Biochimica et Biophysica Acta 1394 (1998) 99^110

Micelle formation of sodium cholate and solubilization into the micelle

Abstract The micellization of sodium cholate (NaC) was studied at 298.2 K by aqueous solubility at different pH values. Using a stepwise association model of cholate anions without the sodium counterion, the aggregation number (n) of the cholate micelle was evaluated and found to increase with the total concentration, indicating that the mass action model worked quite well. The n value at 60 mM was found equal to 16. The membrane potential measurement of sodium ion with a cation exchange membrane was made in order to confirm the low counterion binding to micelle. The solubilization of alkylbenzenes (benzene, toluene, ethylbenzene, n-propylbenzene, n-butylbenzene, n-pentylbenzene, n-hexylbenzene) and polycyclic aromatic compounds (naphthalene, anthracene, pyrene) into the aqueous micellar solution of sodium cholate was carried out. Solubilizate concentrations at equilibrium were determined spectrophotometrically at 298.2 K. The first stepwise association constants (K 1 ) between solubilizate monomer and vacant micelle were evaluated from the equilibrium concentrations and found to increase with increasing hydrophobicity of the solubilizate molecules. From the Gibbs energy change for solubilization at the different mean aggregation numbers and from molecular structure of the solubilizates, the function of sodium cholate micelle for solubilization was discussed and was compared with data from conventional aliphatic micelles. 1998 Elsevier Science B.V. All rights reserved.
Keywords: Sodium cholate; Micellization ; Aqueous solubility ; Aggregation number; Alkylbenzene; Polycyclic aromatic compound; Solubilization

1. Introduction Bile salts, having a large, rigid, and planar hydrophobic moiety of a steroid nucleus with two or three hydroxyl groups, are a special group of biosurfactants, whose properties dier considerably from ordinary aliphatic surfactant molecules [1]. They act as solubilizer and emulsier for cholesterol and lipids in the intestines. Because of the interesting physiological functions, micellar properties of bile salts have

* Corresponding author. Fax: +81 (92) 642-2607; E-mail: [email protected]

been extensively studied by various methods [2^16], and many review articles on physicochemical properties of bile have appeared [17]. Bile salts have lower aggregation number compared with conventional aliphatic surfactants. For the sodium cholate micelle, lower values of 4 [2,3], and higher values of 16 [4,5], have been reported from dierent measurements. The aggregation model suggested by Small was the primary^secondary micelle model. In this model, the primary micelles are formed in such a way that the hydrocarbon backs of the steroid nucleus associate. The secondary micelles are then formed by the aggregation of these primary micelles [1]. This model invokes a stepwise aggrega-

0005-2760 / 98 / $ ^ see front matter 1998 Elsevier Science B.V. All rights reserved. PII: S 0 0 0 5 - 2 7 6 0 ( 9 8 ) 0 0 0 9 0 - 3

100

H. Sugioka, Y. Moroi / Biochimica et Biophysica Acta 1394 (1998) 99^110

tion mechanism, i.e. polydispersity in the aggregates, in which the critical micelle concentration (CMC) appears not as a point but over a certain range. This fact was supported by other experimental data derived from surface tension [6] and uorescence [7] measurements. However, micelle formation takes place in the bulk. In this sense, surface tension, by which the state of surfactant molecules on the surface is observed, is essentially an unsuitable method for investigating physicochemical properties of micelles in the bulk. Although it can be used, just for CMC determination. The uorescence probe method is convenient to study molecular aggregation in a bulk, but the method is not totally recommended in the sense that some other chemicals need to be added in the system. On the other hand, the present solubility method depends just on thermodynamics without any other chemicals. In the stepwise aggregation model based on thermodynamics, the aggregation number should increase with amphiphile concentration. However, very few publications give information on this very important point. The enhanced aqueous solubility of otherwise slightly soluble organic substances brought about by the presence of surfactant micelles is well known as solubilization [18,19]. This process, caused by incorporation of hydrophobic organic substances into micelles, plays a very important role not only in industrial processes, but also in biological processes [20], such as adsorption and transfer of materials in living tissues. A function of bile salts is to solubilize sparingly soluble substances such as cholesterol and bilirubin into bile. The solubilization of bile salts has been extensively studied; for example, the solubilization of phospholipid vesicles [21], and of aromatics [8,22^26] and the solubilization into mixed micelles [27]. Most of the studies were carried out under various conditions in the presence of added salts in buffer solution. However, few basic and systematic studies have been carried out. Mukerjee and Cardinal have reported on the solubilization of naphthalene into sodium cholate micelles [22]. According to them, if the CMC is estimated by the usual procedure of nding the intercept from linear extrapolation of the data below and above the CMC, very

dierent CMC values are obtained. The CMC depends upon which sections of the data are used to derive the interpolated trace. Therefore, it is quite dicult to precisely determine the CMC value of bile salts. Hydrophobic solutes with a rigid structure, such as bile salts are expected to have rather an extended concentration range over which the molecular aggregation number increases from low to high values. The physicochemical studies of bile salts, on the one hand, and their physiological function as solubilizer on the other hand, make it clear that their behavior in vitro and their functions in vivo are closely related. In this sense, thermodynamic research on the solubilization by bile salts is quite important. Unfortunately, solubilization has been treated in most cases as a partitioning of solubilizate molecules between a micellar phase and an intermicellar bulk phase [28^31], which is inconsistent with the phase rule as pointed out by one of the authors [32,33]. If the micelles are regarded as a phase, the number of degrees of freedom in the presence of a coexisting solubilizate phase is two. This would mean that the surfactant concentration could not be changed for a given temperature and pressure. However, the maximum additive concentration changes with total surfactant concentration. So, it is preferable to analyze solubilization data in terms of the thermodynamic parameters of solubilization, which regards a micelle as a chemical species [34,35]. In this paper, a proven method based on aqueous solubility was applied for investigating the micellization of sodium cholate in water at 298.2 K. This thermodynamic method is very simple and useful for aggregates whose aggregation number is relatively small and increases with concentration. The aim of the present study is to clarify the dependence of aggregation number of cholate micelle on the concentration using the method mentioned above. The second objective of this study is to investigate the solubilization of alkylbenzenes and polycyclic aromatic compounds into sodium cholate micelles at 298.2 K by the mass action model. This work serves to clarify the eect of molecular structure of solubilizate and surfactant on the thermodynamic parameters of solubilization and the characteristics of solubilization by sodium cholate.

H. Sugioka, Y. Moroi / Biochimica et Biophysica Acta 1394 (1998) 99^110

101

2. Materials and methods 2.1. Materials Cholic acid (CA) was of guaranteed reagent grade (Nacalai Tesque). The cholic acid was puried by foam fractionation [36,37] of the sodium salt solution around the critical micellar concentration, where more than 15% of the original solution was removed with foam. After foam fractionation, the residual solution was acidied with dilute HCl solution to precipitate the cholic acid from the solution. The precipitate was washed with large amounts of doubly distilled water and dried over P2 O5 in a desiccator under reduced pressure. The purity was checked by elemental analysis, and the observed and calculated values were in satisfactory agreement CAW1/2H2 O: C 68.83 (69.02), H 9.91 (9.90)%, where the values in parentheses are the calculated ones. The stock solution of NaC was prepared by stoichiometric neutralization of cholic acid by sodium hydroxide solution and was diluted to a nal volume with doubly distilled water. The solution was ltered through a 0.2 Wm membrane lter (Millipore FGLP01300) to remove dust from the solution. Cholate solutions for analytical use were prepared by volumetric dilution of this stock solution. Benzene, of guaranteed reagent grade (Nacalai Tesque), was used as received. Toluene and ethylbenzene of guaranteed reagent grade (Nacalai Tesque) were washed four times with concentrated sulfuric acid, four times with diluted sodium hydroxide solution alternately, and nally three times with water. n-Propylbenzene, n-butylbenzene, n-pentylbenzene, and n-hexylbenzene (Tokyo Kasei Kogyo) were distilled once and rinsed with a large amount of water 10 times [38]. Naphthalene, anthracene of guaranteed reagent grade (Nacalai Tesque), and pyrene of similar grade (Tokyo Kasei Kogyo) were puried by repeated recrystallizations from the ethanol solution [39]. The purity of these compounds was checked by elemental analysis. Ethanol of 99.5 vol% (Nacalai Tesque) was used without further purication. Water for use was distilled twice from alkaline permanganate solution.

2.2. Relationship between concentration and pH The cholic acid was dispersed in 15 cm3 of doubly distilled water in a glass vessel. The suspension of solid cholic acid was agitated with a rotor while thermostatically controlled at 298.2 K. The pH of the suspension was monitored with a pH electrode (Horiba), during the stepwise introduction of aliquots of a dilute solution of sodium hydroxide. The total acid concentration (Ct ) was evaluated from the titration with an added amount of sodium hydroxide [40]. The neutralization reaction of the acid with sodium hydroxide is very rapid, and therefore, the concentration of cholate anion should be equal to Na concentration, even when aqueous solubility of undissociated cholic acid would be below the saturation limit. 2.3. Membrane potential measurement The measurements were made for sodium ion with a cation-exchange membrane (Tokuyama, NeoceptaCM1) that had been dipped in 1 mM NaC solution for a few days in order to replace all the exchangeable ions with sodium ion. The membrane potential at 298.2 K was measured during the stepwise introduction of aliquots of a NaC solution into the sample solution, using the following cell:
reference electrode (calomel) M agar bridge M reference solution (C0 = 5 mM NaC) M cation-exchange membrane M sample solution (Ct mM NaC) M agar bridge M reference electrode (calomel)

where Ct denotes the varying NaC concentration. The electromotive force (vE) of the cell was measured with a precision of 0.5 mV by using Advantest digital multimeter (model R6441B). 2.4. Solubilization 2.4.1. Alkylbenzenes [38] A simple glass vessel having eight tubes (Fig. 1) was used as an apparatus for the solubilization of volatile or gaseous substances. Eight surfactant solutions of dierent concentrations were poured separately into eight tubes of the glass vessel; the cholate

102

H. Sugioka, Y. Moroi / Biochimica et Biophysica Acta 1394 (1998) 99^110

solubility of 5.98U1036 mol dm33 . However, the lower detection limit of the spectrophotometer is accurate to 0.005. Therefore, we have used the extrapolated solubility value from the linear relationship between the logarithm of aqueous solubility and the carbon number of alkylchain of the solubilizates. 2.4.2. Polycyclic aromatic compounds [39,41] Four cm3 of sodium cholate solution and a small amount of powdered solubilizate sucient to produce a saturated solution were put together into a 10 cm3 injector tube (Fig. 2). The solutions were stirred for about 24 h until equilibrium was reached at 298.2 K, where the temperature was controlled to within 0.005 K. Filtration of the excess solid was performed through a membrane lter of 0.2 Wm pore-size by applying pressure upon the injector. The ltrate was diluted by adding further surfactant solution so that the absorbance of the diluted solution becomes less than 0.7. The absorbance of the sample was measured spectrophotometrically, and the maximum additive concentration (MAC) was determined using the respective molar extinction coefcient given in Table 1. 3. Results 3.1. Micellization Fig. 3 shows the relationship between the total acid concentration of cholic acid, Ct , and pH. Ct is proportional to 1=aH below the CMC of NaC [36]

Fig. 1. Solubilization apparatus for volatile solubilizates. Eight surfactant solutions of dierent concentrations are poured separately into eight tubes. Inside the apparatus, the volatile solubilizates easily evaporate, and the chemical potential of the gaseous solubilizate molecules becomes constant throughout all the phases.

concentration in the rst four tubes was below the CMC and the last four were above the CMC of sodium cholate. A minute amount of solubilizate was placed on the hollow in the middle of the eight ngers. The glass vessel was set into a thermostat controlled at 298.2 K, while each of the surfactant solutions was stirred for 24 h. After equilibration, each surfactant solution was separately drawn into each injection tube through an injection needle, and the tubes were immediately capped with a silicone rubber. The optical absorbance of each solution was measured with an ultraviolet spectrophotometer (Hitachi model 100-50) in order to determine the equilibrium solubilizate concentrations from the molar extinction coecient. For n-pentylbenzene and nhexylbenzene, approximately 1 cm3 of the liquid was placed on the hollow, because the aqueous solubility is very small and their absorbance below the CMC cannot be observed unless their concentrations reach the maximum additive concentration. However, for n-hexylbenzene, the aqueous solubility was below the detection limit of spectrophotometer. The optical density of n-hexylbenzene in pure water was calculated to be ca. 0.001 using the molar extinction coecient of 150 mol31 dm3 cm31 and the aqueous

Fig. 2. Solubilization apparatus for solid solubilizates.

H. Sugioka, Y. Moroi / Biochimica et Biophysica Acta 1394 (1998) 99^110

103

Fig. 3. Relationship between total cholate concentration and aqueous solution pH at 298.2 K. Monomer concentration at 40 mM is determined by the way ACBCC. Point A, Ct = 40 mM; point B, the value on the straight line at the same abscissa value of point A; point C, the ordinate value [S3 ] of point B.

Fig. 4. Variation of membrane potential change (vE) with sodium cholate concentration at 298.2 K. A NaC solution of 5 mM below the CMC is used as reference, and NaC concentration is measured over the range 5^60 mM. The CMC given in the gure is 12 mM from reference [37].

and departs from a straight line when molecular aggregation starts [40]. Fig. 4 displays the relationship between vE and sodium cholate concentration. The vE values decrease linearly over the whole concentration range 5^60 mM of NaC. All experimental points fall on a straight line over the whole concentration range
Table 1 Molar extinction coecients (O) of solubilizates Solubilizate Benzene Toluene Ethylbenzene n-Propylbenzene n-Butylbenzene n-Pentylbenzene n-Hexylbenzene Naphthalene Anthracene Pyrene
a

above and below the CMC. This strongly indicates that Na concentration increases linearly with sodium cholate concentration. 3.2. Solubilization The solubilization data for n-butylbenzene and an-

O/mol31 dm3 cm31 1.25U10 2.17U102 1.83U102 1.87U102 1.79U102 1.64U102 1.50U102 4.92U103 (5.24U103 )a 1.64U105 (1.49U105 )a 4.02U104 (4.63U104 )a
2

Wavelength/nm 253 260 260 260 260 260 260 276 (274) 254 (250) 273 (271) Fig. 5. Change in concentration of n-butylbenzene with sodium cholate concentration at 298.2 K.

The values in parentheses are those in pure water and are used for determination of solubilizate concentration in NaC solution below the CMC.

104

H. Sugioka, Y. Moroi / Biochimica et Biophysica Acta 1394 (1998) 99^110

celle formation expressed by concentration. The counterion binding to the micelle is assumed negligible, which is conrmed by the membrane potential measurement in this work. This simplication of the present analysis seems reasonable. Thus, the total analytical concentration or the total equivalent concentration of the molecules (Ct ) and the total molar concentration or the total concentration of osmotically active particles (i.e. monomers+micelles) (Cm ) become, Ct
n n X X iSi ibi Si i 1 i1 n n X X Si b i Si i1 i 1

Cm
Fig. 6. Change in concentration of anthracene with sodium cholate concentration at 298.2 K.

where [S] is the monomer concentration. Therefore, the following relation holds: dC m =dS
n n X X ibi Si31 ibi Si =S C t =S i 1 i 1

thracene appear to be best represented by a smooth curve, as Figs. 5 and 6 show respectively. Solubilizate concentration remains almost constant for n-alkylbenzenes and naphthalene below the CMC, while increasing slowly for anthracene and pyrene. The aqueous solubility constants for the aromatics used for the thermodynamic analysis are from our previous data [39,41]. 4. Discussion 4.1. Micellization and micelle characterization 4.1.1. Theory The stepwise association of surfactant molecules (S) for micellization, neglecting the counterion binding to the micelle, can be expressed as 2SS2 3SS3 TTTTTT nSSn
bn b3 b2

and then Z S C t =SdS Cm

4 5 6

n C t 3S=C m 3S

where n is the mean micellar aggregation number.

1
Fig. 7. Change in monomer concentration of sodium cholate plotted against the total concentration at 298.2 K.

[32,42], where Ln is the equilibrium constant of mi-

H. Sugioka, Y. Moroi / Biochimica et Biophysica Acta 1394 (1998) 99^110

105

Eq. 5 means that the total molar concentration can be evaluated from the relationship between the total analytical concentration (Ct ) and the monomeric concentration ([S]). 4.1.2. Determination of mean aggregation number When an excess solid phase of a monobasic acid (HS) coexists with an aqueous phase, the total solubility or the total concentration (Ct ) is expressed as [43] C t HS S3 HS HSK a =cS3 =aH ; where the acidity constant (Ka ) is given by K a S3 cS3 aH =HS 9 and the activity of the undissociated species (HS) is assumed equal to its concentration because it is extremely low. From Eq. 8, Ct of cholic acid is proportional to 1=aH below the CMC of NaC. Although Ct departs from a straight line above the CMC (Fig. 3), the monomer concentration ([S3 ]) can be estimated from the solution pH by the linear extrapolation of the values below the CMC. This is only possible as long as the acid phase coexists in the system [36]. For example, if Ct = 40 mM, construct the broken line from point A for Ct = 40 mM on the curve to point B on the extrapolated line derived from the experimental concentrations below the CMC. The ordinate value (point C) of point B gives the sum of undissociated cholic acid concentration ([HS]) and monomer concentration ([S3 ]). The [HS] value is so low, 0.122 mM at 298.2 K [36], that it can be negligible for the Ct values. Therefore, the ordinate can be regarded as indicating the [S3 ] value. This was conrmed in a similar previous study [40]. In this study, however, undissociated cholic acid may disturb the micellization of cholate anions. Fortunately, the ratio of undissociated to dissociated acid is less than 0.01 at the CMC, as estimated from the aqueous solubility (0.122 mM) and the CMC value (12^15 mM). Such a small amount of undissociated species was shown to have little inuence on the CMC value in our previous study [40]. Sodium cholate micelles solubilize undissociated cholic acid in the form of acid^salt mixed micelles. To examine the extent of the solubilization of undis7 8

sociated cholic acid by the cholate micelle, the following gravimetric analysis was performed. Centrifuge tubes containing both NaOH solution of 60 mM and excess coexisting solid of CA were placed into a thermostat at 298.2 K for 24 h with continuous stirring. After centrifuging the suspension, the supernatant was further ltered with 0.2 Wm membrane lter (Millipore FGLP01300) in order to remove further small solid particles of CA completely. A nite volume of the ltrate was transferred to a weighing bottle, and was dried over P2 O5 in a desiccator under reduced pressure. The weight of the residue corresponding to the sum of dissolved NaC, undissociated CA, and CA solubilized into NaC micelles was measured. The solubilized CA was determined by subtracting the dissolved NaC and solubility of CA from the measured weight. The nal result is that the excess (solubilized) CA is 1.7% by weight of the calculated amount. This value means that the mean number of CA per micelle is only 0.27 at a concentration of 60 mM. Therefore, in this system, the small amount of solubilized undissociated acid is not expected to change the intrinsic micellar properties at all, as determined from this experimental fact and other experimental results from mixed micellizations [32,40]. The solubility below the CMC agreed with the calculated value within the experimental tolerances of a weighing balance.

Fig. 8. Plots of Ct /[S3 ] against [S3 ]. These plots are made in order to determine the molar concentration Cm . The ordinate represents the integrand of Eq. 5.

106

H. Sugioka, Y. Moroi / Biochimica et Biophysica Acta 1394 (1998) 99^110

The monomer concentration change with total acid concentration is shown in Fig. 7. Monomer concentrations increase quite slowly with Ct above the CMC. Plots of Ct /[S3 ] vs. [S3 ] for the numerical determination of the Cm values by Eq. 5 are given in Fig. 8. The molar concentration Cm can be calculated by the integration of Ct /[S3 ] with respect to [S3 ], the calculation was made graphically from the evaluation of the area below the curve in Fig. 8. Thus, the change in integration limits of Ct /[S3 ] with respect to [S3 ] enables us to determine the n value at arbitrary concentrations of Ct from Eq. 6. It is obvious that the n value below the CMC is one, as the above analysis suggests, whereas the value of Ct /[S3 ] starts to deviate upward from unity above the CMC. Thus, this thermodynamic method enables us to determine the aggregation number change with concentration by only pH measurement. This method is very suitable for aggregates like bile salts whose aggregation number is relatively small and increases with concentration. 4.1.3. Dependence of the mean aggregation number on concentration In Fig. 9 the n values are shown plotted against the total concentration. The n value at 60 mM is ca. 16. This value agrees with those by Lindman et al. [3] and Fontell [2]. On the other hand, several authors reported lower n values, 3^4 [4,5], but the dependence of aggregation number on the concentration has not been investigated in their studies. Anyhow, sodium cholate micelles have smaller aggregation number compared with conventional aliphatic surfactant micelles. The n value increases with the total surfactant concentration, as the curved solid line in Fig. 9 shows. This fact strongly indicates that the mass action model in the theoretical section, which regards micelles as a chemical species, is applicable. Lindman et al. estimated the aggregation number from the self-diusion coecient of solubilized decanol and gave information on the change of aggregation number with concentration [3]. Their results agree well with ours. The n values in their work are little larger than those found in this work, which seems to be due to solubilized decanol. On the other hand, according to Coello et al. [13], the dependence of the aggregation number on concentration is not observed over

the concentration range 0.0460 (about 45 mM) to 0.205 (about 203 mM) mol kg31 . Several investigators suggest that bile salts have three limiting association concentrations [8,22]: limit 1, 13^15 mM; limit 2, 45^50 mM; and limit 3, 90^ 110 mM. This is consistent with our data for sodium cholate. According to the primary^secondary micelle model proposed by Small [1], four molecules of bile salts are regarded as the basic unit of aggregation. Judging from the present results, the aggregation number of sodium cholate seems to increase stepwise: 8, 16, and so forth. At any rate, the proposed solubility method enables us to reasonably study micellization of systems whose aggregation number increases with total concentration without any assumption or any additive. This claim cannot be made from surface tension and uorescence methods. 4.1.4. Degree of counterion binding to micelle In the section of micellization theory, the counterion is assumed not to take part in micellization. To examine whether this assumption is correct or not, the membrane potential measurement for sodium ion with cation-exchange membrane was performed at 298.2 K. Ideally, the relation between membrane potential vE and activity of sodium ion aNa obeys the Nernst's equation,

Fig. 9. Dependence of the mean aggregation number on the total cholate concentration at 298.2 K. Mean aggregation number was evaluated from Eq. 6, using Ct , Cm , and [S3 ] values. Plots are tted to a smooth curve.

H. Sugioka, Y. Moroi / Biochimica et Biophysica Acta 1394 (1998) 99^110

107

vE 3RT =F lnaNa =a0 359:2 log aNa const:; 10 where F is the Faraday's constant and a0 is the activity of reference solution. In this study, vE values are plotted against the logarithm of the sodium cholate concentration. If counterion association to the micelle takes place as to Na ions in the bulk above the CMC, the straight line must deviate upwards from the CMC. However, several other papers have reported that this is not the case for the sodium cholate, where a low degree of counterion binding is observed [2,13,15]. This fact implies that the counterion Na is only slightly bound to the micelle, which indicates that counterion binding to the micelle can be neglected. This fact was also substantiated by a linear relationship between specic conductivity and the concentration of sodium cholate in solution. In addition, if the total concentration exceeds 60 mM in Fig. 3, the solid line starts to deviate towards the ordinate. This is an indication for the onset of counterion binding [40]. In this region, the above theory cannot be applied and a modied theory including the counterion binding should be used. In this paper, however, only a lower concentration range was analyzed in order to avoid additional complexity for the moment. 4.2. Solubilization of organic compounds 4.2.1. Theory The micellar aggregation number n is assumed to be constant, i.e. micelles are monodisperse. This is to avoid the mathematical diculties arising from treating them as polydisperse. Even in the case of the polydispersity, the present discussion remains essentially valid [34]. Micelle formation equilibrium between surfactant monomer (S) and micelles (M), whose mean aggregation number is n, is expressed as nSM
Kn

is less than a few times the micelle concentration, incorporation of the solubilizates into micelles is assumed not to change the intrinsic properties of the parent micelles. The stepwise association equilibrium between micelles and solubilizates (R) is M RMR1 MR1 RMR2 TTTTTTTTTTTTTTT MRm31 RMRm
Km K2 K1

12

where MRi is the micelle associated with i solubilizate molecules, K i is the stepwise association constant between MRi31 and a solubilizate monomer, and m is the maximum number of solubilizate molecules per micelle. When an excess solid phase coexists with a surfactant solution phase, the maximum additive concentration (MAC) is xed by specifying the total surfactant concentration for a given constant temperature and pressure. This arises from a degree of freedom of three as dened by the Gibbs' phase rule for this system. In the case for the alkylbenzenes, the chemical potential can be kept constant throughout all the phases, and then the K 1 value can be obtained in the same way as that for solid solubilizates. The assumption described above can lead to the following equations: K j K 1 =j Pi Ri exp3R=i! 13 14

where P(i) is the probability that a micelle is associated with i solubilizate molecules, R is the average number of solubilizate molecules per micelle. Eq. 14 is the Poisson distribution. From Eqs. 13 and 14, it follows that Mt M expK 1 R Rt R K 1 RM expK 1 R R Rt 3R=Mt K 1 R 15 16 17

11

where Kn is the equilibrium constant of micelle formation (Kn and M correspond respectively with Ln and Sn in the theory section of Section 4.1.1). When the equivalent concentration of solubilizate

where [Mt ] is the total micelle concentration, [Rt ] is the total equivalent concentration of solubilizate, [R]

108

H. Sugioka, Y. Moroi / Biochimica et Biophysica Acta 1394 (1998) 99^110 Table 2 The average micellar aggregation number (n), the average number of solubilizate per micelle (R), the association constant, and the Gibbs energy change of solubilization Solubilizate Benzene n 8.6 10 13 16 8.6 10 13 16 8.6 10 13 16 8.6 10 13 16 8.6 10 13 16 8.6 10 13 16 8.6 10 13 16 8.6 10 13 16 8.6 10 13 16 8.6 10 13 16 R 0.026 0.033 0.047 0.065 0.12 0.15 0.21 0.25 0.20 0.26 0.39 0.51 0.22 0.28 0.39 0.58 0.23 0.34 0.55 0.72 0.61 0.84 1.3 1.7 0.44 0.64 0.93 1.2 0.42 0.56 0.78 1.0 0.0076 0.011 0.016 0.022 0.059 0.087 0.14 0.18 K 1 /mol31 dm3 6.2U101 7.7U101 1.1U102 1.4U102 3.4U102 4.2U102 5.7U102 7.1U102 7.7U102 1.0U103 1.6U103 2.0U103 2.4U103 3.1U103 4.8U103 6.7U103 8.1U103 1.2U104 1.8U104 2.5U104 2.6U104 3.6U104 5.4U104 7.0U104 7.3U104 1.1U105 1.6U105 2.0U105 2.0U103 2.7U103 3.7U103 4.8U103 3.6U104 5.2U104 7.7U104 1.1U105 1.5U105 2.2U105 3.5U105 4.7U105 vG0 /kJ mol31 310.2 310.8 311.6 312.3 314.5 315.0 315.7 316.3 316.5 317.2 318.2 318.9 319.2 319.9 321.0 321.8 322.3 323.3 324.4 325.1 325.2 326.0 327.0 327.7 318.9 328.8 329.6 330.2 318.9 319.6 320.4 321.1 326.0 326.9 327.9 328.7 329.6 330.5 331.7 332.4

is the monomeric solubilizate concentration not solubilized into the micelles. Eq. 17 can be rearranged as Rt 3R=R K 1 Mt 18 where Mt C t 3S=n and Ct is the total surfactant concentration. 4.2.3. Molar extinction coecient The molar extinction coecients (O) were determined from the optical absorbances of solubilizates at ve dierent concentrations in micellar solutions. The slope of the absorbance plotted against the solubilizate concentrations was used for the extinction coecient determination. For polycyclic aromatic compounds, the wavelengths of the absorption peaks above the CMC were dierent from those below the CMC, which indicates that the dielectric environments surrounding these solubilizates are dierent below and above the CMC. This observation supports the argument that these solubilizates are solubilized into NaC micelles above the CMC. The O values in pure water were determined by the same method as described above. The O values used for the present solubilizates are summarized in Table 1. 4.2.4. Solubilizate concentration The slight increase in concentration for anthracene and pyrene below the CMC is due to the presence of weak interactions between the solubilizate monomer and cholate monomer. The formation of micelles causes solubilizate concentration to increase comparatively rapidly above the CMC. This can be seen in the CMC value ranges from 12 to 15 mM judging from the data in micellization experiment and Figs. 8 and 9. The measured CMC value agrees with the reference data [1,17,37]. The nal solubilizate concentration in the cholate solutions for alkylbenzenes depends on the their amount placed inside the apparatus. Those for the aromatic compounds are the maximum additive concentration, which is a saturation concentration in the presence of a coexisting solid phase. Therefore, the latter depends on temperature, pressure, and the cholate concentration, as mentioned above. 4.2.5. Thermodynamic parameters We can determine the rst stepwise association constant K 1 at dierent Ct concentrations or at

Toluene

Ethylbenzene

n-Propylbenzene

n-Butylbenzene

n-Pentylbenzene

n-Hexylbenzene

Naphthalene

Anthracene

Pyrene

H. Sugioka, Y. Moroi / Biochimica et Biophysica Acta 1394 (1998) 99^110

109

Fig. 10. Gibbs energy change of solubilization for dierent mean aggregation numbers (n) versus carbon number of alkylchain in n-alkylbenzenes at 298.2 K: numbers in parentheses are total surfactant concentration in mM.

dierent n values from Eq. 18, because the relation between Ct and n, has already been determined. Table 2 shows the K 1 and R values. The K 1 values increase rather rapidly with the n values, i.e. the larger the n value becomes, the larger is the amount of solubilizate that can be stabilized by solubilization. The R values obtained for all solubilizates are small enough to conrm that the incorporation of the solubilizates can not change the intrinsic properties of the parent micelles and that the assumption is reasonable. The Gibbs energy change of solubilization can be expressed as vG 0 3RT ln K 1 19

against the number of carbon atoms. In both cases, the value of vG0 decreased linearly with the number of carbon atoms of the solubilizates, suggesting that the solubilizates were stabilized by incorporation into sodium cholate micelles as their hydrophobicity increased. Although all the solubilizates are stabilized by their incorporation into the cholate micelles with the vG0 values decreasing with increasing mean aggregation number, all the slopes remain constant: 33.0 kJ mol31 for n-alkylbenzenes, and 32.0 kJ mol31 for polycyclic aromatic compounds. This fact is quite characteristic of solubilization in sodium cholate micelles. In other words, the Gibbs energy change per methylene group usually decreases in magnitude with decreasing micellar aggregation number for the conventional aliphatic surfactants [44]. However, this appears not to be the case for sodium cholate micelles. From the above values, six carbons in an arene molecule or a benzene ring are equivalent to 4.0 methylene groups, as far as the Gibbs energy change is concerned. The values of the slopes can be compared with those for 1-dodecanesulfonic acid micelles whose aggregation number is 66 [38,39]: 32.81 kJ mol31 for n-alkylbenzenes and 31.77 kJ mol31 for polycyclic aromatic compounds. The above values strongly suggest that so-

Here, vG0 is made up of the three contributions:

20

where l0 MR1 is the standard chemical potential of 0 MR1 at innite dilution and l0 M and lR are the cor0 responding potentials. The vG values are also given in Table 2. In Figs. 10 and 11, respectively, the vG0 values for solubilization at dierent mean aggregation numbers of sodium cholate for n-alkylbenzenes and polycyclic aromatic compounds are plotted

Fig. 11. Gibbs energy change of solubilization for dierent mean aggregation numbers (n) versus total number of carbon atom in polycyclic aromatic compounds at 298.2 K: numbers in parentheses are total surfactant concentration in mM.

110

H. Sugioka, Y. Moroi / Biochimica et Biophysica Acta 1394 (1998) 99^110 Rauner (Eds.), Handbook of Physiology, Vol. 3, Am. Physiol. Soc., Waverly Press, 1989, pp. 621^662. M.L.E. McBain, E. Hutchinson, Solubilization and Related Phenomena, Academic Press, New York, 1995. P.H. Elworthy, A.T. Florence, C.B. Macfarlane, Solubilization by Surface-Active Agents and Its Application in Chemistry and Biological Sciences, Chapman and Hall, London, 1968. R.A. Kroc, R.L. Kroc, G.D. Whedon, W. Garey, Hepatology, The Physical Chemistry of Bile in Health and Disease, Vol. 4(5), Wiley, New York, 1984. G.A. Ramaldes, E. Fattal, F. Puisieux, M. Ollivon, Colloids Surf. 6 (1996) 363^371. P. Mukerjee, J.R. Cardinal, J. Pharm. Sci. 65 (1976) 882^ 886. E. Kolehmainen, J. Colloid Interface Sci. 127 (1989) 301^ 309. E. Kolehmainen, J. Colloid Interface Sci. 165 (1985) 273^ 277. J.R. Cardinal, Y. Chang, D.D. Ivanson, J. Pharm. Sci. 67 (1978) 854^856. C. Ju, C. Bohne, Photochem. Photobiol. 63 (1996) 60^67. E. Fudim-Levin, A. Bor, A. Kaplun, Y. Talmon, D. Lichtenberg, Biochim. Biophys. Acta 1259 (1995) 23^28. J.W. Larsen, L.J. Magid, J. Phys. Chem. 78 (1974) 834^839. C.A. Bunton, L. Sepulveda, J. Phys. Chem. 83 (1979) 680^ 683. C. Gamboa, A.F. Olea, Langmuir 9 (1993) 2066^2070. H. Uchiyama, E.E. Tucker, S.D. Christian, J.F. Scamehorn, J. Phys. Chem. 98 (1994) 1714^1718. Y. Moroi, Micelles, Theoretical and Applied Aspects, Plenum Press, New York, 1992, Chapt. 4 and 9. Y. Moroi, J. Phys. Chem. 84 (1980) 2186^2190. Y. Moroi, K. Sato, R. Matuura, J. Phys. Chem. 86 (1982) 2463^2468. Y. Moroi, H. Noma, R. Matuura, J. Phys. Chem. 87 (1983) 872^876. Y. Moroi, M. Kitagawa, H. Itoh, J. Lipid Res. 33 (1992) 49^ 53. P. Mukerjee, Y. Moroi, M. Murata, Y.S. Yang, Hepatology, The Physical Chemistry of Bile in Health and Disease, Vol. 4(5), Wiley, New York, 1984, 61S^65S. M. Take'uchi, Y. Moroi, Langmuir 11 (1995) 4719^4723. Y. Moroi, K. Mitsunobu, T. Morisue, Y. Kadobayashi, M. Sakai, J. Phys. Chem. 99 (1995) 2372^2376. Y. Moroi, Y. Sakamoto, J. Phys. Chem. 92 (1988) 5189^ 5192. T. Morisue, Y. Moroi, O. Shibata, J. Phys. Chem. 98 (1994) 12995^13000. P. Mukerjee, A.K. Ghosh, J. Am. Chem. Soc. 92 (1970) 6403^6407. Y. Moroi, R. Matuura, Anal. Chem. Acta 152 (1983) 239^ 245. Y. Moroi, R. Matuura, J. Colloid Interface Sci. 125 (1980) 456^462.

dium cholate micelles with their lower aggregation numbers are better solubilizers for both alkane and aromatic compounds than 1-dodecanesulfonic acid micelles, despite the latter having a longer hydrophobic alkylchain and larger aggregation number. The dierences between 33.0 and 32.8 kJ mol31 for the alkylbenzenes, and 32.0 and 31.8 kJ mol31 for the aromatics are quite noticeable. This strongly indicates that solubilization into sodium cholate micelle brings about larger stabilization for the solubilizates in spite of the small mean aggregation number. The increased stabilization suggests that inner hydrophobic region made of hydrocarbon backs of the steroid nucleus is more hydrophobic than inner micelle made of alkylchains of conventional aliphatic surfactants.

[18] [19]

[20]

[21] [22] [23] [24] [25]

References
[1] P.P. Nair, D. Kritchevsky, The Bile Acids; Chemistry, Physiology, and Metabolism, Vol. 1, Plenum Press, New York, 1971, Chapt. 8. [2] K. Fontell, Kolloid Z. Z. Polym. 244 (1971) 246^252. [3] B. Lindman, N. Kamenka, H. Fabre, J. Uimius, T. Wieloch, J. Colloid Interface Sci. 73 (1980) 556^565. [4] K. Fontell, Kolloid Z. Z. Polym. 246 (1971) 710^718. [5] R. Zana, D. Guveli, J. Phys. Chem. 89 (1985) 1687^1690. [6] C.J. O'connor, B.T. Ch'ng, R.G. Wallace, J. Colloid Interface Sci. 95 (1983) 410^419. [7] G. Li, L.B. McGown, J. Phys. Chem. 98 (1994) 13711^ 13719. [8] K. Fontell, Kolloid Z. Z. Polym. 250 (1972) 333^343. [9] J.P. Kratohvil, W.P. Hsu, M.A. Jacobs, T.M. Aminabhavi, Y. Mukunoki, Colloid Polymer Sci. 261 (1983) 781^785. [10] J.P. Kratohvil, H.T. Dellicolli, Can. J. Biochem. 46 (1968) 945^952. [11] J.P. Kratohvil, W.P. Hsu, D.I. Kwok, Langmuir 2 (1986) 256^258. [12] B. Lindman, N. Kamenka, B. Brun, J. Colloid Interface Sci. 56 (1976) 328^336. n [13] A. Coello, F. Meijide, E.R. Nu ez, J.V. Tato, J. Phys. Chem. 97 (1993) 10186^10191. [14] H. Kawamura, Y. Murata, T. Yamaguchi, H. Igimi, M. Tanaka, G. Sugihara, J.P. Kratohvil, J. Phys. Chem. 93 (1989) 3321^3326. [15] M.C. Carey, D.M. Small, J. Colloid Interface Sci. 31 (1969) 382^396. [16] J.M. Roe, B.W. Barry, J. Colloid Interface Sci. 107 (1985) 398^404. [17] D.J. Cabral, D.M. Small, in: S.G. Schultz, J.G. Forte, B.B.

[26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37]

[38] [39] [40] [41] [42] [43] [44]