknown, it is probable that a intervesicle monomer exchange process is enhanced. The increase in the monomer exchange rate can arise from higher monomer solubilization in the bulk phase and also from decreased shell stability. Preferential solvation of the bilayer by urea resulted in a higher dissociation degree (Figure 3c). The increase in R is necessarily accompanied by the increase in the monomer area, as found for HERQUAT-like monolayers.17 This effect changes the (42) Once R values vary with the relative amount of Cl- and Br- in solution, R was simply taken as the arithmetic mean between CTACl and CTABr values.

Langmuir, Vol. 12, No. 5, 1996 1171

monomer packing parameters from a truncated trapezoid (small headgroup area compared to its tail) to an inverse trapezoid (larger headgroup area compared to the hydrocarbon cross sectional area) resulting in the bilayer decreased meta-stability. In other words the increase in the headgroup area and the resulting geometrical constraint forces the bilayer toward larger sizes. Given that growth resulting from vesicle fusion is unlikely to occur due to the intervesicle electrostatic repulsion, monomer interchange via bulk media is more probable. Although within our experimental limits HERQUAT did not present enhanced solubilization in presence of urea (up to 5 M), a solubility increase by 2- or 3-fold (still below experimental detection limits) can provide the mechanism for the decreased vesicle stability in the presence of urea. Conclusion The effects of urea on thermodynamic, kinetic, and metastability properties of surfactant aggregates were presently investigated. cmc determinations for cationic and zwitterionic micellar aggregates in the presence of urea revealed the usual increase in the transition region, the effect being higher for the zwitterionic ones. Derived micelle dissociation degree for the cationic micelle also increases. Kinetic data ranging from nanoseconds to hours of probes fully associate with the interface show a decrease in the anion distinct interfacial affinity order well established to occur with these biomimetic models. Provided that for either probes bulk effects of urea are either absent (thermal reaction) or changed equivalently with respect to bulk water (fluorescence suppression), results are compelling toward the direct effect of the additive. In other words, from the view point of probes residing close or at the interface region, addition of urea leads to changes in the local aggregate counterion distribution. For the urea-free interface the high bromide affinity (compared with chloride) with the interfacial ammonium headgroups has been associated with its high polarizability. In the presence of urea, which has similar polarity to that of water, the decreased affinity can be interpreted to derive from the preferential headgroup solvation by urea and the squeezing out of nearby anions. This reasoning is equivalent to the effect of reducing, by a few angstrons, the Stern Layer region. The increase in the surface charge density (increase in R) and, more important, the decrease in anion selectivity presently found, point to the direct mechanism of urea interaction. This conclusion stems from the fact that structural changes in bulk water (indirect mechanism) should not affect local interfacial properties, like ion polarizability, which governs ion association with the interfaces presently investigated. Since properties of electrolyte solutions in the studied range barely change from bulk water to water/urea solutions, the direct interaction effect due to the additive is the plausible interpretation. With the metastable aggregates, headgroup solvation by urea, besides pointing to the increased extent of counterion dissociation and reduced anion selectivity, results in their decreased stability. It is tempting to associate vesicle transformation into larger aggregates with urea as a result from the geometrical constraint imposed in the bilayer by the enlargement of the surface area per monomer headgroup. More detailed studies with these aggregates are necessary however to check this hypothesis. Acknowledgment. We wish to deeply express our gratitude to the Brazilian granting agencies FAPESP, CNPq, and FINEP for their financial support. LA9505834